ee es

aoa

ay

meetin
‘Seam

et
eater at

; gi eht a

¥
aket

Ate ns
Pe em

‘ cs
nes

$
4

* a
i

Se

1 ose tet ik:

pa Aon

\

“hte

x ‘ y
é * = '
. vi |
a engage),
J we 4) Re St ea
an ‘ if ST ADO Py. y
| A, WA ih
| Kt) \ « vaatpen edt

i
a be
t i,
We
= ~
i
ms
’
=o j
x ss
’
i
* : Ji
j ; a
-
I $4
}
+
H OP i eh
i ‘ , ie
t ;
: Tey
»s « 1
So
¥ i. My a
i
has f ives
~ ;
‘ _— b th
‘ ' y"
' Fy is r
a 1 Dh P
i beg 4
; “e Wo
j
i y
j
i *
? i : 7
hi
f ‘
y Bi s
Sih al PAE Ca Bk
a nk 3 : Lee ACL
| D ea es ly ae
<5 i ee is
*"e ks ae, 4 ni ah ay a ie
*4 n i ' ite bee
Q = ¢ - atid al
~~. "i 4 hévitg wat
. ’ Y Feist i. ' ; ¥
re an ; a. Fe eee i
> 1 ee DA Saas
" HEL aL det Ne
1 i y in fot fad i
* bits
7 y
f
»

¢~* ®
“ -
of
~
é
4 »
7 i
?
of
s
-

salad:
J fut?
wu. a
+ b| nas
iy 3

/
ers
R
‘
= -_
7 t
i -* }
A t
4 5 :
h
<4
0
, rin
war
‘oS
i a
4
‘ -
¥ y
i ‘
| ; “
‘ aS
ye 4
r ates
—*
| bs 4 n
* ‘Tae
4.5%
ie 4
“aar > f:
rrr j
iw? . Rall .
% | j
I .
dab aiG hi f
ay WA t ; Y
aac A j ;
1) fui i‘ . Ms vl i
at i \ ; LT ey a
Ware nied det vm Hii Lae |
; , n? i¢ ie vt, Wid fa J I; “g a '
Ea NL a
acs aires MW in MN Wik : “4 hy i ae y
, ro : ‘ , c . i E c i ‘
{ cere ‘il
a 4 , 1 rer
in ee Pie ORAT E «

1,4
4

' ;
7 .
@F ch \
; i
i] he i
~
‘
"
: wee
‘
‘i:
; La ' i
gy? J
a “G
2 >5, j
< ¥
¥ os 4 i
y ‘
y lw F ¥ we i)
Pie wy
a ‘ :
Ore i
= =
‘1 + righ é
4 ,
@ i) y
® r ‘ _ A ie i
f
‘
‘ +A
‘ LA. i {
ny dhs
, +?
~¢ 1
> \
*
» ae —
> : at y aon af
Pig i ae 1
iy rd Tal, e

7 ees i
oo a .
PS > totus
he >{ P ‘
> Ss © Wolte

_—-, _ yi
.

; y Ns ing
=a

ay a

Po (4

ae ‘

Oy :

a are os * y vt .
a {i j )
7 ? be be | /

;
a 4 a
: 7 i ; ’
‘Su _ , v ‘ n' : ‘
a j ry - ; et

ie ty
Ne nt

Age

4 ’7 — a
=
1? ay Ls -
' qe aah 9

wis
‘
7
re
‘
fay a,
of
7 ‘
5 ane
&
Y
|
; *
i)
f
e
ri
-
;

ue ye

—
‘“
at
aa 4 .
1
' * ' *
aa ’ a,
' f
i ; , i
ba? d }
AT
/ : ‘ * f
4 1
é
i
7
j
Pry i i
;
e A ' % ~
: =” 1
Fe pi ‘
< 2 ea ; ' :
at ~ “ iT ee i eat ee
bn me . i
’ ot }
aa iY ’
a? 4 ~ «
: é sae
' ” 4 %
y
: "
é é
i] Lé 7 1
wv s
vw) i 7 ~
-_ .
; j )
\f :
+
‘
*
n > ,
;
j
i
i
‘
’ 4 i
dy {
’
i
f) ‘
ae!
rH
’ ey :
rime g
ii é = be
v ‘ee ‘
‘ i
<
UL ~~
‘i 25 7 . tne hk.

im}

ne
ie
Ny
:

Mg

eh lag

d

act

= .
I
‘ f
'
7 te
‘A

e
.

$

ee ao leey
C oH \
Shy ae

REAPPRAISAL OF PROPHAETHON
HRUBSOLEI ANDREWS

_(NATURAL HISTORY)
Oe aa AV Ola tNOE

ae

A REAPPRAISAL OF PROPHAETHON
SHRUBSOLEI ANDREWS (AVES)

BY
COLIN JAMES OLIVER HARRISON
AND

CYRIL ALEXANDER WALKER

Pp. 1-30; 3 Plates; 7 Text-figures

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
GEOLOGY Vol. 27 No.1
LONDON : 1976

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

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

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

This paper is Vol. 27, No. 1, of the Geological
(Palaeontological) series. The abbreviated titles of
periodicals cited follow those of the World List of
Scientific Periodicals.

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

ISSN 0007-1471

© Trustees of the British Museum (Natural History), 1976

BRITISH MUSEUM (NATUBAL FISTORY)

Issued 4 February, 1976 Price 2.85

A REAPPRAISAL OF PROPHAETHON
SHRUBSOLEI ANDREWS (AVES)

By C. J. O. HARRISON & C. A. WALKER

CONTENTS
Page
MBSTRACE 0) ‘ : : ‘ : ; ; ; : 3
I. INTRODUCTION : : : : : : . : 5 3
II. DEscRIPTION : : : : : : : : : 4
a. Skull morphology . : ; ; ; : : : 4
b. Postcranial elements . : : . 4 ; : II
c. Measurements F ; : ; : 4 : é 16
III. COMPARISONS WITH RECENT MATERIAL : P : , 3 18
IV. DIscUSSION AND CONCLUSIONS : 2 - 5 : : 24
V. SYSTEMATIC DESCRIPTION : : : F : : 2 27
VI. ACKNOWLEDGEMENTS ; : ‘ ‘ - : , : 29
VII. REFERENCES : : : : i 3 : ‘ : 29
VIII. INDEX ‘ : : ; : : : ; : ; 29
ABSTRACT

Prophaethon shrubsole1 Andrews of the Lower Eocene, known from an imperfect skeleton, has
been previously assigned to the Phaethontidae within the Pelecaniformes. After further
preparation the specimen is redescribed. In comparison with Recent species of Pelecani-
formes, Charadriiformes and Procellariiformes it was found to share some characters with all
three, and could not with any certainty be assigned to any one of them. It is proposed, there-
fore, that Prvophaethon should be considered as a representative of a new monotypic order,
Prophaethontiformes, and it is suggested that it represents an early link between these Recent
orders. This hypothesis is discussed and a diagnosis of the new order is given.

fl. INTRODUCTION

In 1899 C. W. Andrews described a fossil bird which W. H. Shrubsole had collected
from the London Clay of the Isle of Sheppey, Kent, and had presented to the British
Museum (Natural History). When found, the specimen consisted of a clay nodule
in which one side of the skull and limb fragments were exposed. It was partially
prepared at the Museum and when described and figured it showed the entire dorsal,
posterior and right lateral views of the skull with the lower mandible in place. The
orbit was also partly cleared of matrix back to the quadrate and down to the quad-
ratojugal bar. The left side of the pelvis, which was somewhat obscured by the
overlying skull, was also prepared, and a femur and the anterior portion of a broken
tibiotarsus were laterally exposed.

From the general appearance of the skull Andrews concluded that it belonged to a
pelecaniform bird. In the subsequent discussion of its affinities only pelecaniform
families were referred to, and one may conclude that once he had decided on its
possible ordinal status he did not compare it with material other than that within
this taxon.

4 REAPPRAISAL OF

He compared the specimen systematically with examples of the various pelecani-
form families and concluded that it most nearly resembled the genus Phaethon which
constitutes the family Phaethontidae. In spite of some differences he thought that
it was ancestral to Phaethon and named it Prophaethon shrubsolet, regarding the
general similarities of the skull and deep fronto-nasal hinge with a prominent
frontal ridge as evidence of affinity.

He noted that the nostrils of Prophaethon, unlike the holorhinal ones of Phaethon,
approximated to the schizorhinal condition, but quoted Pycraft (1898) to the effect
that on the skull of the young of Phaethon the nostrils show a nearly schizorhinal
condition. He also noted that the pelvis was narrow and more closely resembled
that of Swla and Phalacrocorax. He regarded the nostril condition as ancestral and
concluded that the narrow pelvis indicated that, unlike the Recent Phaethon,
Prophaethon was probably a good swimmer and diver, and that the extent to which
Phaethon had diverged from this condition was a result of post-Eocene evolution.

During work towards a comprehensive review of the British Lower Eocene birds,
still in progress, we decided that the specimen was suitable for further preparation
by modern techniques. This was undertaken by Mr F. M. P. Howie of the Palae-
ontological Laboratory of the Museum. X-ray photographs were used to trace the
position and extent of the bones present, and the specimen was then carefully
prepared. Asa result of this work most of the elements have now been separated
and cleaned. The skull can be examined in all aspects including the palate (the
lack of access to which Andrews regretted). The lower jaw is now separate and
complete. A previously unsuspected portion of sternum has been revealed, and the
hidden part of the proximal end of the tibiotarsus is now free. The pelvis has been
extensively cleared, but still has attached to it the proximal half of the femur, and
also portions of the ribs. An almost complete coracoid and the blade of a scapula
are also present, and there are fragments of vertebrae and broken portions of limb-
bones partly embedded in more resistant matrix.

In the following sections the various parts have been described. The terminology
follows that of Jollie (1957) for the skull, and that of Howard (1929) for the post-
cranial skeleton. From the characters now available the species does not appear to
be closely allied to Phaethon and some characters are apparently shared with non-
pelecaniform taxa.

II. DESCRIPTION

a. Skull morphology

CRANIAL AND ORBITAL STRUCTURES. Viewed dorsally (Fig. 1; Pl. 1, fig. A) and
laterally (Fig. 2; Pl. 1, figs C & D) the skull as a whole shows a fairly even taper
from its widest part in the temporal region to its narrow termination at the tip of the
rostrum. The frontal rises a little from the dorsal edge of the parietals, with a slight
inflation on either side in the area above the brain. From near the posterior edge
of the orbits the frontal maintains a fairly even width anteriorly, with a shallow
median hollow in the interorbital region, and then expands gradually at the anterior

IPIR(QUAIAALIE IL lel (QIN SIEMR OURS OIL JBI. 5

Fic. 1. Dorsal viewof skull, x 1. Abbreviations: boc, basioccipital condyle ; bps, basi-
parasphenoid ; dps, gland depression, ? salt ; exn, external nares ; fm, foramen magnum ;
fnh, fronto-nasal hinge ; ios, interorbital septum; mx, maxilla; ns, nasal strut; pes,
prefrontal attachment area; pl, palatine; pmx, premaxilla; pop, postorbital process ;
q, quadrate ; rps, rostroparasphenoid ; tf, temporal fossa; z, zygoma.

end before terminating abruptly in a transverse, rounded and prominent brow
ridge just above the fronto-nasal hinge. From the more complete left side of the
frontal end of the specimen it would appear to be thick anteriorly, but above the
orbits the frontal becomes very thin and its outer edge may have been irregular.

The posterior dorsal edge of the orbit curves outwards and terminates in a promi-
nent, posteriorly curved postorbital process, which projects at a level only a little
below the dorsal surface of the frontal. Posterior to this process is a deep temporal
fossa. Viewed from above (Fig. 1; Pl. 1, fig. A), the fossa is rounded and partly
enclosed by the postorbital process. In spite of the depth of the fossa the internal
margin does not extend very far onto the dorsal surface of the cranium and there is
only a relatively small and poorly-defined depression around its inner edge, while

dps

Fic. 2. Left lateral view of skull, x1. Abbreviations as in Fig. 1.

6 REAPPRAISAL OF

posteriorly the main fossa slopes back to the line of the fronto-parietal junction.
The fronto-parietal edge is superficially eroded and the original temporal fossa may
have been more clearly defined.

Posterior to the fossa the squamosal and parietal form an arch over the auditory
meatus with a slight peak marking the posterior edge of the fossa; the quadrate
articulates immediately below it. On the left side the outer edges of the bones are
damaged, while on the right side both the postorbital process and the region above
the quadrate are broken.

The cranium is rather small and projects forward medially with the result that,
within the orbit, the posterior part of the frontal which forms the anterior ventral
wall of the cranial cavity slants away postero-laterally, and also at an angle of
c. 45° postero-ventrally.

In the upper region of the orbit the frontal is pierced, on either side of the inter-
orbital septum, by a large oval fenestra opening into the cranial cavity. At the
lower posterior edge of the frontal (or possibly in the orbitosphenoid) there is a large
optic foramen, piercing the interorbital septum and extending laterally on either
side of it. The septum itself is fragmentary and shows an extensive irregular
central fenestra and various small subsidiary foramina. There are grooves, pre-
sumably to accommodate nerves, along the dorsal and ventral edges of the septum.
The interorbital region of the frontal is thick, the roof of the orbit sloping from the
orbit edge towards the septum.

At its anterior dorsal end the interorbital septum is pierced by a small, vertical
fenestra which extends up into a hollow in the ventral surface of the frontal at this
point. Just posterior to this fenestra the septum edge forms a thickened ridge
descending from the orbit roof and curving forward below the fenestra, becoming
concealed by the matrix which encloses the inner nasal region. The interorbital
fenestra and hollow may have accommodated some structure with an olfactory
function. This area is more complete on the left side of the specimen.

On this side the roof of the orbit shows a peculiar feature. Within the orbit,
along its dorsal edge, there is an elongated oval depression, extending slightly
inwards and terminating internally at an abrupt ridge where the frontal resumes its
normal thickness. It is not clear to what extent the edge of the orbit may have been
damaged, but there has been a reduction in thickness of the orbit roof, towards its
edge, to accommodate a structure which was elongated, broader towards its middle,
and dorso-ventrally flattened to an even thickness.

At the anterior end of the orbit the frontal is thick and abruptly flattened on the
lateral surface. In some of the Recent Pelecaniformes this condition is associated
with the presence of a closely attached but unfused prefrontal, which may fall away
to reveal a similar surface of attachment. It therefore seems likely that a pro-
jecting prefrontal was originally present in Prophaethon, but has been lost. On the
left side, within the orbit, there is an incomplete and laterally projecting flange of
bone at the middle of the anterior edge of the interorbital septum. This might be
the remains of the lateral ethmoid plate associated with it. In either case there is
evidence for the existence of some structure at the anterior end of the orbital cavity,
of a type normally associated with a prefrontal.

PROPHAETHON SHRUBSOLETL a

Fic. 3. Ventral view of skull with left zygoma removed, x1. Abbreviations as in Fig. 1.

The jugal and quadratojugal form a narrow even strut of bone arising a little
dorsal to the posterior tip of the maxillary and extending in a straight line back to
the quadrate, continuing the line of the tomium of the upper mandible. The bar
has been fragmented but appears to be laterally flattened along its entire length
and shows no torsion.

Ventrally (Fig. 3; Pl. 1, fig. B), the basiparasphenoid is a small and triangular
plate, tapering anteriorly to a smooth rostroparasphenoid and at the external
posterior corners curving ventrally to form a pair of blunt projections only a little
anterior to the occipital condyle. The area is damaged and pitted in places and it is
not possible to be certain of the structure of the anterior part of the basiparasphenoid
plate.

The posterior surface (Fig. 4; Pl. 1, fig. E) of the cranium is rather flattened, but
a little inflated in the region of the supraoccipital. The foramen magnum is large
and rounded and tilted slightly downwards. The parietals form a broad arch over
the foramen. They are slightly hollowed dorsally towards the inner end, and the

Fic. 4. Posterior view of skull, x1. Abbreviations as in Fig. I.

8 REAPPRAISAL OF

outer ventral edges are prominent and rounded posteriorly. The surfaces of the
exoccipitals are concave. They have a slightly anterior ventral slant and begin to
taper ventrally before expanding again at the line of fusion with the posterior edge
of the basiparasphenoid plate.

QUADRATE. In Prophaethon the right quadrate is in the normal position with the
shaft vertical, but it has been damaged along its external side. The left quadrate
is also in position, but the posterior end of the pterygoid has become displaced
upwards, allowing the quadrate to tilt inwards at its lower end ; it is held in this
position by matrix on the internal side. The tilt has displaced the dorsal articula-
ting surface and allows a better view of this end.

The shaft is postero-anteriorly flattened, the narrow waist at the level of the
orbital process being about three times as broad as it is thick, and widening towards
the dorsal and ventral ends. The dorsal head is laterally broad, the external
flange of the otic process projecting laterally as far as does the quadratojugal socket
on the ventral head. The otic process slants laterally from the waist to form a
rather flattened process, with a shallow ridge from the base of the orbital process
crossing the anterior surface, but with some of the posterior surface and tip broken
away. The internal facet of the dorsal head appears to terminate as a rounded
structure projecting internally only a little further than the inner edge of the shaft.

The orbital process is blade-like, with an almost horizontal upper edge and a
slightly rounded tip. From this tip it curves down towards the lower part of the
shaft. It appears to project anteriorly at an angle to the quadrate shaft of about 70°.

The ventral head of the quadrate is elongated laterally. The quadratojugal
socket projects laterally and curves forwards so that the entrance of the socket is
more anterior than the lateral alignment. The posterior side of the head curves
back from the articulation socket to form a projecting posterior flange; from the
posterior side of the broken right quadrate of the specimen it would appear that
there is a fairly deep but small groove on the lower end of the mid-shaft, cutting into
the flange. The ventral surface of the quadrate then curves anteriorly again, but
on the specimen the mandibular articulation surface of this region is damaged.

PALATE (Fig. 3; Pl. 1, fig. B). The skull presents an example of a typical schizo-
gnathous palate. The maxillae are broadest posteriorly at the point where they fuse
with the palatines, and posteriorly they then taper rapidly on the external side to the
articulation with the jugal. Anteriorly, they taper a little and extend for most of
the length of the rostrum with a narrowing gap between them, fusion presumably
occurring in the premaxillary region at the tip of the rostrum (now lost). They are
slightly hollowed ventrally and the external edge projects ventrally to form the
tomium of the upper mandible. The palatines arise near the posterior ends of the
maxillae and extend back as shafts of similar width to the maxillae, but tapering
a little, and then expanding in the region below the orbits, to form elongated,
ventrally-hollowed blades, their posterior ends articulating with the pterygoids.
The posterior external sides of both blades are a little broken but together they show
the line of the external and posterior edge.

Je HAALID IL IAOIN Sig lela Osa! 9

The palato-maxillae arise on the internal edge of the palatines near the region
where the latter diverge from the maxillae. They extend posteriorly as narrow
blades lying alongside the internal edge of the palatines, but from the side view of
the skull it can be seen that they are laterally flattened blades with the upper edges
curving dorsally and externally on either side of the vomer. Laterally, in the angle
between the nasal bars and the posterior end of the maxilla, this upper edge of the
palato-maxilla can be seen on the right side of the specimen as a projecting flange
level with the junction of the nasal and maxilla ; on the left side it is broken, re-
vealing the narrow extensions where it joins the internal ventral and internal dorsal
ends of the maxilla.

Dorsally, the internal edge of the palatines extends upwards on either side of the
anterior ventral edge of the interorbital septum, its posterior edge curving inwards
and downwards to meet the posterior dorsal surface of the vomer. This structure
is fairly complete but a little broken along its upper edge on the left side, and more
extensively broken, but still showing curvature, on the right side.

The pterygoids are slender shafts. In ventral and lateral view the anterior end
which articulates with the posterior tip of the palatine can be seen to be expanded,
the sides diverging fairly evenly towards the tip on the ventral surface and with a
prominent rounded dorso-external end visible laterally. The main shaft shows some
lateral flattening and where it articulates with the right quadrate it shows dorso-
external torsion and a rounded, blade-like expansion at the posterior dorsal end.
This appears to increase the extent of the articulation with the internal side of the
ventral end of the quadrate.

Rostrum. Where it joins the skull the rostrum is almost square in transverse
section, but slightly narrower dorsally ; anteriorly it tapers very gradually to a point.
The frontal region of the skull terminates in a bold ridge, bordered anteriorly by the
deep, transverse fronto-nasal hinge. From the base of this ridge a rather flat nasal
surface of the basal rostrum projects anteriorly, merging into the premaxilla, and
tapering and becoming more rounded dorsally towards the anterior end. On either
side it is separated from the more ventrally situated maxillae by an elongated
aperture which also tapers and which extends to near the distal tip of the rostrum.
Two slender bars of the nasal slope posteriorly upwards from the posterior part of
the maxillae just where these widen, and appear to fuse laterally with the edges of
the broad posterior rostral surface just anterior to the fronto-nasal groove.

MANDIBLE (Fig. 6f; Pl. 2, figs A—C). The mandibular rami are slender, elon-
gated and with only a gradual taper towards the tip. They have been twisted and
show a dorsal torsion to the right. The distal end is damaged but the more proximal
and articular regions are complete. The dentary portion of each ramus is moder-
ately rounded externally, more marked towards the ventral edge. The internal side -
of the dorsal edge slants ventro-laterally, and ventral to this a deep groove occupies
the centre of the internal surface, extending along the distal two-fifths of the man-
dible. Proximal to this groove the internal surface is rounded with a more promi-
nent central ridge, continuing the internal tomium edge and gradually descending

Io

REAPPRAISAL OF

pre

pre

Fic. 5. Views of right coracoid, x2. A, dorsal; B, external; C, internal; D, ventral.
Abbreviations: gf, glenoid facet; prc, procoracoid ; sci, sternocoracoidal impression ;
sf, sternal facet.

PROPHAELTHON SHRUBSOLET itt

ventrally and becoming shallower, to disappear near the ventro-posteriorly slanting
mandibular suture. On the internal side (Fig. 6f), the upper edge of the splenial
extends about three-quarters of the way along the ventral anterior edge of the
suture and then extends anteriorly and ventrally as a thin line.

The mandibular suture appears as a deep groove internally, and much of the
lower half is occupied by a narrow, elongated fossa, at the anterior end of which a
narrow slit forms the internal opening of a poorly-defined anterior mandibular
fenestra. Externally (Pl. 2, fig. C) the upper part of the suture is fused, but there is
a narrow, elongated fossa, the dentary edge ventral! to it projecting slightly so that
the fossa is internal to it rather than dorsal. The upper edge of the fossa forms an
external opening for the fenestral slit. Posterior to this external fossa the suture
extends back ventro-posteriorly between the dentary and the anterior ventral parts
of the supra-angular and angular. From the anterior end of the fossa a shallow and
broad but well-defined groove extends almost horizontally along the external face
of the dentary and terminates at a level a little posterior to that of the posterior end
of the groove on the internal face of the mandible.

On the broad, internal face of the posterior part of the mandible, towards the
dorsal edge, there is a large, oval, posterior mandibular fossa between the pre-
articular and supra-angular. Since this is filled with matrix the internal structure
is not visible, but it does not penetrate to the external surface.

Posterior to this fossa the mandible becomes less deep, but broadens rapidly just
before the surface of the articulation with the quadrate. Dorsally, the surface
widens rapidly into a triangular, smooth surface with a small articular prominence
at the external corner. On the posterior side of the surface is a deep hollow which
articulates with the ventral head of the quadrate. It has a complete dorsal rim on
the external side, but elsewhere slopes to a median groove which occupies the centre
of an extension of the hollow opening into the internal side. Posterior to the hollow
the dorsal surface terminates as a narrow edge, broadening at the internal end into
a small triangular surface with a raised and rounded internal tip, and on the inner
side of this, broadening the posterior edge of the hollow, there is a small rounded
foramen.

The posterior surface of the mandible is flattened and forms a modified inverted
triangle. It has a distinct and narrow dorso-internal protrusion, while the lowest
part is more broadly rounded and projects as a narrow, curved lip beyond the main
ventral shaft. The surface is hollowed at the centre. The true position of this
surface is difficult to determine because of the torsion of the specimen, but it appears
to show some internal deflection and a marked posterior-dorsal tilt, the ventral lip
projecting beyond the rest of the structure.

b. Postcranial elements

STERNUM (Fig. 6a—c). The sternum was previously concealed within the matrix
and its presence was not suspected until the present preparation had begun. It is
still partially embedded in matrix and lacks its posterior end and the lateral pos-
terior parts of the sternal plate. The carina, which lacks the posterior end, is
exposed on its right side and along the anterior and ventral edges. In addition the

12

REAPPRAISAL. OF

cs

irs f

Fic. 6. Views of partly embedded sternum, x1; a. right lateral, b. anterior, c. ventral.
Views of imperfect left tibiotarsus, x1; d. oblique anterior, e. oblique internal. f. View
of the internal side of right mandible, x1. Abbreviations: art, articular; c, carina;
ca, carinal apex; cs, coracoidal sulcus; dg, dental groove; dms, dorsal manubrial
spine ; irs, inter-ramal suture; mf, mandibular foramen; occ, outer cnemial crest ;
pra, prearticular ; sur, surangular; vms, ventral manubrial spine.

PROPHAELTHON SHR UBSOLEL 13

ventral manubrial spine and right half of the manubrium and coracoid sulcus
are also present. The right half of the sternal plate is shattered posterior to the
thickened region of the sulcus and lacks the external and posterior parts.

The sternal plate shows slight ventral curvature. The carina is large, projecting
ventrally and anteriorly, with a slight curvature of the ventral margin. It becomes
thicker anteriorly, and at the carinal apex bifurcates to accommodate the ventral
end of the anterior articulating surface. The anterior carinal margin is stout, but
tapers at the dorsal end where it curves up to the underside of the ventral manubrial
spine. The latter is a small sharp projection, an inverted triangle in transverse
section with the wider dorsal surface rising slightly anteriorly and then curving
forwards and downwards, tapering away to a point a little below the level of the
ventral edge of the spine.

The anterior carinal margin curves forwards ventrally, its apex anterior to the
tip of the spine. The upper part, about three-fifths of the whole, curves forwards
and becomes thicker ventrally, and forms a blunt projection at a similar anterior
level to the tip of the manubrial spine. Below this, the remaining two-fifths form
a flattened, elongated anterior facet, hollowed centrally and curving forward ven-
trally. It appears analogous to similar surfaces on the sterna of some pelecaniform
species, which articulate with the furcula in the region of the ventral symphysis.
Viewed laterally (Fig. 6a), this surface on Prophaethon is hollowed to such an extent
that it shows some posterior curvature. Viewed anteriorly (Fig. 6b) it is dorso-
ventrally elongated and wedge-shaped, widening ventrally and with the ventral
portion curving anteriorly. At the ventral end the hollow becomes shallower,
forming a poorly-defined lip below the deepest part of the cavity.

The dorsal lip of the coracoid sulcus appears thickened but is superficially damaged.
The ventral lip arises at the lateral dorsal edge of the base of the ventral manubrial
spine and slants posteriorly across the sternal plate. There is a broad ventral
surface between the anterior edges of the dorsal and ventral lips. The ventral
labial prominence occurs about two-thirds of the way along the sulcus as a bluntly
rounded, thin flange overlapping the sulcus which up to this point appears to be
ventrally exposed. A low ridge across the sternal plate from the posterior part of
the carina terminates at the sulcus after crossing the ventral surface of the ventral
labial prominence.

The sulcus terminates a little lateral to the prominence, the ventrally projecting
ridge of the dorsal lip curving posteriorly towards the end of the sulcus to leave a
thinner, laterally-projecting, area of the sterno-coracoid impression.

Coracorp (Fig. 5; Pl. 2, figs D-G). The right coracoid is preserved, but an-
teriorly it is broken off just above the glenoid facet, the broken surface extending
along part of the ventral edge. The sterno-coracoid process and internal distal
angle are also damaged.

The shaft is smooth and rounded, but towards the distal (sternal) end it is dorso-
ventrally flattened. The sterno-coracoid surface is large and slightly curved
ventrally, and since the dorsal lip of the coracoid sulcus is markedly anterior to the
ventral lip, as already described in discussing the sternum, the ridge marking the

14 REAPPRAISAL OF

articulating surface for the latter appears across the middle of the dorsal surface of
the sterno-coracoid area. It becomes lower and disappears before reaching the
flattened and rather rectangular sterno-coracoid process. The internal distal angle
of the coracoid is broken off at the line of the articulation ridge.

The shaft becomes thicker and more rounded, and laterally narrower, as it ap-
proaches the procoracoid. From the faint scar of attachment of the coraco-brachialis
a poorly-defined ridge crosses the shaft to the distal end of the procoracoid.
The latter projects laterally and curves anteriorly and ventrally to a point (broken
short in the specimen). On the internal side there is a curved hollow between the
procoracoid and shaft, with a small longitudinal ridge on either side of it. The
internal opening of the coracoid fenestra is a small hole near the distal end of this
hollow, and the external opening is a similar hole on the dorsal ae. of the pro-
coracoid, towards its edge.

The scapular facet is shallow and dorsal, and from it a thickened ridge borders the
proximal edge of the procoracoid. The glenoid facet is dorsoventrally aligned.
Ventrally, there is a low ridge from the ventro-external edge of the shaft, which
crosses the sterno-coracoid process and terminates near its distal external extremity.

SCAPULA. The single left scapula lacks the proximal articulating end. It is a
slender rod with a slight ventral curvature, dorso-ventrally flattened. It is a little
thicker and more rounded at its base and towards its tip becomes more flattened and
broader, then tapers to a point.

PELvis (PI. 3, figs A-C). Most of the pelvis is present. At the anterior end of it
a row of laterally crushed vertebrae form the relic of the thoracic part of the vertebral
column. Dorsally the pelvis has been cleared of matrix but some of the more
ventral detail of the left side is still obscured by matrix containing broken ribs.
On the right side it is completely exposed. The anterior lateral edge of the ilium
is broken on the right, but appears complete on the left side. Posteriorly, on the
right, the ischium and pubis are broken off at about the posterior end of the ilio-
ischiatic fenestra, and the ilium posterior to the fenestra is also missing, the dorsal
end terminating at the incomplete caudal end of the synsacrum. On the left side,
part of the iliac portion of the synsacrum is visible and the ventral struts of the roof
of the renal depression can also be seen on the left.

At the anterior end of the pelvis the median-dorsal ridge is thick and prominent,
appearing to fuse with the end of a row of elongated and flattened neural spines.
The ridge projects sharply, the iliac plates sloping down steeply on either side and
curving outwards to form a lateral flange, broadest anteriorly and with a rounded
tip. Posteriorly the dorsal ridge widens and the iliac plates become narrower and
less hollowed laterally. The edges of the ridge diverge as two well-defined edges
of the anterior iliac crest, curving outwards to a slight prominence above and internal
to the acetabulum. From there, two blunt, broad ridges continue posteriorly as the
posterior iliac crest, curving slightly towards each other in the region of the ilio-
ischiatic fenestra, before diverging posteriorly. The shield area between them is
narrow and slightly hollowed, with a median ridge beginning to appear towards the

PROPHAETHON SHRUBSOLEI 15

posterior end; on the left side at the posterior end there is what may have been
the first of a double row of narrow fenestrae from the renal depression. Below the
posterior iliac crests the iliac surface slopes outwards as a narrow upper edge to the
iliac-ischiatic fenestra. The fenestra is large and elongated, becoming wider
posteriorly, with a relatively straight lower edge, a posteriorly-curved upper edge,
and the two meeting anteriorly in a small rounded end just posterior to the lower
edge of the antitrochanter. Anteriorly the surface bordering the upper edge of the
fenestra flares out to a prominent lip over an antitrochanter, the articulating surface
of which has an anterior/external aspect and also an anterior/ventral slant. Below
this is a rounded acetabulum, the lower third of which does not penetrate completely
to the ventral side, but forms a rounded hollow in the bone. There is a small
lateral projection midway along the anterior external edge of the acetabulum.

Just ventral to the anterior edge of the acetabulum a small anteriorly-projecting
pectinal process is present, posterior to which the pubis slopes away as a slender rod
of bone. The anterior end of the ischio-pubic fenestra is narrow but rounded and
terminates posterior to the ventral edge of the acetabulum. Dorsal to it the ischial
bar is nearly twice the width of the pubis. Just posterior to the acetabulum the
external surface of this ischial bar slopes ventro-internally, until it meets a small
ridge which slopes back from the posterior edge of the antitrochanter, at which
point torsion occurs and the external face of the ischium has a ventro-external slant,
and shows a poorly-defined, posteriorly and ventrally slanting, ridge in its surface.

The fused synsacrum has its greatest depth at the anterior end and viewed laterally
the dorsal line of the pelvis and the ventral surface of the synsacrum converge
towards the posterior end, forming a thin elongated wedge that shows a slight ventral
curvature at its posterior tip. The ventral surface of the synsacrum is widest at the
region of the sacral vertebrae, where it shows a well-defined median groove, and
tapers gradually towards the posterior end. Anteriorly it also begins to taper but
becomes broader again at the anterior tip where it articulates with the first free
vertebra. The latter appears to be the last dorsal vertebra. The anterior articu-
lation surface of the synsacrum is posterior to the broad dorsal anterior edge of the
ilium.

Femur. The right femur was exposed in the specimen as originally prepared ;
it appears from the original figure (Andrews 1899) to have had a coating of matrix
and to have been superficially damaged beneath this. The present preparation has
exposed on the distal half, which is detached from the main block but still attached
to the tibiotarsus, an internal distal surface almost to the condyle ; on the proximal
half still 72 situ on the pelvis the external and posterior surfaces are visible.

The distal internal face shows a smooth, slightly rounded surface with a rounded
and posteriorly projecting condyle, the internal condylar surface showing an internal
deflection towards the distal end.

The trochanter shows an abrupt obturator ridge with a hollow beneath it on the
posterior side. The anterior head of the trochanter is rounded, and the external
edge appears to form a blunt, projecting ridge which extends for some way along the
anterior external edge of the shaft.

16 REAPPRAISAL OF

TIBIOTARSUS (Fig. 6d—e; Pl. 3, figs D-E). During preparation the proximal end
of the left tibiotarsus was removed and cleaned. The external side of the shaft had
originally broken away as far as the head, but the internal side is still intact.

The anterior and internal sides of the shaft are present and where they join there
is an abrupt edge which becomes more prominent towards the proximal end and
finally forms the anteriorly-projecting flange of the inner cnemial crest. The inner
cnemial crest is a thin prominent flange arising along the internal edge, projecting
anteriorly and at its outer edge curving a little externally. It extends proximally
well beyond the articulating surfaces. It arises gradually from the shaft, is deepest
at about the level of the proximal edge of the articulating surfaces, and then tapers
to a blunt point proximally. The outer cnemial crest arises on the anterior surface
near the external edge and nearer the proximal end than does the inner crest. It is
thicker distally than the inner cnemial crest and projects at an angle between the
anterior and external surfaces, curving slightly towards the external side. Its
proximal end terminates at an angle a little proximal to the articulation surfaces.
It appears to end abruptly as though broken short but this is also apparent on some
entire examples of Recent species.

The internal articular surface is more distally placed than the inter-articular area
and the ends of the crests. Its surface is damaged on the specimen, but it shows a
posteriorly-projecting, curved lip. The inter-articular area slopes distally on the
external side, curving distally to the broken external surface in a shallow hollow on
the external side of the outer cnemial crest.

RIBS AND OTHER ELEMENTS. The head of a rib and part of another are visible on
the left side of the specimen anterior to the pelvis, together with broken shafts still
embedded in matrix. The visible head is relatively stout ventral to the tubercle,
and the portions of shaft are fairly broad. There is no obvious uncinate process,
but a slender and flat strip of partially embedded bone parallel to the left posterior
iliac crest might be referable to this. In general structure the ribs resemble those
of the larger larids and sulids.

In addition to the material described above, there are also a number of fragments
of bone and matrix removed during preparation. A close examination of these
might make their identification possible, but they are not likely to provide additional
useful information to the present study. Their existence is therefore noted, but no
further study has been made.

c. Measurements

All dimensions are given in millimetres.

SKULL

Overall length (premaxilla—supraoc- Maximum width at postorbital pro-
cipital) 112 cesses 38

Maximum width at posterior end Minimum width of interorbital bar 15

(squamosals) 35°5 Width at anterior end of frontals USE

PROPHAETHON SHRUBSOLET 17

SKULL (continued)

Width at base of rostrum (dorsal)

Width at base of rostrum (ventral)

Mid-rostral width

Width at tip of rostrum as preserved

Length of culmen

Length of lateral nasal aperture

Maximum depth of lateral nasal
aperture

Posterior depth of cranium to occipital
condyle

Maximum cranial depth

Depth from anterior end of frontals to
palatines

Width of temporal fossa at postorbital
process

Minimum width of cranium at tem-
poral fossae

QUADRATE

Quadratojugal socket to otic process

External—internal width of quadrato-

’ articular surface

Anterior— posterior width of quadrato-
articular surface

LOWER MANDIBLE

Maximum length of left dentary to
post-articular process (incomplete)
Maximum length of right dentary to
post articular process (incomplete)

Depth of dentary at tip

Width of dentary at tip

Depth of dentary at proximal end of
internal dentary groove

Width of dentary at proximal end of
internal dentary groove

Maximum depth of dentary at intra-
ramal suture

Maximum width of dentary at intra-
ramal suture

SCAPULA

Overall length (incomplete)
Maximum distal width

CORACOID

Overall length
Maximum distal width
Width of glenoid facet
Width of scapula facet
Width of sternal facet

2

14°5

232

67°7

Length of dorsal hollow in orbit

(? nasal gland) 22
Width of dorsal hollow in orbit

(estimated) 4
Length of palatines 42
Maximum proximal width of palatines 5°6
Length of pterygoid 16
Width of shaft of pterygoid 1-6
Width at quadrato-pterygoid articu-

lation 4°3
Maximum width of posterior end of

basiparasphenoid plate 17
Width of rostroparasphenoid 32
Length of zygoma 44°2
Depth of zygoma 226
Width of zygoma ee

External—internal thickness of shaft

below orbital process 4°8
Maximum depth of orbital process 6-2
Length of orbital process 7°8

Depth at anterior end of articular

surface 6°5
Width at anterior end of articular

surface g'I
Depth at posterior end of articular

surface 9°3
Width at posterior end of articular

surface 11:2
Maximum length of posterior mandi-

bular fossa 14°3
Maximum depth of posterior mandi-

bular fossa 6°7
Proximal width 4°9
Thickness at proximal end Pad |
Width of shaft at coracoidal fenestra 6°8
Internal—external length at sternal

facet (left) 20°2
Proximodistal width of sternocora-

coidal process (left) 85

18 REAPPRAISAL OF

STERNUM
Maximum length to ventral manubrial Length of furcular facet 10'9
spine 51°5 Maximum width of furcular facet 6:8
Length of carinal edge (incomplete) 59 Tip of ventral manubrial spine to
Carinal apex to dorsal edge of ventral ventral labial prominence 22°1
manubrial spine 26°5 Width of coracoidal sulcus 5°7
Tip of ventral manubrial spine to
anterior carinal margin 6:3
PELVIS
Maximum medial length as preserved 67 Minimum distance between antitro-
Anterior border of ilium to anterior chanter and pectineal process 14
edge of acetabulum 35 Ventral length of synsacrum 69°5
Maximum width across antitrochan- Anterior depth of pelvis from median
ters 27° A) dorsal ridge to synsacral thoracic
Anterior width from external edge of vertebra 19°5
ilium to medial ridge Ly Width of articular facet of first syn-
Length of ilio-ischiatic fenestra 23°3 sacral thoracic vertebra 75
Maximum depth of _ ilio-ischiatic
fenestra 10
FEMUR
Proximodistal length 51 Maximum proximal width at trochan-
Anterior/posterior thickness at mid- teric ridge 9°4
shaft 4°8
TIBIOTARSUS
Length as preserved Br3 Width from internal edge to external
Articular surface to tip of internal edge of internal cnemial crest 15'5
cnemial crest 6-1 Width to internal edge of internal
Maximum width of internal cnemial cnemial crest 8-6
crest 5:2 Width to external edge of outer
cnemial crest 14

III. COMPARISONS WITH RECENT MATERIAL

In attempting to place Prophaethon within the framework of avian taxonomy,
using the characters revealed by further preparation, it has been necessary to
compare it with osteological material from various Recent taxa. Since the earlier
claims of pelecaniform similarities were based on skull characters it seemed preferable
for comparative purposes to begin at the other extremity.

PROXIMAL END OF TIBIOTARSUS. The obvious characters on this element are the
prominent inner cnemial crest, which has its widest part just above the level of the
articulation surfaces and tapers proximally, and the inner cnemial crest forming a
blunt projection at a similar level. The proximal articular surface slopes posteriorly
and has a projecting lip with a concave proximal surface at the posterior edge. The
external edge of the proximal surface curves smoothly over onto the shaft in a
broad zone where it borders the inner cnemial crest. |

Stercorarius provides a very close match in all aspects, while other Charadriiformes
also show resemblance to varying degrees. In the Procellariiformes Diomedea

PROPHAELPHTON SHRUBSOLET 19

shows some similarities but lacks the projecting posterior lip, and has a larger and
more proximally situated inner cnemial crest. There is no obvious resemblance to
the Pelecaniformes, where Phalacrocorax retains only a short curved outer cnemial
crest while Sula has small blunt projections. The features are almost entirely
lacking in other pelecaniform genera.

Femur. This is long and narrow, the trochanteric crest present as a prominent
narrow ridge rounded off at the proximal end with a curved muscle scar incised into
the external surface. In the Charadriiformes the ridge is much more developed,
both anteriorly and proximally. Macronectes in the Procellariiformes shows
similarities to the specimen. Swla has a stout, curved femur, the proximal end of
which shows some similarity to that of Prophaethon, but the trochanteric ridge is low
and blunt and the head is more ventrally deflected.

Petvis. The narrow, elongated pelvis is typical of that found today in birds
which swim and dive to catch their prey. It bears no resemblance to that of
Phaethon, which is broad and short.

The anterior shield of the ilium extends forward only as far as the proximal end
of the second synsacral thoracic vertebra, and the third vertebra is unfused. This
condition is typical of the Charadriiformes but not of the other orders examined
here. Sula does, however, show a short anterior iliac shield.

The median dorsal ridge is slightly convex, but posteriorly it is depressed between
well-defined posterior iliac crests. In this respect the specimen resembles Phoebetria,
Diomedea and Puffinus among the Procellariiformes. The Alcidae show a similar pro-
file. In the Pelecaniformes the posterior iliac crests are poorly developed and in the
more aquatic forms such as Sula and Phalacrocorax the posterior median ridge is
level with or dorsal to the lateral crests.

The Pelecaniformes also show more laterally projecting and anteriorly directed
antitrochanteric surfaces than does Prophaethon, and in the latter the iliac surface
immediately anterior to the acetabulum is concave and the pectineal process pro-
minent. The specimen is more similar in these respects to both Procellariiformes
and Charadriiformes, although among the last the Alcidae show the pectineal
process greatly reduced or absent. From fine sutures apparent on the specimen the
ilium would appear not to have been fused with the synsacrum, in this respect
resembling Recent Alcidae and Procellariiformes rather than Pelecaniformes.

VERTEBRAE. The free thoracic vertebrae appear to lack hypopophyses. They
have rounded concavities laterally, posterior to the prominent anterior costal facet.
The diapophyses are short, dorsally broad and have a thin middle ridge ventrally
with a deep concavity at the internal corner of the anterior side. There is a thin,
tapering anterior process at the distal end of the diapophysis. These vertebrae
resembles those of the larger Larus species in the Charadriiformes. Those of the
Procellariiformes are more elaborate with larger concavities and various fenestrae,
and with hypopophyses ; those of the Pelecaniformes show still fewer similarities
to those of Prophaethon.

20 REAPPRAISAL OF

STERNUM. Allowing for the incompleteness of the specimen the carina of the
sternum is deep and long, extending further back than is the case on Recent Pelecani-
formes. The ventral lip of the coracoid sulcus is posteriorly situated so that the
sulcus is ventrally exposed, particularly at the middle region, but towards the outer
end it undercuts a small but broad labial prominence.

In the Pelecaniformes the ventral lip of the sulcus extends almost as far anteriorly
as does the dorsal lip, and the sulcus is a deep, anteriorly-directed groove. In
Prophaethon the groove is more typical of that found in Charadriiformes and Pro-
cellariiformes. From the ventral labial prominence a distinct intermuscular line
slants postero-internally towards the middle of the carina. A similar line is present
in Charadriiformes, but in the Procellariiformes one line arises on the sulcus internal
to the labial prominence and crosses the sternum more anteriorly, while another
rises towards the external end and slants towards the posterior end of the carina.
In Pelecaniformes there is a faint ridge from the labial prominence apparent in
Phalacrocoracidae and Phaethontidae.

The hollow facet for furcular articulation at the anterior tip of the carina is a
distinctive character in Pvophaethon. Structures of this kind are found in some
Recent species of Pelecaniformes and Procellariiformes. A small articulation
surface is present on the larger Podicipitiformes. This furcular surface is more
extensively developed in the Pelecaniformes where the Pelecanidae and Fregatidae
have the furculum fused to the sternum while Phaethontidae, Sulidae and Phala-
crocoracidae have the surface developed to varying degrees. Of the Procellarii-
formes the Pelecanoididae have a sternum with a large furcular articulation facet,
its ventral edge curved anteriorly. The bifurcation of the carina tip, apparent in
Prophaethon, is characteristic of many Procellariiformes but not of the Pelecani-
formes. None of the Charadriiformes show furcular articulation facets on the
sternum.

The shape of the manubrial spine in Prophaethon resembles that of Phaethon, but
this structure differs so markedly in different families and genera that we do not
regard it as taxonomically useful.

Coracoip. The shape of the sternal (distal) end of the coracoid is correlated
with that of the coracoid sulcus. The coracoids of Pelecaniformes show sternal
facets on both sides at the distal end, articulating with the deep sulcus. Prophaethon
resembles charadriiform and pelecaniform birds in having a prominent facet across
most of the dorsal surface ; a small one on the ventral surface towards the external
end is correlated with the position of the ventral labial prominence of the sternum.
From the fit of the specimen there is no reason to suppose that the missing internal
distal angle of the coracoid of Prophaethon would have projected across the midline
of the sternum, as in Phaethon.

A stout and dorsally curved internal distal angle to the coracoid is characteristic
of most Procellariiformes, although less evident in the Pelecanoididae, and in the
Laridae and Alcidae of the Charadriiformes. Although this part of the specimen of
Prophaethon is damaged, the general shape of the surrounding bone and the sternal
fragment of the other coracoid, which is also present, indicates an absence of such

PROPHMAELITON SHRUBSOLET 21

curvature. The specimen most closely resembles the flatter coracoids of Ibido-
rhynchus and Haematopus of the Charadriiformes in this respect ; it also resembles
them in the area of irregular surface on the dorsal side distal to the line of attachment
of the coraco-brachialis muscle, the proportions of the shaft, the procoracoid, and
the position of the coracoidal fenestra.

LOWER MANDIBLE. This, although relatively complete, does not give an indica-
tion of affinity with any particular taxon.

The slender ramus, increasing in depth posteriorly and tapering a little at the
anterior end of the articular portion, has a general resemblance to those of Phala-
crocoracidae and some Procellariidae. The tapering groove on the external surface
anterior to the intraramal suture is similar to that of Procellariiformes and Char-
adriiformes, but a groove of this type is also present in Phaethon.

The arrangement of the component bones around the external fossa of the mandi-
bular suture and the shape and position of the fossa itself are most closely paralleled
by the structure in some smaller Laridae, the Burhinidae and to a lesser degree by
some Procellariiformes. The concave, ventrally tapering and postero-dorsally
oriented postarticular surface is very similar to that of Gaviiformes, but only
resembles to a limited extent those of the other taxa examined here.

The articular facets are very similar to those of Laridae, Burhinidae and Pro-
cellariidae but do not closely resemble the more specialized structures of the Pele-
caniformes.

Prophaethon has a large, rounded posterior mandibular fossa, bordered anteriorly
by a large prearticular which extends anteriorly to border the mandibular suture
and overlaps with, and possibly fuses with, the dentary towards the dorsal edge of
the ramus. This large prearticular is also present in the Procellariiformes and
Gaviiformes. The Charadriiformes have a large posterior fossa but the prearticular
is small and only partially occupies the anterior space, the mandibular slit forming a
distinct fenestra. Phaethon shows a similar condition but has the prearticular
larger posteriorly, reducing the size of the posterior fossa. In the other pelecaniform
families the posterior fossa is tiny and the prearticular has filled the remaining
space.

SKULL. The palate is typically schizognathous and resembles those of Charadrii-
formes and Gaviiformes. Procellariiformes also have this type of palate, but the
distal end is modified by bill shape. The palates of Pelecaniformes appear to be
schizognathous in the very juvenile condition but desmognathous in the adults.
Schizognathous palates are also present in Podicipitiformes, Gruiformes, Galli-
formes, Sphenisciformes and Columbiformes. The distinctive proximal bifurcation
of the vomer, visible between the palatines in Prophaethon, is very similar to that of
larger Larus species and of other Charadriiformes, but does not appear in Pelecani-
formes and Procellariiformes.

The rostrum of Prophaethon is long, slender and tapering. An elongated nasal
aperture runs for almost the whole length, rising posteriorly and tapering to a slit
just proximal to the frontonasal hinge. In general shape the rostrum is similar

22 REAPPRAISAL OF

to that of the Phalacrocoracidae, and if it were argued that the latter retained the
open aperture of the juvenile there would be strong similarity. In other Recent
birds the elongated nasal apertures are typical of Charadriiformes, Gaviiformes,
Podicipitiformes and Gruiformes. Although the nares of Pelecaniformes are
typically closed or minute, Phaethon shows an intermediate condition with a short
but relatively large aperture about a third of the way along the rostrum and a tiny
hole near the frontonasal hinge.

A transverse frontonasal hinge developed to differing degrees in different families
is associated with the greatly reduced nostrils in Pelecaniformes. In Prophaethon
the nasal apertures are schizorhinal in shape, but since the nasal struts join the
rostrum anterior to the frontonasal hinge they are functionally holorhinal. A
holorhinal type of nostril associated with a frontonasal hinge in this position also
occurs in the Charadriiformes in the Burhinidae, Thinicoridae and Pluvianus of the
Glareolidae, although charadriiform nostrils are usually collectively described as
schizorhinal. A deep transverse frontonasal hinge comparable with that of the
Phaethontidae also occurs in Rhynchops of the Charadriformes, but in this instance
the nostrils are schizorhinal. Procellariiformes have an unspecialized transverse
frontonasal structure and holorhinal nostrils.

In the Pelecaniformes the heavy frontal brow associated with a deep frontonasal
groove is present only in the Phaethontidae. In the Charadriformes it is partially
developed in Rhynchops, and more highly developed in the Chionidae. In the
latter the nasal struts lie alongside the rostrum and also terminate just below this
brow, apparently forming a hinge with the rostrum. This structure has some
analogy with that of Prophaethon since a close inspection of the latter reveals that
the nasal lies close alongside the dorsal rostral surface but may not fuse with it
completely, and the transverse line of the frontal brow is not completely straight but
shows paired lateral recesses in the fore-edge which might be comparable with those
of Chionidae. Unfortunately the surfaces of the bone are slightly damaged at this
point in the specimen of Prophaethon.

The frontal shows a flattened lateral surface suggesting the loss of an unfused
prefrontal. Prefrontals are unfused in the Phaethontidae and Fregatidae of the
Pelecaniformes, and in Burhinidae of the Charadriiformes ; when they become
detached they leave flat surfaces similar to those of Prophaethon.

The smooth dorsal surface of the skull of Prophaethon may have pelecaniform
similarities, but only by virtue of the absence of the nasal glands, and were these not
present in other taxa they might also be similar. The feature is therefore a general-
ized one, indicative of absence of specialization rather than affinity. In the deeply
rounded temporal fossae posterior to the prominently curved postorbital process
the specimen resembles the Fregatidae, Uvia of the Alcidae and Stercorarius in the
Laridae, rather than the Phaethontidae, but shows some similarity to the last in
the rather limited development of the fossae dorsally and anteriorly.

In the area of quadrate articulation the ventral edge of the orbitosphenoid curves
forwards, leaving a large cavity anterior and internal to the articulation surface and
tending to divide the latter into two. This is also present to a similar degree in the
Pelecaniformes, but much reduced in the Procellariiformes and Charadriiformes.

PROP HAE HON San Os SOL EL 23

The posterior aspect of the skull of Prophaethon has parallels in all the major taxa
here examined, and there are considerable differences in these at the family level.
The supraoccipital of Prophaethon has a small foramen on either side towards the
ventral edge, with associated grooves bordering the upper edge of the foramen
magnum. Similar structures are apparent in the Procellariiformes, more modified
in the Charadriiformes, and greatly modified or absent in the Pelecaniformes.

Within the orbit of Prophaethon the hollow apparently accommodating the nasal
gland is in a unique position, being an elongated hollow bordering the dorsal edge of
the orbit and on the ventral surface of it. In Pelecaniformes the nasal gland is
usually towards the anterior dorsal end of the orbital hollow and in a more median
position. In some Phaethon species a small elongated hollow extends posteriorly
into the orbit from this anterior site, bordering the orbital septum but separated
from it by the channel of the olfactory nerve. In the Sulidae it is an elongated
hollow lying completely in the roof of the orbit, but still close to the septum. In
Charadriiformes and Procellariiformes the nasal gland hollow either borders the
dorsal edge of the orbit or lies internal to it, but on the dorsal surface of the frontal.
The position of this structure in Prophaethon is therefore intermediate between that
in the two Recent types.

The slender, unspecialized pterygoids of Prophaethon, with a small posterior
lateral expansion of the orbital process of the quadrate, give little useful evidence of
affinity. The Phaethontidae show an even less modified structure, but the ptery-
goids of other pelecaniform families and of the other taxa discussed here are stouter,
with some development of lateral flanges on the shaft, and in several diverse taxa
show the expanded posterior end.

The quadrate of Prophaethon also shows little evidence of affinities, and the deep
groove at the ventral end of the posterior surface of the shaft appears to be a pecu-
liarity of the species.

Tables r and 2 give a simplified summary of apparent similarities. Tables such
as this may oversimplify, exaggerating both similarities and differences, but they
indicate the problem involved in assigning Prophaethon to a known taxon.

TABLE I
Comparison of postcranial elements of Prophaethon with extant orders of birds

Charadriiformes Procellariiformes Pelecaniformes
Proximal end, tibiotarsus * x —
Proximal end, femur -
Pelvis zs
Vertebrae *
Sternum, other than anterior facet *
Anterior edge of carina
Coracoid a!

* *X X *X

xX *

* very similar; x some similarity; — little or no similarity

24 REAPPRAISAL OF

TABLE 2

Comparison of cranial elements of Prophaethon with extant orders of birds

Charadrii- Procellarii- Phaethon- —— :
formes formes tidae alee
formes

LOWER MANDIBLE

Shape x = = t

External intraramal suture area * * x ~_

Articulation * * — —

Posterior fossa * = x —

Prearticular = = - *
SKULL

Palate * x - ~

Rostrum z — — x

Frontonasal hinge = * x

Unfused prefrontals “s —

Temporal fossae = x * =

Supraoccipital area x - x -

Quadrate articulation area x x 3 *

* very similar; x some similarity; — little or no similarity

IV. DISCUSSION AND CONCLUSIONS

From the above data comparing Prophaethon with Recent birds there would appear
to be several possible hypotheses concerning its relationship to known taxa.

The species might be regarded as a charadriiform bird, showing the alcid type of
pelvis and with a head adapted for catching prey in water and showing modifications
which have parallels elsewhere in the taxon. But the specialized sternum with its
facet for furcular attachment is not present in known Charadriiformes, and it would
be necessary to argue that, since it is known in aquatic birds of two other orders, it
might be an adaptation that potentially could have evolved under any similar
behavioural or environmental selection pressures.

Prophaethon might be regarded as a procellariiform bird, in which case there is a
precedent for a modified sternum, but the skull and rostral characters would not be
typical of the known forms of that order.

If it is argued that the possession of both the specialized frontonasal hinge and
the sternum are indications of pelecaniform affinities, then another assumption must
be made, that a saltatory form of evolution has produced the more extremely
evolved characters now present in this evolutionary diverse order while retaining
more generalized characters in other structures which could presumably undergo
subsequent modification. This is at variance with the apparent evidence provided
by a proto-frigatebird from the Lower Eocene, described by Olsen (1974), which
shows modifications in various skeletal elements of the postcranial skeleton but has
not achieved either the specialized skull or the sternum typical of the Fregatidae.

If any of the above suggestions are adopted then, at least in so far as the early
Tertiary is involved, the range of osteological characters used to define any of the

PROPHAETHON SHRUBSOLEI 25

three orders must be extended to include a number of significant characters which
are at present regarded as diagnostic of some other order. The limiting groups of
characters which may be used to assign any species to a particular order will then
no longer be clearly separable ; we shall have to assume it is not possible to identify
a bird from this geological period unless certain critical skeletal elements are avail-
able.

Alternatively, in view of the lack of evidence of affinity with any single order, it
could be suggested that in the early Tertiary these three Recent orders had not yet
diverged, and that Prophaethon was referable to an ancestral stem from which
more than one Recent order had subsequently evolved. There is, however, a
considerable amount of fossil bird material from this period, most of which appears
to be referable to Recent families, suggesting that the degree of evolutionary diver-
gence apparent in Recent families had already occurred by this period. In addition
fossils referred to Pelecaniformes and Charadriiformes are known from the Upper
Cretaceous and from the evidence available it seems that in general the ordinal
divisions of birds had occurred within the Cretaceous.

Some other Lower Eocene species show this intermediacy of affinities. Among the
more specialized forms described from this period is a sea-bird with bony, tooth-like
projections on the jaws, Odontopteryx toliapica of the family Odontopterygidae.
Related forms, usually separated in the family Pseudodontornithidae, are known
from the Miocene. In recent studies of the British Lower Eocene we have found
evidence of more numerous and varied forms of both families at this period. These
birds were at first regarded as pelecaniform, but Howard (1957), in describing
Osteodontornis orrt of the Californian Miocene, pointed out that the skeleton showed
a mixture of pelecaniform and procellariiform characters. The British specimens,
recently prepared, appear to confirm the rather specialized nature of this group.

Avian phylogeny in the Cretaceous is still a mystery and in view of the lack of
evidence speculation is of very limited use. If the Lower Eocene is taken as a base
line there is an array of families known from fossil remains, most of them first known
to occur at this period and subsequently persisting until Recent times. If we
insert into this array both Prophaethon and the bony-toothed birds we may produce
the kind of picture shown in simplified and linear form in Fig. 7.

Let us assume that the base line is the Lower Eocene and the lines rise to the
Holocene, that Ar—A3 represents charadriiform families, B1—B3 pelecaniform
families and Cr—C3 procellariiform families. If we insert Prophaethon as AB, the
Odontopterygidae as BCr and the Pseudodontornithidae as BCz, then we have a
continuous sequence in which each family shares some characters with the adjacent
ones and the whole presents a relatively uniform array which might have arisen as a
complex adaptive radiation from a single ancestral origin, rather than as a divergence
from three separate stems, yet showing some morphological convergence. If during
later periods Prophaethon and the bony-toothed birds became extinct, then the
remaining families would appear to fall into more discrete groups which are then
identified as Recent orders — ordinal definition being aided by the absence of the
intervening forms. The diagram presented is a linear one, but the relationships
should be visualized as three-dimensional, the three Recent orders forming a triangle.

26 REAPPRAISAL OF

Al A2 A3 Bl B2 B3 Cl C2 C3

ay
ow
a

Fic. 7. Hypothetical relationship of families.

The bony-toothed birds would occupy the space intervening between the Pelecani-
- formes and the Procellariiformes, while Prophaethon would come in a more central
position in view of its similarities to all three orders.

We have also indicated by transverse lines two possible distributions of shared
characters resulting from the loss of some of these families. If BC1—2 disappear
then the two linking characters which they share become discrete ordinal peculi-
arities. The character which AB is shown as sharing with families on either side
would, following the extinction of AB, still persist in more than one Recent order
but might be limited to certain families. If we assume A1r—3 to be charadriiform
families and Br to be the Phaethontidae, species of the latter are seen to share a
number of minor morphological characters with the former group and not with other
pelecaniform families. The existence of such characters led to some early sug-
gestions that the Phaethontidae had more affinity with the Charadriiformes than
with the Pelecaniformes, and Mathews & Iredale (1921) created for it a suborder
within their gull order, Lari. These similarities have also been explained as con-
vergence with the terns (Sterninae of the Laridae) brought about by similar methods
of feeding. The latter might be a valid explanation, but on the basis of the hypo-
thesis proposed some at least of these similarities might be explained by a closer
degree of affinity between the families involved at an earlier period and by the
persistence of some shared characters.

This proposed arrangement would appear the most satisfactory to explain the
combinations of characters found in Prophaethon and the bony-toothed birds, and
to present a balanced view of sea-bird evolution in the Lower Eocene based on
present evidence. There has for long been general acceptance that the pelecaniform
and procellariiform birds shared a common ancestry more recent than that of avian
orders in general. The suggested association of the Charadriiformes with these has
been weaker and based mainly on the peculiarities shown by the Phaethontidae.

PROPHAEBTAON SHRUBSOLET 27

The evidence from the skeletal structure of Prophaethon might now justify, pro-
visionally, a stronger linking of the Charadriiformes with the Pelecaniformes.
These three Recent orders, with the two interordinal linking taxa, appear to form a
more unified group, perhaps a single superorder which, for the purposes of arranging
avian taxa of the Lower Eocene and earlier periods, might provide a more convenient
unit than the existing groups.

This idea does not significantly conflict with present views on affinities based on
the study of Recent taxa. These are summarized by Sibley & Ahlquist (1972) in
their study of non-passerine relationships based on egg-white protein electrophoresis.
There are various characters which appear to link the three Recent orders, and in
addition the Gaviiformes are considered to be closely related to the Charadriiformes
and the Sphenisciformes to the Procellariiformes. These two additional orders
should perhaps be included in any proposed major grouping. Sibley & Ahlquist
found evidence of similar starch gel egg-white patterns in Sphenisciformes, Gavii-
formes, Procellariiformes, Charadriiformes, and possibly the Pelecanidae, Fregatidae
and Phaethontidae. There is therefore evidence of a large and interrelated supra-
ordinal group, within which some earlier links may still be undiscovered.

While this hypothesis may help to establish the position of Prophaethon within the
framework of palaeontological phylogeny, the problem of fitting it into the existing
taxonomic hierarchy still remains. As we have already indicated, it does not show
definite affinity with any single existing taxon and to attempt to associate it with
one of these would affect ordinal definitions based on osteology. From a nomen-
clatural point of view the alternative treatment of the specimen seems preferable
whereby its isolated position is recognized. In terms of Recent taxonomy Pro-
phaethon constitutes an interordinal link. For nomenclatural purposes it seems
preferable to treat it as a monotypic order, Prophaethontiformes, with the single
family Prophaethontidae, while recognizing its special character. It is then possible
to insert it into the existing framework without unduly affecting existing taxa, and
it can be merged with another taxon at a later date if new and overriding evidence
should justify this. For the moment we would recommend that its placement
should be between the Pelecaniformes and the Charadriiformes.

Ma SYSTEMATIC DESCRIPTION

Order PROPHAETHONTIFORMES nov.

ORDINAL Diacnosis. Dorsal surface of skull with prominent postorbital process
and deep, rounded temporal fossae. Cranium small. Anterior lateral surface of
frontal flattened for attachment of unfused prefrontal. Deep fronto-nasal hinge
with prominent brow ridge. Lateral nasal struts lie alongside, and may fuse with,
rostrum just anterior to hinge. Elongated, oval depression in roof of orbit bordering
external edge. Zygoma laterally flattened with no dorso-ventral torsion at anterior
end. Rostrum tapering evenly with nares approaching schizorhinal condition but
terminating anterior to fronto-nasal hinge. Elongated nasal aperture extending
almost to rostral tip. Palate typically schizognathous. Quadrate with large
projecting flange on otic process, and posterior lower end of shaft with deep, narrow

28 REAPPRAISAL OF

groove. Lower mandible long and slender. Dentary with horizontal external
groove. Posterior mandibular fossa large, not perforating external wall. Pre-
articular large.

Carina of sternum deep, with curved ventral edge and anteriorly projecting apex
with well-developed manubrial spine. Flattened anterior surface at carinal apex
for articulation of furcula, and carinal tip bifurcated. Coracoid sulcus shallow with
poorly-developed ventral lip. Ventral labial prominence projecting and rounded.

Coracoid with large sterno-coracoid surface, slightly curved ventrally, and sternal
facet wide. Small facet for labial prominence on ventral surface. Procoracoid
stout, projecting laterally and curving anteriorly. Coracoid fenestra present.
Glenoid facet projecting dorso-ventrally.

Pelvis elongated and relatively narrow, with prominent median dorsal ridge, and
narrow posterior shield between raised posterior iliac crests. Fused synsacrum with
greatest depth at anterior end, tapering posteriorly. Last thoracic vertebra not
fused tosynsacrum. Anterior iliac plates sloping down steeply and curving outwards
to form broad, lateral flanges which project a little anterior to end of the synsacrum.

Femur with internal cotylar surface internally deflected towards distal end.
Proximal trochanter has abrupt obturator ridge with hollow on posterior side.
Trochanteric ridge prominent.

Inner cnemial crest of tibiotarsus extending proximally well beyond anterior
surfaces, as a thin flange projecting prominently, with some external curvature at
outer edge. Outer cnemial crest thicker distally than inner crest, with proximal
end terminating at an angle a little proximal to articulating surfaces. Internal
articular surface with a posteriorly projecting curved lip.

Family PROPHAETHONTIDAE nov.
Diacnosis. The only family of its order.

Genus PROPHAETHON Andrews 1899
DiaGnosis. The only genus of its family.
Type Species. Pyvophaethon shrubsolec Andrews.

Prophaethon shrubsolei Andrews 1899
(Pls 1-3; Figs 1-6)
1899 Prophaethon shrubsolei Andrews: 776-785, pl. 51.

Diacnosis. The only species of its genus.

HoLotyre. Imperfect skull, lower jaws and hyoids, sternum, right coracoid,
distal end of left coracoid, left scapula, right femur, proximal end of left tibiotarsus,
II vertebrae, rib fragments and synsacrum. In British Museum (Natural History),
Department of Palaeontology registered number A683.

LocALiTy AND Horizon. Lower Eocene, London Clay (Ypresian) of Sheppey,
Kent, England.

PROPHAETLHON SHRUBSOLET 29

VI. ACKNOWLEDGMENTS

We wish to thank Mr F. M. P. Howie for preparing the specimen, Miss M. L.
Holloway for making the detailed line drawings, and Mr F. Greenaway for taking the

photographs.

VII. REFERENCES

ANDREWS, C. W. 1899. On the remains of a new bird from the London Clay of Sheppey.

Proc. zool. Soc. Lond., 1899 : 776-785; pl. 51.

Howarp, H. 1929. The avifauna of the Emeryville shellmound. Univ. Calif. Publs Zool.,

Berkeley, 32 : 301-394.

1957. A gigantic ‘toothed’ marine bird from the Miocene of California. Bull. Dep.

Geol. S. Barbara Mus. nat. Hist., 1: 1-23, 8 figs.
Jotiiz, N. T. 1957. The head skeleton of the chicken and remarks on the anatomy of this
region in other birds. J. Morph., Boston, Mass., 100 : 389-436, 26 figs.
MatHeEews, G.M.& IREDALE, T. 1921. A Manual of the Birds of Australia, 1. 279 pp. London.
OtsENn, S. L. 1974. New fossil evidence of the origin of Frigatebirds. Abstr. 16th int. orn.

Congy., Canberra : 61-62.

PycraFtT, W. P. 1898. Contributions to the Osteology of Birds. Part I. Steganopodes.
Proc. zool. Soc. Lond., 1898 : 82-101, pls 7-8, figs 1-8.

SIBLEY, C. G. & AunLQuIST, J. E. 1972. A comparative study of the egg white protein of
non-passerine birds. Bull. Peabody Mus. nat. Hist., New Haven, Conn., 39: 1-246,

37 figs.

Vi. UN Dir xX

The page numbers of the principal references are printed in bold type. An asterisk (*)

denotes a figure.

Alcidae I9—20, 22
Andrews, C. A. 3-4

bony-toothed birds 25-6
Burhinidae 21-2

carina 23

Charadriiformes 3, 18—27

Chionidae 22

Columbiformes 21

comparisons with Recent material 18—24

Corcoid 4, 10*, 13—14,.17, 20—1, 23; pl. 2,
figs D-G

cranial structures 4—8, 24

Cretaceous, ordinal divisions of birds within

25
Diomedea 18-19

egg-white protein electrophoresis 27
evolution of birds 24—7

families, hypothetical relationship of 25, 26*
femur 4, 15, 18-109, 23
Fregatidae 20, 22, 24, 27

Galliformes 21
Gaviiformes 21-2, 27
Glareolidae 22
Greenaway, F. 29
Gruiformes 2I—2

Haematopus 21
Holloway, Miss M. L. 29
Elowie, FM. P. 4, 20

Ibidorhynchus 21
jaw, lower, see mandible

Lari 26

Laridae 20—2, 26

Larus 19, 21

limb-bones 4 ; see femur, etc.

Macronectes 19

mandible, lower 4, 9-11, 12*, 17, 21, 24;
pl. 2, figs A—C

measurements 16-18

nasal structures 21—2
nostrils 4

30 INDEX

Odontopterygidae 25
Odontopteryx toliapica 25
orbital structure 4—7
Osteodontornis orri 25

palate4)7*;'$=9) 21; pl>avigeB
Pelecanidae 20, 27

Pelecaniformes 3, 6, 19-27
Pelecanoididae 20

pelvis 4, 14-15, 18-19, 23; pl. 3, figs A-C

Phaethon 4, 19-23

Phaethontidae 3—4, 20, 22—4, 26-7

Phalacrocoracidae 20—2

Phalacrocovax 4, 19

Phoebetria 19

Pluvianus 22

Podicipitiformes 20—2

postcranial elements I1—16, 23

Procellariidae 21

Procellariiformes 3, 18—27

Pyrophaethon shrubsolei 1-29 passim, 28;
pls 1-3

Prophaethontidae fam. nov. 27, 28

Prophaethontiformes ord. nov. 3, 27-8

proto-frigatebird 24

Pseudodontornithidae 25

Puffinus 19

Dr C. J. O.. Harrison, Ph.D.
Subdepariment of Ornithology

BritisH Museum (NATURAL History)
TRING

HERTS

quadrate 8, 17

Rhynchops 22
ribs 4, 16
rostrum 9, 21-2

scapula 4, 14, 17

Sheppey, Isle of 3

Shrubsole, W. H. 3

skull 3, 4-11, 5*, 7*, 16-17, 21-4; pl. I,
figs A-E

Sphenisciformes 21, 27

Stercovarius 18, 22

Sterninae 26

sternum 4, 11-13, 12*, 18, 20, 23

Sula 4, 19

Sulidae 20, 23

Thinicoridae 22

tibiotarsus 4,.12*, 16, 18—10; 23877) ples,
figs D-E

Uria 22

vertebrae 4, 19, 23

X-rays 4

C. A. WALKER

Department of Palaeontology

British Museum (NATURAL HIsTory)
_CROMWELL ROAD

Lonpon SW7 5BD

Accepted for publication 18 July 1975

PLATE «

Prophaethon shrubsolei Andrews

BM(NH) Pal. Dept. no. A683.

Views of skull, x I.

HOOP

dorsal (p. 5)
ventral (p. 7, 8)
left lateral (p. 4)
right lateral (p. 4)
posterior (p. 7)

Bull. By. Mus. nat. Hist. (Geol.) 27, 1 JPL ANIC IB, a

PE AWGE. 2

Prophaethon shrubsolei Andrews

BM(NH) Pal. Dept. no. A683. Views of lower mandible, x I. (p. 9)
A. dorsal
B. ventral
C. left lateral

BM(NH) Pal. Dept. no. A683. Views of right coracoid, x 2. (p. 13)
D. dorsal
E. external
F. internal
G. ventral

Bull. Br. Mus. nat. Hist. (Geol.) 27, 1 PLATE 2

ee Ad) Ea
Prophaethon shrubsolei Andrews

BM(NH) Pal. Dept. no. A683. Views of pelvis, x 1. (p. 14)
A. dorsal
B. ventral
C. right lateral

BM(NH) Pal. Dept. no. A683. Views of proximal end of left tibiotarsus, x 2. (p. 16)
D. anterior
i internal

Bull, Br. Mus. nat. Hist. (Geol.) 27, 1 PLATE 3

arms
I ine

oe if

gph cee? aA
SL a - ‘
a Yt

A

as

n 2

ee

x Mines ar

LO:

LY

. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya.

. APPENDIX. DAVEY, R. J., Downiz, C., SARJEANT, W. A. S. & WILLIAMS, G. L?

. RHODES, F. H. T., Austin, R. L. & DRUCE, E. C. British Avonian (Carboni-

. Cuitps, A. Upper Jurassic Rhynchonellid Brachiopods from Northwestern Ee

. Goonys Po Cz lhe relationships of certain Upper Cretaceous Teleosts with — i

. Owen, H.G. Middle Albian Stratigraphy in the Anglo-Paris Basin. Pp.164;

. Sippigul, Q. A. Early Tertiary Ostracoda of the family Trachyleberididae A .

A LIST OF SUPPLEMENTS
TO THE GEOLOGICAL SERIES
OF THE BULLETIN OF S
THE BRITISH MUSEUM (NATURAL HISTORY) soy

Pp. 213; 30 Plates; 2 Text-figures. 1965. OUT OF PRINT.

. Ex-NacGar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper

Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley, ; -
Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. £11. na

. Davey, R. J., DowntE, C., SARJEANT, W. A. S. & WitttaMs, G. L. Studies on

Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 Plates; 64 Text-
figures. 1966. £8.20.

Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 24.

Wee

1969. 95P- Boe

. Etyiott, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) af a
the Middle East. Pp. 111 ; 24 Plates; 16 Text-figures. 1968. 6.10. rhe

ferous) Conodont ans and their value in local and continental correlation.
Pp. 313; 31 Plates; 92 Text-figures. 1969. £13.10. Pa Re

Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £5.25.

special reference to the Myctophoids. Pp. 255; 102 Text-figures. 19609. a
£7.70. ee

3 Plates; 52 Text-figures. 1971. £7.20.

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. f{9.60.
Forey, P. L. A revision of the elopiform fishes, fossil and Recent. Pp, 2225.

92 Text-figures. 1973. £11.35. ne
WitiAms, A. Ordovician Brachiopoda from the Shelve District, Shropshire. ake

Pp. 163; 28 Plates; 11 Text-figures; 110 Tables. 1974. £12.80.

Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS45NU _ hues

V , d
AS l ey X

GENERAL *
29 JUL 1976

A. J. SUTCLIFFE

ae. CAND,

_ BULLETIN OF.

NATURAL HISTORY
Vol. 27 No.

: 1976

Peas tOCENE RODENTS OF THE
BRITISH ISLES

BY
ANTONY JOHN SUTCLIFFE
British Museum (Natural History), London
AND

KAZIMIERZ KOWALSKI

Institute of Systematic and Experimental Zoology,
Polish Academy of Sciences, Krakéw, Poland

Php. 31-147 ; 31 Text-figures ; 13 Tables

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
GEOLOGY Vol. 27 No. 2
LONDON: 1976

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

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

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

This paper 1s Vol. 27, No. 2, of the Geological
(Palaeontological) series. The abbreviated titles of
periodicals cited follow those of the World List of
Scientific Periodicals.

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

ISSN 0007-1471

© Trustees of the British Museum (Natural History), 1976

BRITISH MUSEUM. (NATURAL HiSiaO RR Ye)

Issued 29 July, 1976 Price £7.40

i.

PEEISTOCENE RODENTS OF THE
BRITISH ISLES

By. J. SUTCLIBFPE & K: KOWAESKI

CONTENTS

SYNOPSIS

INTRODUCTION
A. HISTORY OF STUDIES
B. THE GEOLOGICAL BACKGROUND .

LOCALITIES IN THE BRITISH ISLES WITH FOSSIL RODENTS

A. Deposits OF East ANGLIA.
(i) Red Crag.
(ii) Icenian Crag : :
(ii1) Cromer Forest Bed Series .

(a) Pastonian of East Runton acd Happisburgh.

(b) Beestonian : ;

(c) Cromerian, sensu ee

(d) Anglian : 5 : :
B. TERRACE AND SOLIFLUXION DEPOSITS OF THE RIVERS THAMES

AND LEA
(i) Lower Thames .

(a) High Terrace of Susnsigese

(b) Clacton

(c) Early Middle Samene at hades Gene iaeeck

and Ilford ‘ ,

d) Late Middle Terrace at Gar cord ane Ban ;
e) Floodplain Terrace complex
f) The Middle Sai aa cae Beeaee aie
pper Thames . , ‘
(a) Isleworth .
(b) Marlow

(a) Water Hall aera Cel Pit

(b) Nazeing, Ponders End, Edmonton and See
(iv) Solifluxion and melt-water deposits ‘

(a) Northfleet, Ebbsfleet and Baker’s Bole

C. Cave DEPposits
(i) Westbury Fissure

) Kent’s Cavern
) Tornewton Cave

) Other caves with pre- Tpswichian? epesae
v) ‘Ipswichian’ cave deposits :

) Caves with post-Ipswichian deposits .

) Caves in Scotland
(viii) Caves in Ireland

D. OTHER LOCALITIES ;
(i) Cromerian localities .

Page

34 PLEISTOCENE RODENTS

) Hoxnian localities

) ‘Ipswichian’ localities
(iv) Last Glaciation localities
) Holocene localities

(vi) Scottish localities

III. CLASSIFICATION AND DISTRIBUTION OF RODENTS IN THE PLEISTOCENE
OF THE BRITISH ISLES
Family SciuRIDAE Brandt
Genus Sciuvus Linnaeus
Sciurus whiter Hinton
Sciurus vulgaris Linnaeus
Genus Spermophilus Cuvier ‘
Spermophilus (Urocitellus) phe earns :
Spermophilus (Colobotis) superciliosus (iKkaup)
Family GLIRIDAE Thomas
Genus Muscardinus Kaup .
Muscardinus avellanarius (Linnaeus)
Family CasTORIDAE Gray
Genus Tvogontherium Fischer
Trogontherium minus Newton .
Trogontherium boisvillett: (Laugel)
Genus Castor Linnaeus
Castor fiber Linnaeus
Family MurIDAE Gray .
Genus A podemus Kaup
A podemus sylvaticus (anaes)
A podemus flavicollis (Melchior)
Genus Micromys Dehm :
Micromys minutus (Pallas)
Genus Mus Linnaeus.
Mus musculus Linnaeus .
Genus Rattus Fischer.
Rattus vattus (Linnaeus) .
Rattus norvegicus (Berkenhout)
Family CricETIDAE Rochebrune
Genus Cricetus Leske.
Cricetus cricetus (Linnaeus)
Genus Allocricetus Schaub .
cf. Allocricetus bursae Schaub .
Genus Dicrostonyx Gloger . :
Dicrostonyx torquatus (Pallas) .
Genus Lemmus Link .
Lemmus lemmus (Linnaeus)
Genus Clethrionomys Tilesius
Clethrionomys glareolus (Schreber)
Genus Pliomys Méhely
Pliomys episcopalis Méhely
Genus Mimomys Major
Mimomys pliocaenicus Major
Mimomys rveidi Hinton
Mimomys newtont Major.
Mimomys savini Hinton .

OF BRITISH ISLES 35

Page
Genus Arvicola Lacépede . ‘ ‘ : , : 3 99
Arvicola cantiana (Hinton) , j : ‘ : : 102
Arvicola cantiana-terrestyvis 5 ; : ; : ; 103
Avrvicola tervestvis (Linnaeus) . ; : 5 : : 103
Genus Pitymys McMurtrie : : “ : : ; 105
Pitymys arvvaloides Hinton : ; ; : ‘ ; 106
Pitymys gregaloides Hinton , : : ‘ ; 107
Genus Microtus Schrank . ‘ ‘ : : . ‘ 107
Microtus avvalinus Hinton ; : ‘ : : : 108
Microtus agrestis (Linnaeus) . : : : ; : 109
Microtus arvalis (Pallas) : : : : A : £12
Microtus nivaloides Major ‘ : s 5 : , 114
Microtus nivalis (Martins) ‘ ‘ ‘ : ; 4 114
Microtus vatticepoides Hinton . ; 3 : , ; 116
Microtus oeconomus (Pallas) . : ‘ ‘ ; : eu)
Microtus gregalis (Pallas) ; 5 : P : ; 118
Genus Lagurus Gloger : : : : : : ; 120
Lagurus lagurus (Pallas) . : : ; : F ; 120

IV. HISTORY OF THE RODENT FAUNA OF THE BRITISH ISLES DURING THE
PLEISTOCENE . z : : : d : : : m2
Red Cragin. : 3 ; : : : E22
Icenian Crag and Bes cate Beret, Bed. ; 5 : : 522
Cromerian sensu stricto . ‘ ‘ ‘ : s . : 122
“Westbury stage’ . ; . : < ; : ; : 122
Mundesley Arctic Bed . ; : : : : : : 123
Hoxnian deposits . ‘ : : ‘ . ; ; 123,
Comparison of the rodent conn ‘ 123
Notes on the Wolstonian- Sp eelia pane of fe succession : 124
Middle Terrace of the Thames interglacial : : : : 125
Tornewton Cave Glutton Stratum cold stage . ‘ : é 125
Joint Mitnor interglacial : : . ‘ : 3 : 126
Last Glaciation . : : ‘ : ‘ : : , 127
Elolocene .. ‘ 3 ; : A : 5 : : 128
Ireland : : : : ‘ ; ; é . : 129
V. ACKNOWLEDGEMENTS : 2 : : ‘ : ‘ : 129
VI. REFERENCES . : 2 : , ; : ; : F 130
VS “INDEX . : : ‘ : : : : : , : 027

SYNOPSIS

For nearly half a century, since its publication in 1926, M. A. C. Hinton’s Monograph of the
Voles and Lemmings has remained the only comprehensive work on British Pleistocene rodents.
Subsequent advances in Quaternary studies and the availability of new fossil material have made
some updating of this publication necessary. In the present work a brief historical review of
the studies of British Pleistocene rodents is followed by a description of the rodent assemblages
from the principal fossil localities. Rodent species recorded from the British Pleistocene are
then discussed in systematic order and, finally, an attempt is made to relate a generalized sequence
of rodent faunas to the climatic sequence usually employed today as the basis for British
Pleistocene chronology.

36 PLEISTOCENE RODENTS

I. INTRODUCTION

NEARLY half a century has elapsed since the publication by the British Museum
(Natural History) of M. A. C. Hinton’s Monograph of the Voles and Lemmings in
1926. Today it is still the only comprehensive work on British Pleistocene rodents
and remains the standard reference work on this subject. In the meantime there
have been great advances in many related fields of study. Hinton did not attempt
to relate his sequence of rodent faunas to the glacial—interglacial sequence of climatic
events, which was already gaining favour at the time of his studies, and which is
known today to be even more complicated than the four glacial and three interglacial
stages for so long employed as a basis of Pleistocene chronology.

Excavations in the British Isles since 1926 have produced a wealth of new fossil
rodent material which was not available to Hinton. Findings from the Westbury-
sub-Mendip Fissure (Somerset), Swanscombe (Kent), Tornewton Cave (Devon) and
a series of Late Glacial cave sites in the Peak District (Staffordshire and Derbyshire),
excavated and published by various workers, are of specialimportance. The advent
of carbon-14 (14C) dating has permitted the accurate dating of some Upper Pleisto-
cene rodent remains. Studies of pollen and insects are further, relatively new,
stratigraphic aids.

On the continent of Europe there have been great advances in rodent studies
which provide a basis for the interpretation of rodent migrations into the British
Isles and assist in identifying stages of geographical isolation. W.von Koenigswald’s
recent work (1972, 1973) on the Mimomys-—Arvicola lineage, based initially on
continental European remains and subsequently extended to the British Isles,
throws new light on the British Pleistocene rodent sequence.

In the present work the authors have cooperated on a reappraisal of the Pleistocene
rodent faunas of the British Isles. One of us (K. Kowalski of Krakow, Poland) is
mainly responsible for the systematic part and for comparison with continental
evidence ; the other (A. J. Sutcliffe of London) for the stratigraphic sequence of
British rodent faunas. Discussions with W. von Koenigswald (Tubingen) about the
Mimomys—Arvicola lineage have contributed fundamentally to the last-mentioned
part of this work.

Although we cover 37 species of rodents from over 100 sites, many problems re-
main which can only be resolved by further careful collecting. Additional material
is needed from critical deposits such as the early and late parts of the Forest Bed
Series of Norfolk and from Hinton’s Middle Terraces of the Thames, or from deposits
of equivalent age. Some of the faunal assemblages which we have listed, such as
that from the Otter Stratum of Tornewton Cave, are unique and cannot be compared
with deposits of the same age in other parts of the British Isles. We have shown
scarcely any evidence of differences between contemporary faunas in different areas,
nor of temporal faunal changes within the broad climatic phases used for reference.
The systematic relationship between some of the rodent species is still far from clear.
All these aspects of rodent studies and many others offer great scope for future
research. If this Bulletin provides a stepping stone for such work it will have been
worth while.

OF BRITISH ISLES 37
A. History of Studies

Although publications of the eighteenth century and, in greater number, of the
first decades of the nineteenth contain references to the occurrence of remains of
beaver in British Quaternary deposits, the first scientific description of fossil rodents
from this country did not appear until the middle of the nineteenth century. In
1846 R. Owen published a detailed description, accompanied by excellent drawings,
of the remains of Tvogontherium cuviert and Castor fiber from the Forest Bed deposits
of East Anglia. He also mentioned the presence of the remains of Avvicola ‘which I
have been unable satisfactorily to distinguish from Avvicola amphibia, or common
Water-rat’ (1846 : 202) in the sediments of many caves. The bones of the same
species were present, according to Owen ‘in newer pliocene deposits’. He considered
(1846 : 205) the specimens from the ‘older pliocene crag near Norwich . . . indicated
a species of Avvicola intermediate in size between the Water-vole (Arvicola amphibia)
and the Field-vole (Avvicola arvalis)’.

The next important contribution to the study of British Pleistocene rodents was
that of W. A. Sanford. His papers on rodents from caves of Somerset (1870a, b)
contain not only descriptions of forms still living in Britain (Arvicola terrestris,
Clethrionomys glareolus and Microtus agrestis) but also of species now extinct in this
country (Microtus oeconomus, Lemmus ‘of the type of L. norvegicus’ and Dicrostonyx
torquatus). It is interesting to note that Sanford correctly determined the skulls of
D. torquatus but ascribed jaws of the same species to Avvicola gulielam. He men-
tioned also the occurrence of a small hamster, ‘Cvicetus songarus (Pallas)’, and
Spermophilus erythrogenoides, a species previously discovered and described by H.
Falconer (in Murchison 1868).

The first attempt to make a systematic classification of fossil voles, one of the
most difficult groups of rodents, was made by H. P. Blackmore & E. R. Alston
(1874). They failed to recognize the difference between Recent and fossil species,
but were the first to record Mzicrotus nivalis in Britain.

A series of works on the ‘preglacial’ deposits of East Anglia by E. T. Newton
(1881, 1882a, b, 1890b, 1891 and 1909) provides an important contribution to the
knowledge of the rodents of this area. Newton was the first to state that the vole
remains were specifically different from Recent forms and he named a new species,
Arvicola intermedius Newton 1881. He also published diagnoses of two other new
species of rodents, a hamster, ‘Cricetus vulgaris Runtonenstis n. subsp.’ and a beaver
‘Trogontherium minus n. sp.’ His work is of importance in regarding the vertebrate
remains, including many rodents, from the ‘pre-glacial’ deposits of East Anglia as
specifically distinct from living forms. He also published the first scientific descrip-
tion (1894, 1899a, b) of the vertebrate remains, including rodents, from the rich
late Pleistocene fauna of Ightham Fissures, Kent.

The greatest contribution to the study of British Pleistocene rodents was that of
M. A. C. Hinton (1883-1961). His first note on this subject appeared in 1900 and
later this group of mammals became the main object of his interest. His work
culminated in 1926 with the publication of volume 1 of his Monograph of the Voles
and Lemmings. Winton soon became the world authority on Microtinae and his
monograph (unfortunately unfinished) is still the standard work in this branch of

38 PLEISTOCENE RODENTS

mammalogy. He was the first to recognize that, since rodent species have short
ranges in time, their remains are of special stratigraphic value in the study of the
Pleistocene. They thus help the geologist attempting to correlate scattered or
isolated cavern deposits with others to which ordinary stratigraphical methods can be
applied. The importance of this concept was further expanded by Kowalski (1966).
Hinton (1926a ; 1926b : 126-136) recognized a series of rodent faunas of different
ages at British localities. In order of increasing antiquity these were as follows :

Third Terrace of the Thames, in the valley of the River Lea.

Ightham Fissure (Kent) stage.

Late Middle Terrace of the Thames (typical locality Crayford and Erith, Kent).
Early Middle Terrace of the Thames (typical locality Grays Thurrock, Essex).
High Terrace of the Thames (Ingress Vale, Swanscombe, Kent).

Upper Freshwater Bed at West Runton, Norfolk.

Shelly Crag at East Runton, Norfolk.

Norwich Crag and Weybourne Crag of the Norfolk Coast.

Although Hinton did not attempt to relate this division to any sequence of climatic
fluctuations, his faunal stages were far ahead of any other palaeontological division
of the British Pleistocene at that time.

Hinton also laid the basis for systematic study of the Microtinae (Savage 1963).
After 1926 he abandoned nearly completely his study of the British fossil rodents,
being involved in problems of pest control and the study of other mammals. He
nevertheless published in 1952 a thorough description of rodent remains from the
Late Glacial deposits of the Lea Valley.

Hinton’s work is not only an important chapter in the history of zoology, but it
still forms the standard reference on British Pleistocene rodents. For this reason it
seems necessary to mention some of its limitations. He was a monoglacialist, or
strictly an antiglacialist, as he postulated only one very late and only moderately
cold period in the Pleistocene history of England. This opinion, which he regarded
as the established truth, probably also influenced his zoological views. When he
came across the fossil remains of species associated today with arctic climate he
was inclined to determine them as forms different from Recent ones, since they
could not then be used as evidence of particular climatic conditions.

Hinton was a prominent typologist. Although he was not more of a ‘splitter’
than most of his contemporary zoologists, he nevertheless used subspecific names
extensively and was inclined to determine each new variant as a new species. For
this reason, even in Hinton’s own collection, many intermediate specimens were left
undetermined.

Finally, Hinton accepted the multituberculate origin of rodents, a theory now
generally rejected. He accepted that microtine evolution has always led from
forms with complicated to forms with simple teeth. Strangely enough this did not
hinder him in the correct reconstruction of some of the lines of evolution in voles,
since among the different tendencies in the late evolution of the teeth of this par-
ticular group some lines did indeed progress from more complicated to simpler
tooth-patterns.

QF IIR Ist MSILIBS 39

Relatively few contributions to the study of British Pleistocene rodents have
been published since the appearance of Hinton’s monograph. These include the
work of J. W. Jackson (1932, 1934); L. S. Palmer (1934); A. Schreuder (1929,
IQ3I, 1950, 1951); J. N. Carreck (1957, 1966) ; D. Bramwell (1960, 1964, 1970) ;
meen outcine & I BE. Zeuner (1962); J. C. Pernetta (1966); R. J. G. Savage
(1966) ; K. Kowalski (1967) ; W. von Koenigswald (1973) and M. Bishop (1974,
1975). Most of these papers contain site lists of rodents with relatively little general
discussion of their stratigraphical or systematic position. The papers by Schreuder
(1929, I93I, 1951) are worthy of special mention for containing a thorough re-
description and systematic discussion of the beaver remains from Britain. The
paper of 1950 contains a list of the voles from the stratigraphically important
Middle Pleistocene deposits of Swanscombe, Kent. More recently A. J. Stuart, ina
general review of the British Pleistocene fauna (1974), listed the rodent faunas from
key sites and discussed their stratigraphical occurrence.

Since the last synthetic work of Hinton (1926b) there has been great progress in
other fields of study related to the problem of the history of British rodents. Investi-
gations on the continent of Europe (not without influence from Hinton’s work) have
brought to light many rich local rodent faunas of Pliocene and Pleistocene age and
have made it possible to reconstruct in greater detail the stratigraphical position
and evolutionary history of many of these mammals. Recent contributions in the
field include the work of J. Chaline (1972), who gives a detailed description of the
rodents of the Middle and Upper Pleistocene of France, and of W. von Koenigswald,
who studied the phylogenetic lineage of the genus Avvicola. In addition, knowledge
of the systematic position, ecology and geographical variability of Recent rodents
has greatly improved throughout the whole of the Palaearctic region.

In the British Isles there have been extensive geological and palaeobotanical
studies (e.g. West 1968), which have provided a more detailed stratigraphic back-
ground for rodent studies. In addition the stratigraphic position of many old
localities has been clarified and new rodent localities have been discovered. A
new aspect of investigation has been provided by zoological studies of insular races
of small mammals (Corbet 1961).

In view of the increasing importance of rodents in stratigraphic studies and of all
the additional information which has become available since the publication of
Hinton’s monograph, it is timely to review the existing data on British Pleistocene
rodents. In the description which follows the stratigraphic age of crucial rodent
localities of the British Isles will be further examined, the systematic position and
synonymy of the rodent species, currently overburdened with too many names
connected with insufficiently characterized forms, will be discussed and the history
of the rodent population of Britain during the Pleistocene reviewed. Many gaps
nevertheless remain in the known sequence of events. This paper does not contain
morphological descriptions of the various species of British Pleistocene rodents, a
task, important for further stratigraphic and systematic studies, which needs to be
carried out in the future. In the meanwhile subspecific designations are not used
in the present paper. The nomenclature of living species of rodents here used is
based mainly on the work of Ellerman & Morrison-Scott (1966). Information about

40 PLEISTOCENE RODENTS

morphology, biology and distribution of Recent rodents in Britain is available in
Miller (1912), Matthews (1952), Southern (1964) and Corbet (1966). Beirne (1947)
has discussed the possible arrival dates of some species of rodents into the British
Isles.

B. The Geological Background

As previously mentioned, great progress has been made towards a more detailed
understanding of the British Pleistocene sequence since Hinton published his mono-
graph in 1926. Some deposits considered to be Pliocene at that time are now
regarded as Lower Pleistocene, and there now exists a detailed picture, still being
further elaborated from current studies, of alternating glacial and interglacial
stages in the British Isles. A recent contribution of special importance is that of
West & Wilson (1966), who demonstrated from palaeobotanical studies that the
Cromer Forest Bed Series of Norfolk in fact represents two interglacial stages with
an intervening cold stage.

At the present time the most generally accepted correlation for the British
Pleistocene is that recommended by the Geological Society of London (Mitchell e¢ al.
1973), which is shown, with some additional information from other sources indicated,
in Table 1.

While in general it seems possible to relate most British rodent faunas to this
sequence, there are some parts (notably the Cromerian and Wolstonian—Ipswichian
stages) which seem inadequate to account for all the rodent stages currently
attributed to them. There can be little doubt that they are more complex than is
indicated in the table opposite.

Although, in the 1973 correlation of the Geological Society, the Wolstonian and
Ipswichian are recognized as two stages only, an earlier correlation published by the
same Society (Evans 1971) proposed a more detailed chronology for this part of
Pleistocene time which must be mentioned here, since it has important application
to our rodent studies.

From a consideration of information derived from deep-sea cores and other lines
of evidence Evans argued that there had been more than one warm phase since the
Hoxnian (Holsteinian) Interglacial. Whilst accepting the period from about 100
to 70 thousand years ago (which he regarded as the true Last Interglacial, the
Ipswichian or Eemian, equivalent to Zeuner’s Late Monastirian shoreline) as the
only fully warm period of considerable length since the Holsteinian, he drew atten-
tion to other lesser mild phases of post-Holsteinian age about 170 000 and 130 000
years ago, possibly equivalent to the Danish Vejlby I and Vejlby II mild stages, the
possible relationship of which is shown in Table 2.

Evans also drew attention to some problems related to the interpretation of this
part of the Pleistocene sequence. He pointed out that Zeuner, who regarded the
Last Interglacial as double, used the term ‘Last Interglacial’ for two different ranges
of time, that lasting from 180 to 120 thousand years ago (cycles 5w and 4) and also
for that lasting from 130 to 70 thousand years ago (cycles 4w and 3). He considered
that some confusion had arisen in archaeological circles where ‘Last Interglacial’
had been used as a reference datum and preferred to restrict this term only to zone 3w.

41

OF BRIGISH ISLES

TABLE t

Generalized British Pleistocene sequence

(‘qenrseg = q ‘eNeH = 5H ‘urewig :ozeu[D Jopuy))

Il9y} ye uMOYS a1e (gz61) sa8ejys JUBPOI s,UOPUT}Y ‘€Z6Y 4saA\ ® uoZOYS ‘Auueg ‘[[ayouPY pue ‘zZ61 ysayq ® syIedS ‘gg6I ysoAA UO paseg

*21q42} 9} JO UUINTOS 4seI OY} UT UOT}ISOd AjexT] ysOu

Ee en en a ae pa Sa. ee
DVD
ANITIVUOD

oe LILLIA PSAP AI AIPA

sseI9
auinoqAej\ pue YOIMION

uojuNY yseq ye BeID AT[EYS

U0JUNYT FSO
ze peg Jozemyso1,y reddy

(Arepunoq vua90}sIa[q-Ol[q $,U0}UTH)

sowley oy} JO soelIey ysl

souleyy

ay} JO sovIIOy, app ATIeA
souley

OY} JO 9OVIIOL S[PPHAL 93eT

O8E}S IINssLy WeYYST
AgyeA eaT 94}
UI SOWeY YT, OY} JO 9OeIIOT PIL

uortsod orydeisyers a1quqoig
(9z61) seune.y JUapo, s,u0pUTET

(uotsolea)

OVA AIY

(sovag
HOIMUON

x»
ANUNOPGATM)
Dvug
NVINGOT

SaIaaS

aag
ISTO
UAWOUD

SLisodaqd
TVIANTIY
x

aovura
AAV)

s}isodap

¥

(Aje}] JO ueysy)

‘dwoy UeIIGANOYT HNHOOVId
RviNOnAVAN (Aye}z] JO ueryourszeltA “T)
UeI[SIL-9Ig
‘duay, NVINVHGNT io ee i
_-_- é _ —
12) (ee) NVINYAH (Ayeaz Jo
uvIyourssertA “1)
‘dura NVIINY ueHeEL
uetuoiIngy ANAHOOLSIW Id
(d) UAMOT
PIoD NVILNAAVG LenB
‘duioy NVINOLSVq uvideueyyl
(d) NVINOLSaag vee
PIOD
dua NVINANOUD
(q+) (NvILd01SaM07T) ueleys[q ¢ =ANAOOLSIATd
10) (el) NVITIONY URITELS ¢ HICGCCIN
‘dworp NVINXO}] UeIUIOJS[OHL
(q+) (NVIDNIddI5) (sxed) uerqees ;
1) (9) NVINOLSTOM
‘duo, NVIHOIMSd] uelulay
ANAOOLSIA Id
a waddf
(d+5)
NVISNAAG uvITISYIO
plop v d TOSYSTOM,
‘dway NVIYANV TL euss0[0H ANHOOTOH
og) ee SS
urequg urezyIg S edoing MN
Deyewlyg > SO8PYS 1 S98PIS

ystig

42 PLEISTOCENE RODENTS

TABLE 2

Possible relationship of continental climatic divisions.to British mild phases (after Evans 1971)

Evans’ half-cycles Years ee: Continental Mediterranean
(c = cold, w = warm) ago 6 climatic divisions sea levels
names
2c 70 000 Early Weichsel, Early Wirm
3W 100 000 IpSWICHIAN Eemian Late Monastirian
3c I20 000 Saale 2, Riss 2, Warthe
4W 130 000 Vejlby 2 Main Monastirian
4C Minor cold phase
5w 180 000 Vejlby 1, ? Domnitz
5c 200 000 Saale 1, Riss 1
ow Hoxnian Holsteinian

It follows from Evans’ chronology that the terms Saale 2, Riss 2 and Penultimate
Glaciation of some authors have been used for periods of time later than Last Inter-
glacial of some other authors, and that great confusion can arise here if stage names
such as these are applied incautiously to our studies of British Pleistocene rodents.
Likewise the terms Penultimate Glaciation or Wolstonian (when applied away from
the type locality) could confusingly be used to mean any period of time from that
considered by some writers to be the middle of the Hoxnian/Holsteinian (for ex-
ample, Mullender’s suggestion in Wymer (1974) that the upper part of Lower Loam
of Swanscombe is Wolstonian) to that immediately preceding Evans’ Interglacial
half-cycle 3w. Wymer (1974) has suggested a warm stage during the Wolstonian.
Bristow & Cox (in Mitchell e¢ al. 1973) have argued, from their study of glacial
deposits in East Anglia, that the interglacial deposits at the Hoxnian type locality
of Hoxne belong to the last and not penultimate interglacial and they referred the
Ipswichian and Hoxnian deposits, which they accepted may have been separated
by a cold oscillation, to a single interglacial between the Devensian and Anglian.

In the present paper we will try to overcome such problems, as far as possible,
by concentrating on establishing a relative chronology for the rodent faunas of the
British Isles which will not be affected by future refinements in the naming of the
British Pleistocene sequence, rather than to attempt to refer these faunas too rigidly
to a chronology which is clearly incomplete. Let us now make a detailed examina-
tion of the rodent faunas from the various British Pleistocene localities.

Il, LOCALITIES IN THE BRITISH TSEBRBS With. POSSIEsRODENES

Remains of Pleistocene rodents have been found at many localities in the British
Isles, notably in the early Pleistocene marine Crag and Forest Bed deposits of East
Anglia and in later river terrace and cave deposits. In Ireland they are known only
from cave deposits of late Pleistocene age. The location of the principal localities
is shown in Fig. 1.

A. Deposits of East Anglia

Extensive areas of Suffolk and Norfolk are covered by Crag and Forest Bed
deposits from which rodent remains have sometimes been recovered.

OF BRITISH, ISLES 43

Although recent research (West & Wilson 1966, Norton 1967, West 1968) has
greatly increased understanding of the early Pleistocene sequence of East Anglia,
all the rodent remains from this region available for the present study are unfor-
tunately from old collections, many of them with imptecise stratigraphic information.
A series of rodent faunas can nevertheless be distinguished.

(i) RED CraG. This is a marine shallow-water shore deposit typically laid down
in land-locked bays. It is best developed in Suffolk where it consists mainly of
shelly sands. There is also a basal nodule bed with rolled and polished fossils,
including mastodon teeth, apparently derived by the Crag sea from earlier deposits.
The occurrence of these derived fossils, which include mammalian remains of Eocene,
Miocene and Pliocene age, makes the study of the Red Crag mammalian fauna
extremely difficult. For a long time the deposit was considered to be of Pliocene
age and it was not until 1948 that its Lower Pleistocene date was accepted and the
occurrence of relatively unmineralized contemporary mammalian remains, associated
with the derived fossils, was recognized.

Only a few rodent remains have been found in the Red Crag. Tvogontherium
minus (known from Astian localities on the European continent) and Hystrix,
represented by teeth (Spencer 1966) and by gnawing on part of a deer antler (Sutcliffe
& Collings 1972), are probably derived from pre-Pleistocene deposits and cannot be
included in the contemporary Red Crag faunal list.

Castor fiber, recorded from a number of Red Crag localities, including Sutton and
Woodbridge (47, Fig. 1) is probably a contemporary species. A rolled microtine
tooth from the Red Crag has been provisionally referred by Spencer (1964) to
Mimomys sp.

(ii) IceENIAN Crac. A further series of marine deposits, the Icenian Crag (in-
cluding the Norwich and Weybourne Crags), occupies a basin on the north of the
Red Crag outcrop. It was apparently laid down in a more open sea. Its exact
relationship to the Red Crag is not fully understood, but its age appears to range
from Thurnian, through Antian and Baventian, to Pastonian. According to West
(1968), these stages may be equivalent to the Tiglian, Eburonian and Waalian in
the stratigraphy of the Netherlands. The relationship between the Norwich and
Weybourne Crags is also not clear. The Weybourne Crag, typified by the mollusc
Macoma balihica, is found at more than one stage. The youngest part of the
sequence, including part of the Weybourne Crag, is of Pastonian age and is con-
temporaneous with the lowest part of the Cromer Forest Bed series. Part of the
Norwich Crag of Suffolk may also be Pastonian.

Rodent remains, which have been found in the Norwich and Weybourne Crags of
Bramerton (55, Fig. 1), Covehithe (51), Easton Bavents (50), Sizewell (49), Thorpe
(48), Trimingham (61) and other localities represent Mimomys pliocaenicus, M.
veidi, M. newtont, Castor fiber and Trogontherium boisvilletti. Since conditions were
unsuitable for the preservation of very small remains, it is not surprising that no
teeth of glirids and murids have been found in the deposits. M. pliocaentcus and
M. reidi are typical Tiglian (Upper Villafranchian) elements and confirm a Lower
Pleistocene age for at least part of this series.

44

PLEISTOCENE RODENTS

82
81
e
80
e
77
76e
©
79 75
e78
74
AS
68 vA €73
067 . e@72
ee 65
& 270
6g 56
oor
e
52
sae >> $3)
20 oa ri
47
@40 46
el?
39 ° eo)
AG . 42
¢ 037 43
e
15 21 e 360 3, 3
6 25 5; HY
= or:
14 e
969 91) 23 ue” 2
ry HOeP
o8
se
22
a.
Zoe
Je x

1

Fic. 1. Location map of the principal Quaternary rodent localities in the British Isles.
Key to numbers opposite.

I Joint Mitnor Cave,
Buckfastleigh, Devon

2 Levaton Cave, Torbryan,
Devon

3 Tornewton Cave,

Torbryan, Devon

Brixham Cave, Brixham,

4
Devon

5 Happaway Cave,
Torquay, Devon

6 Kent’s Cavern, Torquay,

Devon

7 (i) Cow Cave, Chudleigh,
Devon

7 (ii) Chudleigh Fissure, Devon

8 Huntspill Cut, Huntspill,
Somerset

9 (i) Brean Down, nr Brean,
Somerset

g (ii) Uphill Cave, Uphill,

Somerset
Hay Wood Rockshelter,
Hutton, Somerset

(
(
(
Somerset
10 (ili) Picken’s Hole,
nr Bleadon, Somerset
(
(
(

10 (iv) Hutton Cave, Hutton,
Somerset

10 (v) Banwell Cave, Somerset

Ir (i) Rowberrow Cavern,

Burrington, Somerset
II (ii) Aveline’s Hole,
Burrington, Somerset
WZ Gough’s Cave, Cheddar,
Somerset
13 Westbury-sub-Mendip
Fissure, Somerset

14 Clevedon Cave, Clevedon,
Somerset

15 Alveston Fissure, Alveston
Gloucestershire

16 (i) Minchin Hole, nr Penard,
Glamorganshire

16 (ii) Bacon Hole, nr Penard,
Glamorganshire

17 (i) King Arthur’s Cave,
nr Whitchurch,
Herefordshire

17 (ii) Great Doward Cave,
Whitchurch,

’ Herefordshire
18 Merlin’s Cave, Symond’s

Yat, Herefordshire
19 Beckford,
Worcestershire
20 Upton Warren,
nr Droitwich,
Worcestershire
21 Sugworth, Oxfordshire
22 Fisherton, Salisbury,
Wiltshire
23 Thatcham, Berkshire
West Wittering, Sussex
24 (ii) Selsey, Sussex

QUA JE RAEITIES ISL IUSIEIBS

KEY TO FIG. I

25 Marlow, Buckingham-
shire

26 Isleworth, Middlesex

27 Crayford and Erith, Kent

28 Ightham Fissures,
Ightham, Kent

29 Northfleet, Kent

30 Swanscombe, Kent

31 Upnor, Kent

Be Murston, Kent

33 Grays Thurrock, Essex

34 Aveley, Essex

35 Ilford, Essex

36 (i) Hackney, London

36 (ui) Angel Road, Middlesex

(north London)

36 (111) Ponders End, Middlesex

37 Nazeing, Lea Valley,
Essex

38 Water Hall Farm,
Hertfordshire

39 Hitchin, Hertfordshire

40 Barrington,
Cambridgeshire

41 Cambridge Fens
(including Burwell and
Swaffley), Cambridge-
shire

42 Copford, nr Colchester,
Essex

43 Clacton, Essex

44 (i) Harkstead, Suffolk

44 (li) Stutton, Suffolk

45 Felixstowe, Suffolk

46 (1) Bobbitshole, nr Ipswich,

Suffolk

46 (ii) Stoke Tunnel Beds,

47 (i)

Ipswich, Suffolk
Sutton, Suffolk

47 (ii) Woodbridge, Suffolk
47 (ili) Kyson, Suffolk

48

49
50
51
52
3)3)

Thorpe, Norfolk
(see p.g6)
Sizewell, Suffolk
Easton Bavents, Suffolk
Covehithe, Suffolk
Kessingland, Suffolk
Geldeston, nr Beccles,
Norfolk
Hoxne, nr Diss, Suffolk
Bramerton, Norfolk
Happisburgh, Norfolk
Ostend, Norfolk
Bacton, Norfolk
Paston, Norfolk
Mundesley, Norfolk
Trimingham, Norfolk
Overstrand, Norfolk
Cromer, Norfolk
East Runton, Norfolk
West Runton, Norfolk
Swanton Morley, Norfolk
Lynx Cave, Denbighshire
(exact location not
known)

7I (i)

45

Gwaenysgor Cave,
Prestatyn, Flintshire

Elder Bush Cave,
Wetton, Staffordshire

Harborough Cave,
Brassington, Derbyshire

Fox Hole Cave, High
Wheeldon Hill, Earl
Sterndale, nr Buxton,
Derbyshire

71 (ii) Etches’ Cave, Dowel Dale,

Earl Sterndale, nr
Buxton, Derbyshire

71 (iii) Dowel Cave, Dowel Dale,

Zz

80

81
82

83

84

85
86

87
88

89

90

gI

Earl Sterndale, nr
Buxton, Derbyshire

Langwith Cave, Upper
Langwith, nr Scarcliffe,
Derbyshire

Pin Hole Cave, Creswell,
Derbyshire

Hessle, nr Kingston-
upon-Hull, Yorkshire

Staple Howe, nr
Wintringham,
Yorkshire

Star Can) mr
Scarborough,
Yorkshire

Kirkdale Cave, Helmsley,
Yorkshire

Cowside Cave No. 3,
Settle, Yorkshire

Dog Holes Cave,
Warton Crag, nr
Carnforth, Lancashire

Middlestots Bog, Edrom
Parish, Berwickshire,
Scotland

Corstorphine, nr
Edinburgh, Midlothian

Loch of Marlee, Kinlock,
Perthshire

Creag nan Uamh Cave,
Inchnadamph,
Sutherland

Keshcorran Caves,
Ballymote, Co. Sligo,
Republic of Ireland

Edenvale Caves, nr Ennis,
Co. Clare

Red Cellar Cave, nr
Lough Gur,
Co. Limerick

Castlepook Cave,
Doneraile, Co. Cork

Castletownroche Cave,
Connaberry, Co. Cork

Kilgreany Cave,
Cappagh,
Co. Waterford

Ballynamintra Cave,
Whitechurch, Co.
Waterford

Nornour, Isles of Scilly

40 PILE TSTOGE NE RODIN aS

gham
sMundesley

HP Felixtowe

Weybourne Crag & Forest Bed @
Norwich Crag A

Red Crag @
Fic. 2. Location map of Crag and Forest Bed rodent localities in East Anglia.

(111) CROMER Forest BED SERIES. This differs from the earlier Crag deposits in
being predominantly estuarine and freshwater, with some beach deposits. Coastal
exposures occur at many localities along the Norfolk coast, though these are un-
fortunately becoming increasingly obscured by sea defence work. Abundant
remains of mammals (including rodents) and of plants occur in the Forest Bed. The
most important rodent localities are West and East Runton (65, 64), with fewer
remains from Ostend (57), Cromer (63), Overstrand (62), Mundesley (60), Paston
(59), Bacton (58) and Kessingland (52).

It has long been recognized that all the Forest Bed deposits are not contem-
poraneous but that they accumulated over a considerable period of time, during
which there occurred changes of climate and fauna. West & Wilson (1966), basing

OF BRITISH ISLES 47

their conclusions on a study of plant remains, identified four climatic stages within
the Forest Bed Series of deposits. In descending order these are:

Early Anglian (cold, followed by glacial conditions
during the later Anglian)

Cromerian sensu stricto (warm)

Beestonian (cold)

Pastonian (warm)

Rodent remains are known from all horizons except the Beestonian. The Ostend
Forest Bed presents a problem which is further discussed on pp. 48, 122.

(a) Pastonian of East Runton and Happisburgh. Only the earliest part of the
Forest Bed sequence is represented at East Runton (64) ; deposits of Cromerian age,
sensu stricto, apparently being absent. Azzaroli (1953) observed that the large
mammals from this locality are Villafranchian species and that mammals present
in the later part of the Forest Bed are absent. West (personal communication,
1972) considers, from palaeobotanical studies, that no Cromerian deposits, sensu
stricto, have yet been proved on the foreshore at East Runton. He found, in
descending order, the following sequence :

c. Clay Conglomerate, being reworked estuarine sediment, probably of Pastonian
age, redeposited during late Pastonian or Beestonian times.

b. Shelly Crag, of Pastonian age, regarded by Hinton (1926b : 365) as Weybourne
Crag.

a. Flint Bed, relating to a pre-Pastonian land surface.

Unfortunately the East Runton section has not been seen really well for many
years and all the rodent remains available for the present study are from old collec-
tions. Three species of Mimomys are recorded; these are M. newtom, M. plio-
caenicus and M. savini. The occurrence, together in the Shelly Crag, of the last
two species, which represent successive stages in the development of the same
phyletic line of voles, suggests the natural or accidental mixing of elements from
different layers. M. pliocaenicus and M. newtoni are represented in both the Shelly
Crag and among the ‘clay pebbles’ of East Runton, the holotype of MW. newtont being
from the former horizon.

The precise stratigraphic range of these three species is far from clear. M. savini
(represented by a few specimens from the Shelly Crag) is very abundant in the later,
Cromerian sensu stricto, deposits of West Runton. The occurrence of remains of
M. pliocaenicus and M. newtoni (known also from the earlier Norwich Crag) in the
‘clay pebbles’ could mean that these rodents persisted into Pastonian times. Such
an interpretation would not be out of harmony with the ‘Villafranchian’ megafauna
of this locality. The rolled condition of some of the remains could alternatively
indicate their derivation from an earlier deposit.

Trogontherium borsvillettt and Castor fiber have also been found at East Runton.

A second Forest Bed locality, where the deposits are apparently of Pastonian
age only, is Happisburgh (56). Remains of Castor fiber and also fir cones bearing

5)

48 PLEISTOCENE RODENTS

marks which appear to indicate gnawing by squirrels (Newton 1882a) have been
found there.

(b) Beestonian. No rodent remains are known from Beestonian deposits.

(c) Cromerian sensu stricto. The Upper Freshwater Bed of West Runton (65),
defined by West (1961) as the type deposit of the Cromerian interglacial, contains
abundant rodent remains. The following species are represented: Apodemus
sylvaticus, Trogonthertum boisvilletti, Castor fiber, Cricetus cricetus, Mimomys savini,
Clethrionomys glareolus, Pitymys arvaloides, P. gregaloides, Microtus arvalinus, M.
nivaloides and M. ratticepoides. Muscardinus recorded by Hinton (in Barrett-
Hamilton & Hinton 1910-21, 2 : 351) from the ‘Forest Bed’ can possibly be added
to the above list.

The rodent fauna with at least 11 species is unusually rich. It is predominantly
of forest and meadow type, testifying to a mild climate; no arctic elements are
present. The presence of Cvicetus in this fauna is difficult to explain.

Sciurus whiter is represented from the marine ‘Monkey Gravel’ at West Runton,
which overlies the Upper Freshwater Bed.

(d) Anglian. Newton (1882b) described some isolated worn teeth of Spermophilus,
found in association with remains of arctic plants, in the Arctic Freshwater Bed
near Mundesley (60). This deposit, which is immediately overlain by glacial till,
has been interpreted as an indication of oncoming cold conditions in early Anglian
times.

The rodent fauna of the Forest Bed deposits at Ostend (57) near Bacton presents
a stratigraphic problem, since it cannot be accurately related to the above sequence.
The remains include Avvicola bactonensis and A. green (Hinton 1926b : 386, 3890),
now both regarded as synonyms of A. cantiana, which led Hinton to conclude (p. 391)
that the Ostend deposit is later than the Upper Freshwater Bed of West Runton.
The specimens were collected over a century ago by the Rev. C. Green and un-
fortunately lack reliable stratigraphic information.

B. Terrace and Solifluxion deposits of the Rivers Thames and Lea

The richest series of rodent-bearing deposits related to a British river system is
that of the River Thames and River Lea. The deposits of these rivers have been
extensively worked commercially for gravel and brickearth and there have been
many deep excavations for building foundations in the London area, leading to
frequent discoveries of rodent remains (Fig. 3).

Three main series of deposits occur in association with these rivers, the exact
relationship between which is often difficult to determine. Firstly, there are the
terrace deposits of their upper courses, including Marlow and Isleworth in the Upper
Thames, and Water Hall Farm Pit, Nazeing, Ponders End, Edmonton and Hackney
on the River Lea. Deposits of both interglacial and glacial age occur here, those
representing glacial stages merging with buried channels in the Lower Thames
Valley province, where they are inaccessible for study in consequence of later rise
of sea level.

ORS BRS Hy TSS 49

sClacton
on Sea

Waiter Hall

Ponders End
Edmonton

Hackney @ !Iford
Marshes

Avele
@ ¥ River Thames

Erith @
Crayford @ Upnor

= ® s 4
Swanscombe Northfleet KENT peak

Fic. 3. Location map of rodent localities in the Thames estuary and the Lea valley.

Secondly, there are terraces of the Lower Thames, including Swanscombe,
Clacton, Ilford, Aveley and Grays, Crayford and Erith. Hinton (1926b : 126-131)
classified these as the High Terrace of the Thames (Swanscombe), the Early Middle
Terrace (Grays Thurrock), and the Late Middle Terrace (Crayford and Erith).
Whereas each terrace of the upper part of the river is of approximately constant
height above the present-day bed of the river and thus becomes progressively lower
as it is followed downstream, these terraces of the Lower Thames (which are related
to former high sea levels) have approximately horizontal aggradation surfaces
and they do not extend further upstream than the head of the contemporary tidal
limit of the river. All are of interglacial or interstadial age and they can be cor-
related with raised beaches of similar heights along the open coastline. The cold
stages which occurred between the accumulation of these terraces of the Lower
Thames are represented by deposits submerged in the buried channels of the present-
day river, previously mentioned.

Thirdly there are solifluxion and hillwash deposits, the heights of which are un-
related to the terrace system, with the deposits of which they are often interbedded.
Northfleet is the only rodent locality, probably of this category, known to the
writers.

It follows that, in an area as complicated as the valleys of the Rivers Thames and
Lea, the exact relationship between the various rodent-bearing deposits is difficult
to interpret. The literature on this subject, which goes back for nearly a century
and a half, is voluminous. The general sequence was discussed by King & Oakley
(1936) and in a number of important papers by Zeuner (e.g. 1954), who produced a
schematic section from the Ebbsfleet Valley to Swanscombe, Kent, showing his
interpretation of the relationship between the various deposits there. He inter-
preted the solifluxion and loess deposits with Levallois industries of Baker’s Hole

50 PLEISTOCENE RODENTS

and Ebbsfleet as earlier than deposits of both Main Monastirian and Late Monastirian
age, which he regarded as Last Interglacial, although only the Late Monastirian is
accepted as Last Interglacial by Evans (p. 40).

During recent years J. N. Carreck has made a detailed further study of many of
the mammalian localities in the Thames estuary, especially those in Kent, together
with related museum collections and previous literature. On a basis of all lines of
evidence he relates (7m litt.) the deposits of some of the more important localities,
here given in ascending stratigraphic order, as follows :

1. Interglacial deposits of Swanscombe, with Clactonian and Acheulean industries.

. Interglacial deposits of Grays Thurrock and Little Thurrock (including the
Orsett Road brickearths, but not those of West Thurrock, which are slightly
later), Ilford and Aveley. The Ilford mammoth is a transitional form between
Mammuthus trogontheri and M. primigemus. The Ilford fauna does not
include any cold elements.

3. Deposits of the Baker’s Hole cold stage at Ebbsfleet/Northfleet. The first
true mammoth, M. primigenius, intermediate in form between the mammoths
of Ilford and Crayford, appears at this locality in non-estuarine deposits over-
lying coombe rock (solifluxion) deposits. There are several associated Leval-
lois industries.

4. Interglacial deposits of Crayford and Erith. Carreck found that the mammoth
from the Lower Crayford Brickearth is more advanced than that from stage 3,
though not so advanced as Last Glaciation mammoths from such localities as
Ponders End, in the Lea valley, north London. He suggested that a layer of
shattered chalk at the base of the Crayford sequence may be equivalent to the
Baker’s Hole cold stage and that the Crayford Gravel and Lower Brickearth,
which are interglacial, follow immediately in time. The Crayford fauna
suggests more open grassland conditions than indicated at Ilford or Grays, and
there is a Levallois industry more advanced than that at Baker’s Hole.

5. Other interglacial deposits, with hippopotamus, suggest that there may have
been a further amelioration of climate. Carreck considers that the hippopo-
tamus may have survived to a very late stage during the Last Interglacial.
The Last Glaciation was characterized by a more advanced mammoth than
that from Crayford.

iS)

Carreck pointed out that most of the Grays mammals were collected over a century
ago, possibly from more than one terrace, which makes the study of this locality
difficult ; he drew attention to the hazards of confusing the Baker’s Hole cold stage,
which he regarded as occurring within the Ipswichian Interglacial, with the Wol-
stonian/Saale Glaciation.

Carreck’s chronological conclusions are of very far-reaching importance, with
two points of special interest. Hinton’s supposition that the Crayford deposits are
later than those at Grays receives further support, but unlike the Geological Society
(Mitchell e¢ al. 1973), which regards the Northfleet coombe rock as Wolstonian and
earlier than Ilford (which is regarded as Ipswichian), Carreck regards the Ilford
deposits as being earlier than those of Northfleet.

OF BRITISH TSEES 51

Let us now consider in more detail the various deposits in the valley of the Thames
and its tributaries where remains of fossil rodents have been found.

(i) LowER THAMES.

(a) High Terrace of Swanscombe. One of the most important British rodent
localities is the world-famous Swanscombe skull site (see Ovey (1964) for a fuller
account). A series of estuarine deposits is aggraded to a height of approximately
31°5 m (103 ft) O.D. and is believed to have accumulated when the sea level rose to
this height. Solifluxion deposits bring this figure to a total of 35:5m (1173 ft).
Rodent remains have been found in two gravel pits, separated by a distance of about
a third of a kilometre and now both disused. These are Barnfield Pit, where skull
fragments of Acheulean Man were found, and Dierden’s Pit, Ingress Vale. The
following sequence of deposits has been described from Barnfield Pit :

6. The Upper Gravel. This is regarded as a solifluxion deposit, later than the
terrace proper. No mammalian remains have been found in it.

5. [he Upper Loam, with an Acheulean industry but no mammalian remains.

4. The Upper Middle Gravel, occupying a channel in the underlying deposits and
containing an Acheulean industry. The top of this stratum, which is at about
31°5 m (103 ft) O.D., is regarded as the true surface of the terrace deposits
proper. This deposit contains the ‘Homo layer’ in which the human remains
were found.

3. The Lower Middle Gravel, with an Acheulean industry.

2. The Lower Loam, with an im situ Clactonian activity horizon at its base and a
knapping floor higher up. A zone of subaerial weathering, with many well-
preserved animal footprints, is present on the surface of this deposit.

1. The Lower Gravel, with an early Clactonian industry, resting on a bench of
underlying rock at 21-27 m (70-90 ft).

The deposits contain a rich molluscan fauna which has recently been re-examined
by Kerney (1971). He concluded that the Lower Gravel, Lower Loam and the very
base of the Lower Middle Gravel accumulated under temperate conditions, after
which the climate became cooler.

Rodent remains have been found in the Lower Gravel (Carreck 1959), in the Lower
Loam (remains recently excavated by Waechter), and in a silt bed in the Upper
Middle Gravel at a level slightly higher than the ‘Homo layer’ (Schreuder 1950). A
few rodent remains have also been found at Ingress Vale. The distribution of rodent
species at these two localities is shown in Table 3.

The exact relationship between the deposits of Barnfield Pit and Ingress Vale is
uncertain. Avvicola cantiana occurs at both the latter site and in the Lower Loam
of Barnfield Pit, suggesting a possible correlation of the two. Kerney (1971 and
personal communication) found at Ingress Vale a temperate molluscan fauna associ-
ated with an industry which is culturally more advanced than that in the Lower
Gravel and Lower Loam, suggesting to him a possible equivalence to the base of the
Lower Middle Gravel.

The Swanscombe rodent remains are insufficiently abundant to permit reconstruc-
tion of any environmental and climatic changes, but the species of voles found in the

52 PEEISTOCENE RODENTS

TABLE 3
Distribution of rodent species at Barnfield Pit and Ingress Vale, Swanscombe, Kent

Barnfield Pit

~"—-- SO —-’- -
Lower Upper Ingress

Lower. Dower" (Middle ) , Middle’ |!) aie
Gravel Loam
Gravel Gravel

Trogontherium boisvilletti x
Castor fiber x
Apodemus sylvaticus* x
Clethrionomys glareolus x
Ayvicola cantiana x x
Lemmus sp. x
Microtus avvalinus x x
M. vatticepoides x x
Microtus sp. (avvalis-agrestis group) x
Pitymys avvaloides x x

* Cited in earlier papers as A. white.

Lower Loam suggest that during the deposition of this layer a meadow environment
may have prevailed. There are no arctic and no forest species, but such negative
evidence is not decisive.

Although the Swanscombe deposits are generally regarded as being of Hoxnian
Interglacial age, there is increasing evidence that a substantial interval may have
elapsed between the deposition of the Lower Gravels/Lower Loam complex and of
the overlying Middle Gravels. In general the rodent fauna of the lower unit quite
closely resembles that of the Upper Freshwater Bed of West Runton. Puitymys
arvaloides, Microtus arvalinus and M. ratticepoides occur in both the Upper Fresh-
water Bed at West Runton and in the Lower Loam at Barnfield Pit. The two above-
mentioned species of Microtus persist into the Upper Middle Gravel. Mimomys
savini, characteristic of Cromerian assemblages sensu stricto, is nevertheless absent
from Swanscombe, where it is replaced by Avrvicola, absent from the Upper Fresh-
water Bed. It is of interest that by Upper Middle Gravel times Lemmus had
appeared and A. cantiana may have disappeared. At the present day Lemmus is a
rodent of northern latitudes. Whereas we have no means of demonstrating that its
habit is not a recent adaptation, the occurrence of this genus could be interpreted
as further evidence of the cool conditions indicated by the mollusca.

Recently Mullender (in Wymer 1974) has examined pollen from the Lower Loam
of Barnfield Pit and found a marked break in the profile with a great increase in pine
and near disappearance of alder. He equated the lower part of the Lower Loam
with the Hoxnian Late-temperate Zone II and the upper part with the Wolstonian.
If this interpretation were to be accepted then the later deposits of Barnfield Pit
could no longer be referred to the Hoxnian, which would lead us to fresh problems
of nomenclature for this part of the sequence.

(b) Clacton. The exact age relationship between the deposits of Clacton and
Swanscombe is not clear, though Clacton is generally regarded as a further Hoxnian

OF BRITISH ISLES 53

locality intermediate in age between the Lower Loam and Middle Gravel of Barnfield
Pit. The Clacton deposits, which le at about sea level, are at a lower altitude than
those at Swanscombe. West (1972) suggests that slight downwarping has occurred
in the Clacton area, which may explain this difference. There is a Clactonian in-
dustry. Plant remains from Clacton, studied by Pike & Godwin (1952), suggest
about one-third of an interglacial sequence, during which mixed deciduous forest of
the warmth maximum was replaced by coniferous forest.

Only a few rodent remains have been found in the Clacton deposits. These are
Trogontherium cf. cuviert (= T. borsvillett:) and Clethrionomys sp. (Singer et al. 1973),
Castor sp. and Mucrotus of the agrestis group (Hinton 1923b), and Arvicola cf.
praeceptor (= A. cantiana) (Warren 1955, quoting Hinton).

(c) Early Middle Terrace at Aveley, Grays Thurrock and Ilford. Hinton (1926b:
129) defined Grays Thurrock (33) as the type locality of this terrace of the Thames,
with which the present writers also include Ilford (35) and Aveley (34). Table 4
lists those rodent species whose remains have been found at these localities.

The age of the deposits at these localities has been the subject of much discussion
which is yet to be concluded. Hinton (1926b: 129-131) considered that the
species of Avvicola from Grays (A. praeceptor, = A. cantiana) was not closely related
to modern species of that genus. He considered that the latest Forest Bed deposits
(presumably Ostend), the High Terrace of the Thames and the Early Middle Terrace
were close in time. Zeuner (1945) considered Grays to be Hoxnian.

Hoxnian terrace deposits do indeed occur in the Grays area, though remains of
fossil mammals do not seem to have been found in them. Wymer (1957) records
terrace deposits of Hoxnian age with Clactonian implements resting on a bench at
15 m (49 ft) O.D., but points out that the Grays brickearth is later than this deposit.
Most of the Grays rodent remains were found in the brickearth in a small pit near
Orsett Road (Hinton 1gor), about 650 m (700 yds) west of Wymer’s site.

West (1969) studied plant remains from the three sites mentioned above and
concluded that Aveley and Ilford are of Ipswichian age, and that Grays is inter-
glacial, probably also Ipswichian. He pointed out that there was at this time
apparently an important aggradation phase which resulted in the spreading of
alluvium up to levels of between 12 and 15 m (40-50 ft) O.D.

The age of the Grays-Ilford—Aveley deposits will be further discussed below.
From the evidence of the Avvicola remains, Grays Thurrock would appear to be
relatively early.

TABLE 4

Distribution of rodent species at Grays Thurrock, Aveley and Ilford, Essex

Grays Thurrock Aveley Ilford
Castor fiber x x
A podemus sylvaticus x
Clethrionomys glareolus x x
Ayvicola cantiana x x x
Microtus agrestis x Microtus sp. x

54 PLEISTOCENE RODENTS

(d) Late Middle Terrace at Crayford and Enth. A rich rodent fauna has been
collected from the Thames terrace deposits of Crayford and nearby Erith; it has
been described by Kennard (1944). Owing to the lenticular character of the deposits
at these localities there is little constancy in the details of the succession. Three
broad divisions can, however, be identified. In descending order there are :

3. The Upper Brickearth, apparently not fluviatile but the result of sludging.

2. The Lower Brickearth and Corbicula Bed, up to 9m (30 ft) O.D., laid down
in a sluggish stream and in more strongly running water respectively. Leval-
lois artefacts have been found in this deposit. Most of the rodent remains
are from the Corbicula Bed.

1. The Lower Gravel, deposited by a fast-flowing river. The top was a land
surface occupied by Levallois man, who left many artefacts.

Most of the rodent remains have been found in the Corbicula Bed, which is a
sandy development of the Lower Brickearth. The following species are represented :
Spermophilus primigenius, Microtus oeconomus, M. nivalis, M. agrestis, Arvicola sp.,
Lemmus lemmus and Dicrostonyx torquatus.

The age of the Crayford deposits is a topic requiring extensive further study.
Hinton (1926b: 131) considered Crayford to be later than Grays, with a major
intervening change of fauna. This view is also supported by Carreck (in litt.). At
the present time Crayford is widely considered to be of Ipswichian age and both the
freshwater and land mollusca found in the Corbicula Bed suggest conditions warmer
than at the present time. The associated rodent fauna, however, is a typical
assemblage of the ‘penultimate’ Glaciation. Spermophilus primigenius is unknown
from the sediments of the Last Glaciation in Europe and, although Mzcrotus nivalis
was present during the Last Glaciation in many parts of Europe, and still survives in
the mountains of central and south Europe, there is no evidence that it survived
into the Last Glaciation in Britain.

It has been suggested that the lemmings burrowed into the Corbicula Bed from a
later land surface, or alternatively that the remains were derived from an earlier
deposit. Field evidence, however, suggests that the rodent remains are contem-
porary with the Corbicula Bed.

A total absence of forest species of land mollusca indicates that the country was
open grassland. Living species of Spermophilus are also predominantly grassland
animals. This could be interpreted as the beginning or end of an interglacial stage,
though it is pointed out by Turner (in prep.) that riverside trampling by large
herbivores can also give rise to clearances in otherwise wooded areas.

The Crayford deposits apparently date from some time during Wolstonian-—
Ipswichian times. Their precise stratigraphic position will be further discussed
below and in Section IV (pp. 125-126).

(e) Floodplain Terrace complex. A series of lower terraces of Last Interglacial
and Last Glaciation age in the Thames estuary have produced an abundance of
remains of large mammals but, with one exception (Castor from the Upper Flood-
plain Terrace of the River Medway, a river flowing into the Thames estuary, at
Upnor, Kent), no rodent remains are known from them.

OF BRITISH ISLES 55

(f) The Middle Terrace/Floodplain Terrace problem. Although no stratigraphically
significant rodent remains have been found in the Floodplain Terrace complex
of the Thames estuary, and although Hinton did not mention this terrace in his
monograph (his ‘Third Terrace of the Thames’ is actually a terrace of the upper part
of the River Lea), abundant rodent remains which are probably of Upper Floodplain
Terrace age have nevertheless been found in many British cave deposits. The
present section would be incomplete without some discussion of the status of this
terrace.

It has already been pointed out that various workers have presented evidence
that the period of time regarded by the Geological Society (Mitchell et al. 1973) as
Hoxnian—Wolstonian—Ipswichian was, in fact, more complicated than this sequence
suggests. Zeuner (1945) described evidence for a minor cool phase dividing what
he regarded as the Last (Ipswichian) Interglacial into an earlier and a later part.
He considered that there were two stages of high sea level during the Last Inter-
glacial, the Main Monastirian or 18m and subsequent Late Monastirian or 8m
shorelines, interrupted by an intra-Monastirian fall of sea level. These stages were
accepted by Evans (1971 ; see pp. 40-42), though he argued that the earlier of these
two stages should not be called Last Interglacial and observed that some confusion
seemed to have arisen in archaeological circles where ‘Last Interglacial’ had been
used for both.

Zeuner correlated the so-called Taplow Terrace of the Thames estuary (our
Middle Terrace complex belongs here ; the term “Taplow’ is best abandoned in the
estuary since Taplow is in the upper part of the Thames and the terrace there is not
estuarine) with the Main Monastirian sea level and the Upper Floodplain Terrace
with that of the Late Monastirian. Both these supposed terraces are highly fossili-
ferous, with apparently different mammalian megafaunas.

When Zeuner described his interpretation of the Last Interglacial terrace sequence
of the Thames estuary very little palaeobotanical evidence was available to him.
During recent years, however, fossil plant remains have been found at a number of
Last Interglacial localities there, the most important being Trafalgar Square
(Franks 1960), Seven Kings Station, Ilford (West e¢ al. 1964) and Aveley (West
1969), and here problems have arisen. The plant remains from these localities
were assigned to the pollen zones shown in Table 5.

To those who had been studying the terraces and mammalian faunas of the
Thames estuary, these results were surprising. Whereas Ilford and Aveley had
appeared to be part of the Middle Terrace of the Thames and Trafalgar Square part
of a lower terrace, the Upper Floodplain Terrace (i.e. Zeuner’s two terraces of the
Last Interglacial, representing two distinct periods of time), only one climatic
fluctuation, with the climatic optimum of Zone IIb represented at all three sites,
could be recognized from the pollen evidence. Was Zeuner wrong in separating the
two stages of his supposed Last Interglacial, the Middle and ‘Upper Floodplain’
Terraces of the Thames estuary really being only one terrace, or are the floral
remains from two separate climatic events so similar that these terraces cannot be
distinguished on palaeobotanical evidence? Mammalian and morphological evi-
dence suggests that, in agreement with the views of Zeuner and Evans, the latter

56 PLEISTOCENE RODENTS

TABLE 5

Pollen zonation of some ‘Ipswichian’ interglacial sites in the Thames estuary

Pollen Zones Flora Ilford Aveley | Trafalgar Square

h-i
&

Oak, hornbeam,
silver fir

lib
e Ila
Ib
la

eh

Mixed oak forest

Birch, pine
c
‘PENULTIMATE’
GLACIATION

N.B. The Trafalgar Square deposits are of Zone IIb age, and should be shown somewhat higher than
indicated.

Last INTERGLACIAL

d

alternative is the more likely even though palaeobotanical evidence supports only a
single terrace (Fig. 4).

Even the chronological interpretation of the Ilford sequence is not without
problems. West ef al. (1964) referred the gravel under the Ipswichian brickearth
and plant deposit at Seven Kings Station to the Gipping cold stage, but Carreck
(in litt.) has pointed out that there is considerable lateral variation of the Ilford
deposits with interglacial species of mollusca and mammals frequently present in the
gravels and sands.

Let us now consider the large mammals of these Thames sites in some detail. At
Trafalgar Square (supposedly Upper Floodplain Terrace) Zone IIb contains remains
of hippopotamus, straight-tusked elephant, a rhinoceros which is probably Dicero-
rhinus henuitoechus, fallow deer, red deer, giant ox, bison, lion and other animals.
No mammoth or horse remains were found. This is a faunal assemblage which
occurs commonly in British cave deposits with localities as widely spread as Joint
Mitnor Cave, Devon, and Kirkdale Cave, Yorkshire (Sutcliffe 1960).

The Ilford—Aveley fauna appears to be entirely different. At Ilford most of the
elephant remains are of an early form of mammoth with affinities to the Middle
Pleistocene Mammuthus trogonther, though straight-tusked elephant is also
represented. Two species of rhinoceros are present, D. hemistoechus and D. kirch-
bergensis. This last species is also common in the preceding Hoxnian Interglacial.
There is an abundance of horse; both hippopotamus and fallow deer are absent.

QUA IBIS el IS ILILS 57

At Aveley straight-tusked elephant in Zone IIb is replaced by mammoth in Zone III.
Comparison of rodent species unfortunately cannot be made, since no diagnostic
remains have been found in the Upper Floodplain Terrace. Avrvicola cantiana is
present in the Middle Terrace at Aveley, Ilford and Grays Thurrock.

The megafaunas of Aveley and Ilford on the one hand, and of Trafalgar Square
on the other, are so different that it is difficult to believe that they are contemporary,
though consideration must be given to the possibility of their representing different
stages within a single climatic fluctuation.

The evidence does not seem to support the latter alternative. The hippopotamus
level at Trafalgar Square is Zone IIb. A vast amount of Zone IIb clay was exca-
vated for cement-making at Aveley but no hippopotamus remains were found. The
Ilford plant remains cover Zones IA—IIb but there were no hippopotamus remains
there ; two specimens formerly attributed to this animal have since been shown to
have been incorrectly identified.

It has been pointed out, in support of there having been only one temperate stage,
that most of the Ilford mammal remains were recovered over a century ago, during
excavations for brickearth which were relatively shallow and did not extend into

THAMES ESTUARY IPSWICHIAN TERRACE SEQUENCE
SCHEME A (Morphological & Mammal Evidence)

Middle Terrace

14m. ss
(45°) O.D. Lo S Upper Floodplain Terrace
gee NDS ;
~, (Lower Floodplain Terrace)
Om. O.D.

SCHEME B (Palaeobotanical Evidence)

Ilford Mammals

a ee EC

22 Erosion
Sara Trafalgar Square Mammals

Fic. 4. Schematic section of Thames estuary Ipswichian terraces, showing alternative
interpretations of the stratigraphic succession. Reproduced from Sutcliffe & Bowen

(1973 : 18).

58 PLEISTOCENE RODENTS

underlying gravel deposits, which were of no commercial value. The organic
deposit described by West e¢ al. (1964) at Seven Kings Station underlay the brick-
earth, so it is possible that the ford mammalian fauna is later than Zone IIb
(perhaps Zone III), that is the Trafalgar Square fauna might have been found in the
basal part of the Ilford terrace had excavations been carried deeper. According to
this theory Trafalgar Square represents an earlier stage than Ilford ; according to
the twin terrace theory it is later. The former interpretation presents several
difficulties. In spite of extensive commercial excavation of sand and clay in the
London area, no such relationship has ever been observed ; the Upper Floodplain
Terrace does appear to be a good morphological feature ; the Ilford mammoth is
relatively primitive ; the Ilford fauna has never been recognized in any British cave,
with the possible exception of Hutton Cave, whereas the Trafalgar Square fauna
with hippopotamus occurs frequently, suggesting that such deposits have been
subject to less denudation and are more recent.

In the absence of diagnostic rodent species in the Upper Floodplain Terrace of the
Thames the above problem cannot at present be resolved from studies in the Thames
area. It will be further critically examined below, in the section on hippopotamus
faunas in caves and in Section IV.

(ii) UPPER THAMES.

(a) Isleworth. Remains of Microtus oeconomus and M. gregalis were found at
Willment’s Gravel Pit, Isleworth, in a deposit where they were associated with
remains of temperate mollusca and insects (Coope, 1975) and with plant remains
dated by 14C to 43 140+1520 or —1280 years B.P. (Birmingham 319). This deposit
was overlain by sands and gravels containing remains of reindeer and cut by many
fossil ice wedges. A description of the site, by J. Simons, is in preparation. The
sequence is interpreted as indicating deteriorating climatic conditions after an inter-
stadial during the Last Glaciation. It is probably the earliest occurrence of M.
gregalis in Britain ; this species appears to be restricted there to the Last Glaciation
and to have disappeared before the end of that stage.

(b) Marlow. A small sample of brickearth, found in the Treacher collection at
the British Museum (Natural History) and labelled as coming from a brickearth pit
1 mile north-east of Marlow, was found to be rich in rodent remains, predominantly
dissociated teeth. They were examined by Dr G. B. Corbet who found that most of
them were apparently of Microtus arvalis. There was also one first lower molar of
M.oeconomus. The lack of any second upper molars of M. agrestis and the occur-
rence of only one first lower molar of M. oeconomus among the many first lower
molars (which teeth distinguish these last two closely related species from J.
arvalis) suggest M. arvalis was the principal species present.

The age of the Marlow rodents is uncertain, but it is likely to be Last Glaciation.
Neither M. arvalis nor M. oeconomus survives on the mainland of Britain at the
present day.

(1) RIvER LEA. .

(a) Water Hall Farm Gravel Pit. An important sequence of Pleistocene deposits
occurs at Water Hall Farm, Hertfordshire. Here the valley of the River Lea has

OF BRITISH ISLES 59

cut through earlier glacial deposits which still cap the hills on either side. In the
valley bottom the River Lea is flanked on one bank by its present-day floodplain,
on the other by a slightly higher terrace in which have been found abundant remains
of interglacial mammals, including Hippopotamus and Palaeoloxodon antiquus.
Remains of Mammuthus have also been found. At the base of the sequence of
deposits in this terrace there was formerly exposed a white marl with remains of
amphibia and of the rodents Mzcrotus agrestis, M. oeconomus and M. nivalis. Since
most of the commercial excavation of the interglacial terrace, since concluded, was
carried out by mechanical excavators, the exact stratigraphic relationship between the
hippopotamus layer and the rodent layer remains unproved, but nearly the whole of the
ossiferous gravel deposit for which the pit was worked overlay the rodent-bearing
marl and there seems no doubt that the rodent remains antedate those of the hippo-
potamus. Unfortunately no rodent remains have been found in the upper deposits.
The sequence of events at this locality is provisionally determined as follows :

1. Boulder clay and glacial deposits. Not later than ‘penultimate’ glaciation.

2. Retreat of ice and valley deepening by River Lea.

3. Deposition of marl with remains of amphibia and rodents.

4. Deposition of interglacial deposits with Hippopotamus and P. antiquus over-
lying the rodent marl. Last Interglacial.

5. Further deepening of Lea Valley and formation of present floodplain. Last
Glaciation to present day.

(b) Nazeing, Ponders End, Edmonton and Hackney. A complex series of deposits,
mainly of late Last Glaciation and Holocene age, occupies the valley of the lower
part of the River Lea. Rodent remains have been found at a number of localities
there. The most important of these is at Nazeing, where a series of channel, lake
and marsh deposits have been palaeobotanically dated as ranging from Late Glacial
to post-Glacial vegetation Zone VII, providing the latest known survival dates in
Britain for a number of rodent species (Hinton 1952). The late Pleistocene part
of the sequence produced remains of Microtus oeconomus, M. gregalis, Arvicola
terrestris, Lemmus lemmus and Dicrostonyx torquatus. Of these the last-mentioned
species and M. gregalis were still present in Late Glacial pollen Zone III and M.
oeconomus appears to have survived possibly until post-Glacial Zone V. Arvicola
terrestris also persisted into the Flandrian and Apodemus sylvaticus, Clethrionomys
glareolus and M. agrestis appear in Zones V—VI, by which time post-Glacial re-
afforestation had reached an advanced stage.

Remains of Dicrostonyx torquatus have been found at three other late Pleistocene
localities in the Lea Valley : Ponders End (Warren 1916), Angel Road, Edmonton
(Hinton 1912) and at Hackney.

(iv) SOLIFLUXION AND MELT-WATER DEPOSITS.

(a) Northfleet, Ebbsfleet and Baker’s Hole. Burchell (1935) recorded finding rodent
remains associated with those of mammoth and rhinoceros in non-estuarine deposits
filling a channel cut through a coombe rock (solifluxion) deposit at Baker’s Hole,
between Northfleet and Swanscombe, Kent. As previously mentioned (p. 50)
Carreck (in litt.), from the study of the mammoth remains from this site, considered

60 PLEISTOCENE: RODENTS

that the deposits, which indicate an ameliorating climate following the deposition
of the coombe rock, are later than those of Ilford but earlier than Crayford. Stuart
(in ltt.), who has recently re-examined the Burchell specimens in the British Museum
(Natural History) together with a few specimens collected more recently by Kerney
and Sieveking, gives the following list of species: Clethrionomys glareolus, Microtus
arvvalis/agrestis and Arvicola cantiana. He found that in teeth of the last-mentioned
species the enamel is clearly differentiated, as in Mimomys, with no A. cantiana-
terrvestyis intermediate forms (see notes on Avvicola, pp. 99-102). This suggests a
relatively early date for the Northfleet rodent remains, a conclusion in agreement
with Carreck’s observations on the mammoth remains. Burchell (1935) lsted
Microtus arvalis (here interpreted as M. arvalis/agrestis group), M. nivalis and M.
anglicus (= M. gregalis) from his excavations, but the last two have not been
confirmed by Stuart’s re-examination of the available specimens.

C. Cave Deposits

Remains of Pleistocene rodents have been found in many caves in the British
Isles. Most such deposits accumulated in consequence of animals accidentally
falling down shafts or of their remains being carried into caves by birds of prey.
Since, however, the entrance parts of caves are vulnerable to destruction by processes
of denudation, many formerly-existing rodent deposits have since disappeared ; it is
unusual to find any rodent remains in caves of earlier than Upper Pleistocene age.
Kent’s Cavern and a recently discovered fissure at Westbury-sub-Mendip, Somerset,
are probably the only exceptions. A Lower Pleistocene cave deposit with megafauna
has also been found at Dove Holes, Derbyshire (Crag age), but unfortunately no
contemporary rodent remains have been recorded. In contrast, cave deposits with
Last Glaciation and Holocene rodent remains are very numerous.

Few of these sites have produced stratified sequences of rodent faunas, either
because the deposits accumulated during only a short period of time, because they
had been disturbed before excavation, or because they were excavated before the
importance of stratigraphy had become adequately appreciated.

The following are the most important rodent caves in the British Isles.

(i) WESTBURY FIssuRE (13). The recently discovered fissure infilling at Westbury-
sub-Mendip, described by Bishop (1974, 1975), is of special importance, since its
rich mammalian fauna indicates a stage slightly later than the type Cromerian of
West Runton, not otherwise well represented in the British Pleistocene. Bishop
described, in descending order, the following sequence of deposits :

III. ‘Rodent Earth’ (layer 10, the upper part of the ‘Calcareous Group’, layers
2-9 below). A deposit with an abundance of remains of rodents and other
small mammals, possibly an accumulation of pellets of birds of prey.

II. ‘Calcareous Group’ (layers 2-9, excluding the ‘Rodent Earth’). Pre-
dominantly limestone breccias with abundant remains of bears. The cave
was probably a bear den at this stage.

I. ‘Siliceous Group’ (layer 1). Water-laid deposits with some rolled bones and
teeth.

Oisy Iie Slel MSIL 61

TABLE 6

Stratigraphic distribution of rodent species in the Westbury Fissure
(after Bishop 1974, with personally communicated additions)

WH
o
18)
oO
4 = a q a
& a oO s 2 9S
is Es fe) = =
S 2) Oo ~» Se S
SS ed S| ae in
SS rs] “—
& S
S S wH ae; 7)
aS = a = <
S| Ss = S S aS S
Se al ier: Ef en ee Coe
- s ; = “< Ss SS) SS
cD) ~ jer = a, Ss S
oe 38 % SS D S S S Sb ©
2 SON Ae ae ae 25s ees A Ne
= s = Z S s Sy 8 Sy 3S
= = S S S > iS ‘'S Sees
o 8 Q Ss S =~ Ss = < cS
aa O) ze Q nN Ss) Q 55 AY =
(common)
III. ‘Rodent Earth’ fe) x x x x x x x x

II. ‘Calcareous Group’
(excluding 2-9 x x x x x
“Rodent Earth’)

I. ‘Siliceous Group’ I x

The distribution of rodent species in the Westbury fissure is shown in Table 6.

The age of the deposit has been discussed in detail by Bishop (1974). From the
‘Calcareous Group’, excluding the ‘Rodent Earth’, he recorded a fauna including
Homotherium latidens, Felis gombaszoegensis, Ursus deningert, Xenocyon lycaonotdes,
Canis lupus mosbachensis, Dicerorhinus etruscus and Equus mosbachensis. Rodents
from this level are Apodemus sylvaticus, Lemmus sp. (the earliest record of this genus
in the British Pleistocene), Avvicola cantiana, Pitymys gregaloides and Microtus
arvalinus. The same species are represented in greater abundance in the ‘Rodent
Earth’, which Bishop regarded as the last stage of the ‘Calcareous Group’. Remains
of Plhiomys episcopalis (the first record in the British Pleistocene), Dicrostonyx sp.
(the earliest record in the British Pleistocene) and Clethrionomys were also found in
the ‘Rodent Earth’.

Bishop regarded the Westbury fauna as later than the type Cromerian of West
Runton but not later than Elsterian. He drew attention to its similarity to that of
the classic ‘late Cromerian’ sites of Europe (in particular Mauer, Hundsheim, Tark6
and Mosbach) and equated the Westbury ‘Rodent Earth’ with the Avvicola fauna
group I of Koenigswald (1973) (see pp. 100-101). This stage was previously
unrecognized in the British Pleistocene sequence.

(i) KenT’s CAVERN. Unfortunately most of the rodent remains found in Kent’s
Cavern were excavated during the nineteenth century and only sparse stratigraphic
information is associated with them. The following species have been found there
(Hinton 1915, Kennard 1945-6, British Museum (Natural History) collection) :

62 PLEISTOCENE RODENTS

Castor fiber, Apodemus flavicollis, Clethrionomys glareolus, Arvicola terrestris, Dicro-
stonyx torquatus, Lemmus lemmus, Pitymys gregaloides, Muicrotus agrestis, M.
oeconomus and M. gregalis. In addition, Campbell & Sampson (1971) recorded a
specimen of Arvicola greent (regarded here as a synonym of A. cantiana) among
specimens collected from the cave by J. MacEnery between 1825 and 1829. This is
a typically Last Glaciation—Holocene fauna, with the exception of P. gregalotdes,
not recorded elsewhere in Britain in deposits later than those of the Westbury
Fissure, and A. cantiana. The occurrence of P. gregaloides and also sabre-toothed
cat, Homotherium, in Kent’s Cavern has been interpreted by both Hinton (1926b)
and Campbell & Sampson (1971) as evidence of a Cromerian stratum somewhere in
Kent’s Cavern.

The sequence of deposits in the cave varies from one chamber to another (for
details see Campbell & Sampson). The principal deposits, in descending order, are
as follows :

5. Black Mould, with Mesolithic and later artefacts and fauna.

4. Granular Stalagmite, with Mesolithic and Neolithic artefacts and fauna.

3. Cave Earth, including a local area of hearths known as the Black Band. Middle
to Upper Palaeolithic industries. Mammalian fauna including woolly mam-
moth, woolly rhinoceros and hyaena.

2. Crystalline Stalagmite.

1. Breccia, with many bear remains and with a Lower Palaeolithic industry.

The upper part of the above sequence is typically Last Glaciation (the Cave
Earth) to Holocene. Most of the rodent remains probably come from the upper
levels. Pitymys gregaloides, on the other hand, is a Middle Pleistocene species
probably derived from the Breccia, the age of which is at present uncertain. Homo-
therium, P. gregaloides and A. cantiana are all recorded from the Westbury Fissure.
At the present time it seems most likely that the earliest remains from Kent’s
Cavern are of Westbury Fissure age. There is no conclusive evidence of fauna as
early as Cromerian sensu stricto. It may be inferred that the Crystalline Stalagmite
represents a major break in sedimentation, possibly because the cave entrance had
become sealed.

(iii) TORNEWTON CAVE (3). The most important sequence of stratified cave
deposits with rodent remains known from the British Isles is that of Tornewton Cave,
south Devon (Sutcliffe & Zeuner 1962, Kowalski 1967). Deposits in the shaft-like
Main Chamber of this cave and of the talus deposits outside span a cold—interglacial—
cold sequence which is unique in showing differences between the rodent faunas of
the two cold stages concerned. The relationship between this sequence and that of
the terraces of the River Thames will be considered in Section IV, pp. 124-127.

The principal deposits in the Main Chamber of Tornewton Cave (excluding some
superficial deposits excavated during the nineteenth century and a series of stalag-
mite floors), in descending order, were found to be as follows :

5. ‘Diluvium’. Most of this deposit was excavated during the nineteenth century.
The remaining part contained a few remains of wolf, hyaena, bear, bovid and

OF BRITISH. TSLES 63

reindeer. A Holocene molluscan fauna has been identified by M. P. Kerney
from this horizon and it is probable that the mammalian remains are a mixture
of derived Pleistocene and contemporary Holocene species.

The Reindeer Stratum. Most of this deposit was excavated from the Main
Chamber of the cave during the nineteenth century. Abundant rodent
remains, associated with remains of wolf, hyaena, horse, rhinoceros, reindeer,
a bovid and a sparse Upper Palaeolithic industry were nevertheless recovered
more recently from an extension of this stratum in the talus outside the cave.
The stratum is of Last Glaciation age.

. The Hyaena Stratum. A deposit with abundant remains of spotted hyaenas,

which animals apparently occupied the cave as a lair, associated with occasional
remains of hippopotamus, narrow-nosed rhinoceros, red and fallow deer, lion,
wolf and bear. This is an interglacial assemblage, tentatively referred to the
‘Ipswichian’.

. The Bear Stratum. A deposit with abundant remains of brown bears, which

animals used the cave as a lair at this stage.

. The Glutton Stratum. The earliest fossiliferous deposit in the cave, with

numerous remains of brown bears associated with occasional remains of glutton °
(wolverine) and reindeer. This deposit represents an earlier cold phase.

Table 7 shows the stratigraphic distribution of rodent remains in the Main
Chamber and (Reindeer Stratum only) in the talus of Tornewton Cave. It will be

seen

from the table that, in the Glutton Stratum, Muicrotus oeconomus and M.

miwalis make up the bulk of the rodent fauna. Lagurus lagurus (from the only
British locality with this genus), Allocricetus bursae and Cricetus cricetus are typical
species of the ‘penultimate’ glaciation. Diucrostonyx torquatus and Lemmus lemmus

TABLE 7

Number of rodent specimens from the main deposits of Tornewton Cave

Glutton Bear Hyaena Reindeer

Stratum Stratum Stratum Stratum Diluviumy’ Total

A podemus sylvaticus I - I - 8 10
Cricetus cricetus 4 — — = - 4
cf. Allocricetus bursae 4 _ - ~ - 4
Dicrostonyx torquatus 7 2 _ F I 7
Lemmus lemmus I I - = = 2
Clethrionomys glareolus II 3 - 12 4 30
Ayvicola sp. 36 I 2 14 2 55
Microtus agrestis 170 4 6 13 Fig | 380
Microtus nivalis 298 4 - - = 302
Microtus oeconomus 1066 15 — 19 9 IIO0g
Microtus gregalis — _ - 43 12 55
Lagurus lagurus 62 I - — - 63
Total 1660 31 9 218 LE3 2031

The count is based on the number of determinable specimens, not individuals. The apparent greater
abundance of remains in the Glutton Stratum is the result of more extensive sampling of this deposit.
Remains from minor deposits not included above bring the total number of specimens determined to

2383.
6

64 PLEISTOCENE RODENTS

are present but not abundant. Mucrotus gregalis is absent. Sylvan species are very
rare, though a few remains of Clethrionomys glareolus occur with this predominantly
meadow and steppe assemblage. A single tooth of Afodemus sylvaticus probably
represents contamination of the deposit. The rodent fauna of the Glutton Stratum
has affinities to that of Crayford, Kent, from which it was probably not greatly
separated in time.

Only a few rodent remains were collected from the Bear Stratum. The fauna
does not differ in composition from that of the underlying Glutton Stratum though,
with the exception of one tooth of Lagurus lagurus and four of Microtus nivalis,
the typical ‘penultimate glaciation’ elements are lacking. Their absence could be
accidental, however, since they are also rare in the Glutton Stratum. The absence
of lemmings could indicate a slight amelioration of climate.

Rodents are poorly represented in the Hyaena Stratum. Apodemus sylvaticus,
represented by only one specimen, is a forest species. Mucrotus agrestis and Arvicola
sp. are ecologically neutral.

The Reindeer Stratum has a predominantly cold fauna with the typically Last
Glaciation Microtus gregalis and with Dicrostonyx torquatus.

The apparent mixture of arctic and sylvan species in the “Diluvium’ is in agree-
ment with this deposit being of Holocene age, with derived Pleistocene remains.
Apodemus sylvaticus reappears here. The climate was milder than before.

The sequence of rodent faunas described above is of great importance as 1t makes
possible, for the first time, distinction between mammalian assemblages, indis-
tinguishable on the basis of the larger mammals, of the two cold stages concerned.
Arctic elements (for example, lemmings) appear twice in the sequence, but are lack-
ing from the interglacial Hyaena Stratum. The lower cold stage is characterized
by, among other rodents, Cricetus cricetus, cf. Allocricetus bursae, Lagurus lagurus
and Microtus nivalis, all absent from the Reindeer Stratum, where they are replaced
by large quantities of M. gregalis which, on the other hand, is absent from the
Glutton Stratum.

Distinction is also possible between the teeth of Avvicola from the Glutton Stratum
and those from the Reindeer Stratum. W. von Koenigswald (1973 and personal
communication) has found that although the form from the latter deposit is A.
terrestris, that from the Glutton Stratum is an intermediate form tending towards
Arvicola cantiana, known from the Middle Pleistocene deposits of Swanscombe (see
pp. 100-102).

In addition to the deposits of the Main Chamber of Tornewton Cave two important
rodent deposits were found in positions where they could not be directly related to
the main sequence. These are the Otter Stratum, in a small chamber adjoining the
Main Chamber, and the Upper Rodent Stratum, on the rock platform outside the
cave mouth.

The Otter Stratum was found to contain an abundance of remains of brown bears
associated with some mammalian species unrecorded elsewhere in the Pleistocene of
the British Isles, notably Cyvnaonyx, the clawless otter, and Crocidura, the white-
toothed shrew (Rzebik 1968). It was composed of a mixture of broken stalagmite

OF BRiIitism Isles 65

blocks, some containing faunal remains, in an earthy matrix, suggesting some dis-
turbance of the deposit. The following rodent specimens were collected from it.

Cricetus cricetus 2 specimens
Lagurus lagurus I specimen

Dicrostonyx torquatus I specimen

Lemmus lemmus I specimen

Mucrotus nivalis 4 specimens
Microtus oeconomus 82 specimens
A podemus sylvaticus 2 specimens
Clethrionomys glareolus 19 specimens
Microtus agrestis 5 specimens
Arvicola sp. 8 specimens

This fauna appears to be a mixture from two originally separate layers. Remains
from a warm period, including Microtus oeconomus, predominate. They point toa
last interglacial age. No Hoxnian elements are present. Lagurus lagurus, Cricetus
cricetus and Mucrotus nivalis, on the other hand, are species of the ‘penultimate’
glaciation. In an attempt to separate the two faunal assemblages some of the
stalagmite blocks were dissolved in acid. Only temperate species (Muicrotus
agrestis, 1 determinable specimen ; M.oeconomus, 6; Arvicola sp., 1 ; Clethrionomys
glareolus, 4; Apodemus sylvaticus, 1) were found in the stalagmite. The exact
stratigraphic relationship between the Otter Stratum and the deposits of the Main
Chamber of Tornewton Cave is unfortunately uncertain, since the overlying deposits
were excavated during the nineteenth century and are no longer available for
examination. The cave was apparently still inhabited as a lair by bears, making it
probable that the deposit antedates the Hyaena Stratum. On the other hand, it
seems to be later than the Glutton Stratum since Koenigswald has found that,
whilst the Arvicola from both the Glutton Stratum and the Otter Stratum is an
intermediate form between A. cantiana and A. terrestris, that from the Otter Stratum
is slightly more advanced, though not so advanced as the form from the Reindeer
Stratum, which is A. terrestris. The Otter Stratum fauna appears to be a unique
assemblage, unknown from any other locality in the British Isles.

The deposit known as the Upper Rodent Stratum occupied a rift on the rock
platform outside the main entrance to Tornewton Cave. It was only a short distance
below the surface and is probably of modern date. The following species are
represented: Apodemus sylvaticus, 58 specimens; Clethrionomys glareolus, 14;
Microtus agrestis, 86; Arvicola terrestris, 2; Muicromys minutus, 1. All these
species live near the cave at the present day, no extinct species are represented ; and
M. minutus is a late arrival probably associated with neighbouring plough-land.

In conclusion the relative proportions of the various ecologic groups of rodents
represented in Tornewton Cave are shown in Table 8.

(iv) OTHER CAVES WITH PRE-‘IPSWICHIAN’ DEPOSITS. In light of the faunal
sequence demonstrated at Tornewton Cave, it would be surprising if other rodent
faunas dating from the time interval between the Hoxnian and the warm stage
represented by the Hyaena Stratum of Tornewton Cave were not to be found in
caves from time to time. At present this period is one of some obscurity.

66 PLEISTOCENE RODENTS

TABLE 8

The relative proportions of the various ecologic groups of rodents found in Tornewton Cave

Layer ae Boreal Forest Neutral
and steppe
Upper Rodent Stratum fo) o) 44°7 55°53
‘Diluvium’ II'5 8-0 10-6 69°9
Reindeer Stratum 22°9 o7 575 62°8
Hyaena Stratum fo) fe) 2:3 97°7
Bear Stratum 25°8 48°4 9°6 16-1
Glutton Stratum 22-77 64:2 O-7 12°4

Tundra and steppe elements include Dicrostonyx torquatus, Lemmus lemmus, Microtus gregalis, M.
nivalis, Lagurus lagurus, Cricetus cricetus and Allocricetus. Boreal elements include M. oeconomus. At
the present day this species lives mainly in the taiga belt but it also occurs, under milder climatic con-
ditions, in marshland. The forest element is represented by Apodemus sylvaticus and Clethrionomys
glaveolus. Arvicola sp. and M. agrestis are ‘neutral’ in their ecologic requirements. Micromys has been
omitted from the table since it was probably introduced by man.

Some isolated pre-‘Ipswichian’ rodent-bearing cave deposits have nevertheless
apparently been found. Hinton (1926b) regarded the rodents of Clevedon Cave (14)
and Banwell Cave (10), Somerset, as probably contemporary with those of Crayford
which, we have observed, are similar to those of the Glutton Stratum of Tornewton
Cave. The rich material from Clevedon Cave is of special interest, as it contains
only two species of rodents, Microtus nivalis and M. oeconomus. Hinton (19072)
wrote: ‘Dr H. C. Male very generously presented me with a small series of the
numerous jaws of Microtus which he found in the Clevedon deposit.’ In the material
described by Hinton, which is still preserved in the British Museum (Natural
History), there are exclusively remains of the species mentioned above. It is pos-
sible that Dr Male did not present Hinton with the material of the other rodents,
but it does not seem possible that he retained jaws of Mzicrotus agrestis, as they are
difficult to distinguish from those of M. nivalis except by a specialist. It seems
probable that the Clevedon Cave remains accumulated at a time when the rodent
fauna was composed exclusively of M. oeconomus and M. nivalis.

Hinton’s specimens from Banwell Cave have not been seen by the present writers.
The only available specimens, representing Avrvicola sp. and Mucrotus sp., are
stratigraphically inconclusive.

The Gough’s Cave (12) rodent fauna is composed of Arvicola sp., Microtus agrestis,
M.oeconomus and M. nivalis. The last-mentioned species suggests an age equivalent
to the Glutton Stratum at Tornewton Cave, though the occurrence of a late Upper
Palaeolithic industry in the cave indicates a more recent date for most of the
deposits.

The occurrence of M. nivalis among various rodent remains excavated by J.
Simons from Cow Cave, Chudleigh, Devon (7), suggests that there may be a pre-
‘Ipswichian’ deposit there.

Hutton Cave, Somerset (10), may provide a further example of a pre-‘Ipswichian’
rodent locality. One of the rodents found there, cf. Allocricetus bursae, is known
elsewhere in the British Isles only from the Glutton Stratum of Tornewton Cave.

OF. BRITISH ISLES 67

A primitive form of mammoth further supports a relatively early date for the fauna
of this cave.

(v) ‘IPSWICHIAN’ CAVE DEPOSITS. Sites, in addition to Tornewton Cave, include
the hippopotamus-bearing deposits of Joint Mitnor Cave, Devon (1), with Avrvicola
cantiana-terrestris transition form (determined by W. von Koenigswald) and Microtus
agrestis ; Minchin Hole, Glamorganshire (16), with the same species; the nearby
Bacon Hole (16) recently excavated by C. Stringer (1975) with Arvicola sp.,
M. agrestis, Clethrionomys glareolus and M. oeconomus; and Alveston Fissure,
Gloucestershire (15), with M. agrestis and Clethrionomys sp. Rodent remains from
the interglacial site of Kirkdale Cave, Yorkshire (77), include Avvicola sp., M.
agrestis, Clethrionomys sp., Apodemus sylvaticus and Dicrostonyx torquatus. This
cave was excavated over a century ago. Ducrostonyx is out of place and suggests
some mixing of material.

The rodent fauna of this stage is typically sparse, with Arvicola cantiana-terrestris
transition form and Microtus agrestis the most common species.

(vi) CAVES WITH POST-IPSWICHIAN DEPOSITS. Cave deposits dating from the
Last Glaciation and Holocene are far more numerous than earlier deposits, since
there has been less time for them to be destroyed by denudation. Many examples
are known. The distribution of rodent species at some of the more important cave
sites believed to date from this period is shown in Table 10 (pp. 70-71 ; for refer-
ences, see Section III).

The interpretation of these rodent faunas must unfortunately be undertaken with
caution since many of the deposits in which they were found had been disturbed by
burrowing animals; some important rodent caves were excavated during the
nineteenth century, so that adequate stratigraphic information is lacking. A fairly
consistent Pleistocene faunal assemblage is nevertheless apparent, with Sczurus
vulgaris, Muscardinus avellanarius, Apodemus flavicollis and Rattus rattus probably
not arriving until Holocene times. A series of excavations recently conducted by
the Peakland Archaeological Society in Derbyshire and Staffordshire (Bramwell
1960, 1964, 1970; Pernetta 1966) is of special interest for the light thrown on the
rodent faunas of the end of the Pleistocene and early Holocene. The two lemmings,
Lemmus and Dicrostonyx, appear to have been common until the end of the Pleisto-
cene after some of the large Pleistocene mammals, such as woolly rhinoceros and
hyaena, had disappeared ; Miucrotus oeconomus seems to have survived into post-
Pleistocene times. The apparent absence of lemmings at Levaton Cave (2, a Pleisto-
cene site with mammoth, woolly rhinoceros, red deer, reindeer and hyaena) suggests
that, during part of the Last Glaciation (probably an interstadial), Lemmus and
Dicrostonyx may have been absent from the fauna, at least in the south of England.
The Ightham fissures, Kent (Abbot 1917; Newton 1894, 1899a, b), are further
important cave sites with rodent remains of Last Glaciation age. All the fissures
do not appear to be contemporary, however, and there may have been some mixing
of Holocene material. ;

(vii) CAVES IN SCOTLAND. A series of small caves near Inchnadamph in Scotland
(83) are of special interest as including the most northerly British fossil rodent

68 PLEISTOCENE RODENTS

locality. From one of them, Creag nan Uamh Cave, Newton (in Peach & Horne
1917) recorded Dicrostonyx torquatus, Microtus agrestis and M. ratticeps (= M.
oeconomus). Remains of brown bear, reindeer and lynx were also found at the same
site. The age of these remains is uncertain, but the caves are situated in wild
mountainous country and the possibility that they are of relatively late age needs
to be taken into consideration. Some bones of brown bear found in a nearby cave
have recently been dated by 1!4C as only about 2700 years old (Burleigh 1972).

(viii) CAVES IN IRELAND. Most rodent species living in Ireland today are probably
post-Pleistocene arrivals. Only four possibly Pleistocene species need to be con-
sidered here. Their distribution at cave sites is shown in Table 9.

The exact stratigraphic position of these various remains is uncertain, although it
seems unlikely that any of the Irish mammalian faunas are earlier than Last Glaci-
ation. Mitchell (1969) considered that the mammals could not all be attributed to
one phase, suggesting that, although the remains of giant deer and reindeer found
beneath the peat bogs were known to be of Late Glacial age, the Castlepook Cave
fauna dated from an (earlier) interstadial of the Last Glaciation. A recently obtained
14C date for part of a mammoth bone from Castlepook Cave (33 500 + 1200 years B.P.,
University of Dublin 122) confirms Mitchell’s supposition.

Both Lemmus and Dicrostonyx are northern species which might be expected to
have reached Ireland in advance of other rodent species. Associated species in

TABLE 9

Distribution of rodent species at Irish cave sites

Keshcorran

Edenvale caves,
Co. Clare (85) Caves, Co.
: ee Sligo (84)

— SF

Ballynamintra Cave, Co. Waterford (90)
Kilgreany Cave, Co. Waterford (89)
Castlepook Cave, Co. Cork (87)
Castletownroche Cave, Co. Cork (88)
Red Cellar Cave, Co. Limerick (86)
Alice and Gwendoline Caves

a) 8 2
2 > >
zy eS, aS
a" iS a
fs = 4 O 35
S a oO »
om eats 2 ee
s z : & =
aes) 2
iss) oO ise} (2) —_
© Z ai O a
A podemus
ee 3 x x x x x x x x x
sylvaticus
Dicrostonyx
y x x x x x x x x
torquatus
Lemmus lemmus x x“

Microtus agrestis ?

OF BERIEISH ISLES 69

Castletownroche Cave include mammoth, both mammoth and spotted hyaena being
present in Castlepook Cave.

Although Afodemus sylvaticus has been recorded from several cave sites, Barrett-
Hamilton & Hinton (1910~—21) observed that in some instances it was most abundant
in the upper layers, implying a possible late arrival in Ireland. Scharff et al. (1918),
on the other hand, considered that at Castlepook Cave Apodemus was part of the
Pleistocene fauna.

An imperfect skull of Mzcrotus agrestis from Kilgreany Cave presents a problem,
since this is the only possible fossil record of Microtus from Ireland, where this genus
does not occur even at the present day. Savage (1966) observed that the skull is
notably fresh and unaltered and drew attention to the occasional occurrence of vole
skulls in north-east Ireland in owl pellets dropped by passing Scottish owls. There
are no other records of Microtus in Ireland and the chance introduction of this
specimen by owls or human activity needs to be taken into consideration.

D. Other Localities

(1) CROMERIAN LOCALITIES. Stuart (1974) recorded Mimomys cf. savini associated
with Dicerorhinus cf. etruscus from a site at Sugworth, near Oxford (21). He con-
sidered that the site is unlikely to be later than Cromerian.

(ii) HOXNIAN LOCALITIES. Carreck (1959) recorded Pitymys sp. from an inter-
glacial tufa near Hitchin, Hertfordshire (39). This locality is unlikely to be later
than Hoxnian in age. Other Hoxnian localities are Copford, Essex (42, Brown
1852; Turner 1970), with Trogonthertum boisvilletti, and the Hoxnian type site of
Hoxne with Tvogontherium boisvilletti (Spencer 1956), Apodemus sp., Lemmus sp.,
Arvicola sp. and Microtus sp. (personal communication from Dr R. G. Wolff).

(1) ‘IPSWICHIAN’ LOCALITIES. Six important East Anglian mammalian localities,
usually regarded as of Ipswichian age (including the type Ipswichian locality,
Bobbitshole) have produced rodent remains as shown in Table Ir.

It has been claimed by Koenigswald (pers. comm.) that although the Avvicola
from Barrington and from the Stuart collection from Swanton Morley (two hippo-
potamus localities) is A. cantiana-terresiris transition form, that from the Stuart
collection from Stutton (a locality without hippopotamus) is A. cantiana, suggesting
affinities with the Grays—Ilford—Aveley fauna of the Thames estuary. Stuart
(1974), on the other hand, considers that the Barrington remains include both
‘cantiana’ and ‘terrestris’ and not merely intermediate forms, suggesting that the
above interpretation is oversimplified. Stuart (im litt.) records A. cantiana from
Harkstead, a locality near to and probably of the same age as Stutton.

Other ‘Ipswichian’ localities with rodent remains are Hessle (74) with Arvicola
and Selsey (24) with Castor.

(iv) LAST GLACIATION LOCALITIES. The Last Glaciation brings us within the range
of radiocarbon dating. Four sites, two of them thus dated, will be mentioned here.
At Upton Warren, Worcestershire (20), remains of Dicrostonyx (determined by
Carreck) were found in association with mammoth, woolly rhinoceros and reindeer

7O

Distribution of rodent species at B

Somerset

Devon

d[OH] S,oulpoAy

UIOACD MOIIOG MOY

a[OP] S,uoxoIg

IoqyOYsyooy, poo Avy

eae) TrydQ,

oAey) UOpRog

aIMssKy YstorpnyD

ueAIqIoy, ‘aAeD) U0VeADT

Aenbioy, ‘oaeg Aemeddeyzy

dAeD) WPYXIG

yuoy ‘SoINssLy wWieyYyST

Sciurus vulgaris*

Spermophilus superciliosus

Muscardinus avellanarius*

Castor fiber*

A podemus sylvaticus

A. flavicollis*

Rattus vattus*

Clethrionomys glareolus

Avrvicola terrestris

Dicrostonyx torquatus

Lemmus lemmus

x

Microtus agvestis|arvalis group

M. oeconomus

M. gregalis

syuopor [eroe[s-ysod jo o1n}x1uIpe
Y}IM s}USpor pue euNeyesoul 9u900}sT9[q

eUNPJOIONW 9U990}SI90[q
euneyesoul sus00jsI0[q 1oddq
e}eIYS posnjuod ‘ausd0[OFT
VUNPJOIONW 9U990}SI9[q
PUNVJOIOU 9U9904S19[q
SUNVJOIOUWI IU9I04SI9[G
euneyesoul prep sieldl pozerloossy
9us00}jste[q-3sod Ayjsour ATqeqorg

CUNLJLSOW DUIIO}SIO[ Po}erIoossy

sjuoulsyea weayAs [eror[s-ysod Jo o1nzxruIpe
Y}IM s}UNpoOI pue euUNeyesoUI 9U990}sSI9[q

post-Pleistocene.

* Probably

Wee

+ Glaciation and Holocene cave sites

Derbyshire

ailysysiqued ‘oAeg xuAT

dITYS}UIP, ‘oaeg IossAueem4y

oITYSYIO XK “€ ‘ON VALD IpPISMmOD

oITyseoueyT ‘oAed SofoFT 80q x

dITYSPIOFeAS ‘oaey Ysng IJopyy

DARD IOPFT UG x) xX

DARD IJOH{ XO

—

aAey) ented
L

adaey [OMOC

>
x

dAeD UYIMSUCT x xX

dAeD YSsnoi1oqsiefy

DAC) OAM §=9ARD S,UTPIOP

SKC "Go Nga CUTSE

SUIVUIOI YSI}PIIG-Oueulory
pue o1yqzljoor,eg 93e] sTqeqoid poyeroossy

SUNVJLSOU 9UII0O}SIO]Y poyeIoossy

PUNPJOIOIW JUI90}STO[

PUNPJOIONU JUII0}STOT

eunryesoul ouss0}siofq 1oddq

sjuepor [eroe[s-ysod jo o1n}xruIpe

YJIM s}uNpoI pue euNneyesour 9u990}ST9[q

[OAD] (98% Ioyeoq) ousD0][0F{

S[9AIT 9UD90}SIO[G 932]

S[OAV] DUIIOTOFT

S[OAOT 9U990}STOTq 9} e'T

S[OAO] OUDDOTOF{

S[OA9T 9U990}STO[ J 9}e'T

SUNVLJESOU 9UB0}SIO[q po}eIOOssy

9U990}STO[q-}SOq

AjUO PUNFOIOTIAL

SUIeUIOI UPUIOYT

oe a (TS =

72 PLEISTOCENE RODENTS

TABLE IL

Rodent species from East Anglhan ‘Ipswichian’ sites

Suffolk ©
4 nt
\ ee
ee ys
i)
S + 12, Z
So eS i
ah q =
5) ra ©
n 0) ite)
ch ee es
ida) <
=~ (aa) ay = Se
> aa =
ia eee S| @ a Ss
+ U q fe) 5
(=| ~ wa g op fe)
2 ee wat, oe MEE Be
GS) Gere ea” 2e* aaa)
nn an Nn mM ea DN
A podemus sylvaticus x x x
Clethrionomys cf. glareolus x
Ayrvicola cantiana x x
Avrvicola cantiana-terrestyis transition form x x x
Avvicola sp. x
Microtus agrestis x x x x
IM. oeconomus x

List based on specimens in the Ipswich Museum (Stutton, Stoke Tunnel Beds, Harkstead and Bobbits-
hole), British Museum (Natural History) (Barrington), and Stuart Collection from Swanton Morley
(Stuart 1974).

in terrace deposits of the River Salwarpe (Coope et al. 1961). Associated insect
remains indicate cold continental conditions. There are !4C dates of 41 500+ 1200
years (GRO 595) and 41 900 + 800 years (GRO 1245).

At Beckford, Worcestershire (19), remains which are probably Muicrotus arvalis
have been determined by Dr G. B. Corbet from specimens separated by Coope
from material collected by D. J. Briggs from deposits of the Carrant Brook, in a
terrace equivalent to number 2 terrace of the Warwickshire Avon or the main
terrace of the River Severn. Associated mammals include mammoth, woolly
rhinoceros and horse. There is a cold insect fauna and “C dates of 27 650 + 250
years (Birmingham 293) and 27 300 + 500 (Birmingham 595).

At Fisherton, Wiltshire (22), Dicrostonyx torquatus, Microtus oeconomus and
Spermophilus superciliosus were associated with remains of large Pleistocene
mammals. In spite of earlier suggestions Microtus nivalis is absent. This is a
typical Last Glaciation faunal assemblage.

At Brean Down, Somerset (g), remains of Dicrostonyx were found in association
with arctic fox and reindeer in talus deposits regarded by Apsimon e¢ al. (1961) as
being of Late Glacial age (early Zone II).

(v) HOLOCENE LOCALITIES. Two rodent species, now extinct in the British Isles,
survived into post-Pleistocene times. These are Castor fiber and Microtus oeconomus.
Localities for Castor fiber include the Cambridge Fens (41), Staple Howe, Yorkshire

OF BRILISH TSS 73

(75), Star Carr, Yorkshire (76), and Thatcham, Berkshire (23). Mucrotus oeconomus
has been found in the Huntspill Cut (8) in the Somerset levels and on Nornour,
Isles of Scilly (91).

(vi) SCOTTISH LOCALITIES. Other Scottish rodent finds include Dicrostonyx from
Corstorphine, Edinburgh (81), and Castor fiber from the Loch of Marlee, Perthshire
(82), and Middlestots Bog, Berwickshire (80). The various ages of these remains
have not been determined.

i Clas > Ee iON SANDS DISTRIBU LLON OF RODENTS IN DE
IPILJDILS WOK ISAN ITs (Oey ANS IRIN US al IS IOINS

In this section rodents represented in the British Pleistocene are discussed in
systematic order. For completeness of taxonomic reference authors and dates of
both families and genera are given, though these are generally omitted from the
references. In the few cases where the currently used name of an extinct species
is based on non-British material the original reference, marked **, is given in the
synonymy. With these exceptions, only names used for fossil specimens from the
British Isles are cited in the synonymy and each name is given only once, with
reference to the first paper where it appeared. The type localities of extinct species
of rodents are indicated * in the locality lists or, if non-British, in the sections on
general distribution.

Locality lists for each species are divided into (a) English and Welsh, (b) Scottish
and (c) Irish sites; they are given in ascending stratigraphic order, as precisely as
possible on the basis of current information. Where several records are believed
to be of similar age they are grouped together, in alphabetical order, at the appropri-
ate position. Numbers refer to site locations in Fig. 1. Published records at
generic level only are generally not included in the locality lists. Though coverage
is basically restricted to the Pleistocene, Holocene records of Castor fiber and Mztcrotus
oeconomus are also included since both these species are now extinct in the British
Isles. Post-Pleistocene arrivals in the British Isles are briefly mentioned for the
sake of completeness.

Family SCIURIDAE Brandt 1855
Genus SCIURUS Linnaeus 1758
Sciurus whitei Hinton 1914
Squirrel, extinct
IQI4 Sciurus whiter Hinton : 193-195, fig. 10.
LocaLity : *West Runton, Norfolk (65) : Hinton 1914.

Hinton described a new species of squirrel, S. white1, from a unique fourth upper
premolar (BM(NH), M 10720) from the marine ‘Monkey Gravel’ overlying the
Cromerian Upper Freshwater Bed at West Runton, Norfolk.

74 PLEISTOCENE RODENTS

VA
ee 5 Ver cae : : :
7 eae Baa Fic. 5. Distribution of remains of Sciurus
os ines whitei Hinton in the British Isles.

No other fossil remains of this species have been found in the British Isles, though
a number of fir cones from Cromerian deposits of the Forest Bed of Norfolk bear
marks suggesting that they had been gnawed by squirrels (Newton 1882a). Newton
(1881, 1882a, 1891) described and figured a humerus from Ostend, Norfolk (57),
which agreed closely in form with that of the living red squirrel, S. vulgaris Linn.,
but he was not certain whether it came from the Forest Bed or from a Recent
alluvial deposit.

On the continent of Europe, remains of Sciwrus are known from a number of
Early and Middle Pleistocene localities in France and Hungary, though they have
usually been left without specific determination. Janossy (1962) discovered remains
of a squirrel approaching S. white: at the Middle Pleistocene locality of Tark6 in
Hungary, and described them as belonging to a new subspecies, S. whiter hungaricus
Janossy 1962. In his opinion S. white: may represent an ancestral form of S.
vulgaris.

Sciurus vulgaris Linnaeus 1758
Red squirrel

The red squirrel is a widespread Palaearctic species, still common in parts of the
British Isles. It occurs principally in woodland, but also in scrub beyond the
Arctic circle. Its remains have been found in Late Pleistocene sediments of Europe
from France to the Ukraine, but are nowhere common.

There are no indisputable records of S. vulgaris in the Pleistocene of the British
Isles. Its remains are known from Dowel and Langwith Caves, Derbyshire (71, 72 :
Bramwell 1960, Mullins 1913). The Dowel Cave record is from a Holocene stratum.
The stratigraphic position of the remains from Langwith Cave was not determined,
but the occurrence of Rattus rattus Linn., the black rat, suggests that some Holocene
mammalian remains were present in addition to the rich Pleistocene fauna of the
cave.

OlM” Is S ial MSI IaS 75

Genus SPERMOPHILUS Cuvier 1825 [= Citellus Oken 1816}!

It was long suspected that the ground squirrels of the British Pleistocene represent
two species. Hinton (in Barrett-Hamilton & Hinton 1910-21 : 724) wrote:

Dr Forsyth Major studied the material [of ground-squirrels] with great care many years ago
and we believe that he concluded that at least two species occur in the British Pleistocene ;
unfortunately his results were never published. The writer in turn has made some progress
with a similar investigation, but has not been able to complete his work yet. In his view also
there are two species at least, both extinct, one being allied to the living Cztellus [= Spermo-
philus| erythrogenys, the other more nearly related to C. [= Spervmophilus] eversmanni.

Following studies of specimens from the British Museum (Natural History) the
two forms have recently been distinguished by I. M. Gromov of Leningrad (im Itt.)
as Citellus [= Spermophilus| superciliosus, from the Upper Pleistocene, and C.
[= S.] primigenius from deposits of earlier age.

Spermophilus (Urocitellus) primigenius Kormos 1934
Ground squirrel, extinct

1876 Spermophilus ; Cheadle : 70-71.

1882b Spermophilus Altaicus ?; Newton: 51-54, pl. 2.

1885 Spermophilus erythrogenoides Falconer ; Lydekker : 212-213.
1934 ** Spermophilus primigenius Kormos.: 314-315, fig. 45.

1947 Citellus erythrogenoides Falc.; Jackson : 168.

LocaLities : Mundesley, Norfolk (60) : Newton 1882b.

Crayford and Erith, Kent (27): Cheadle 1876, Lydekker 1885-87, Whitaker
1889, Newton 1890a, 1894, Barrett-Hamilton & Hinton I9g10-21, Kennard 1944,
Jackson 1947.

DISTRIBUTION IN THE BRITISH ISLES. Remains of this species of ground squirrel
have been found abundantly in the Middle Terrace deposits of the Thames at Cray-
ford and Erith, Kent. Cheadle (1876) attributed remains which he had found at
Erith to Spermophilus sp. Subsequent writers (Lydekker, Whitaker, Newton,
Barrett-Hamilton & Hinton, Kennard, Jackson) regarded the form represented at
these localities as S. erythrogenoides. A skull from Erith was subsequently deter-
mined as Citellus [= Spermophilus] primigenius by Gromov, who wrote as follows
(in litt., translation) :

Specimen M 9605, a deformed skull with a fragment of mandible, belongs to a large form of
Citellus (Urocitellus) primigenius Korm., approaching in its dimensions the Polish C. (U.)
polonicus Gromoy, though it preserves the basic characters of the former. The skull also has
some characters typical for C. nogaici Top.: a well-marked basin which divides the hypo-
and entoconid parts on the grinding surface of M;, a weakly developed posterior crest on M3,
a weakly developed narrowing on the hypoconid of P,, and a foramen mentale shifted upwards.
The second of these characters, as well as the presence of tiny cusps on the bottom of the

1 Although the name Cifelius Oken 1816 has been extensively used hitherto for these ground squirrels,
Oken (1815-16) is not consistently binominal (Ellerman & Morrison-Scott 1966 : 3) and was rejected as
unavailable by fiat of the International Commission on Zoological Nomenclature (Opinion 417, 1956).

76 PLEISTOCENE RODENTS

external valleys of M,—M3;, occurs also in C. polonicus. Finally, peculiar to the British
remains is the forward shifting of the crests of the masseteric plate on the mandible. It is
probably a distinct subspecies.

Some isolated worn teeth of Spermophilus described by Newton (1882b : 54) as
close to S. altaicus (= S. eversmannt) were found in deposits now regarded as of early
Anglian age at Mundesley, Norfolk. They provide the earliest record of Spermo-
philus in the British Isles and are provisionally referred to S. (Urocitellus) primigentus.

GENERAL DISTRIBUTION. S. primigenius (type locality Kalkberg, Nagyharsany-
berg) is known from Villafranchian as well as from later (Giinz, Gtinz—Mindel)
deposits of Hungary. It is also present in the Middle Pleistocene of the Ukraine
and Germany (Gromov 1965). According to Chaline (1972), ground squirrels,
which he was unable to determine specifically, were present during the Mindel and
Riss glaciations in France.

The living representatives of the subgenus Uvocitellus — the long-tailed ground
squirrels S. undulatus (Pallas 1779) and S. parryi Richardson 1827 — inhabit steppes
as well as meadows in the tundra-zone of central and eastern Asia and North
America.

Spermophilus (Colobotis) superciliosus (Kaup 1839)
Ground squirrel, extinct

1839 6** Spermophilus superciliosus Kaup : 112.

1866 Spermophilus erythrogenoides (Falc.) ; Dawkins & Sanford : xxxix (nomen nudum).
1866 Spermophilus citillus Pallas; Dawkins & Sanford : xxxix.

1868 Spermophilus erythrogenoides ; Falconer, in Murchison : 452-454, pl. 35.

1974 Citellus sp.; Stuart: 246.

Localities: Langwith Cave, Derbyshire (72) : Mullins 1913, Barrett-Hamilton
& Hinton 1910-21.

Mendip Hills Caves (Bleadon Cave and others) : Dawkins & Sanford 1866,
Falconer 1868, Stevens 1869, Sanford 1870a, b, Newton 1882a, Barrett-Hamilton
& Hinton 1910-21, BM(NH).

Fisherton, Wiltshire (22): Falconer 1868, Stevens 1869, Barrett-Hamilton &
Hinton 1910-21.

Ightham Fissures, Kent (28) : Newton 1894, 1899b, Barrett-Hamilton & Hinton
Ig10—21, BM(NH).

Picken’s Hole (layer 3), Somerset (10) : Stuart (1974).

Pin Hole Cave, Derbyshire (73) : Jackson 1947.

DISTRIBUTION IN THE BriTIsH IsLEs. S. swperciliosus is a characteristic element
of the fauna of the Last Glaciation.

GENERAL DISTRIBUTION AND SYSTEMATIC REMARKS. S. superciliosus was described
by Kaup from Eppelsheim in Germany. It is an element of the Late Pleistocene
(Wiirm) fauna, present in almost the whole of Europe, including France, Germany,
Poland, the European and Asiatic part of the Soviet Union, Czechoslovakia and
Hungary. Its living relatives inhabit the steppe zone of eastern Europe and Asia.

OF BRITISH ISLES 77

~ > Fie. 6. Distribution of remains of Spermo-
: 0

\ philus (Urocitellus) primigenius Kormos
x ; S (Vv) and S. (Colobotis) superciliosus (Kaup)
) (@) in the British Isles.
a a
ae yp

In Britain it has usually been recorded as S. erythrogenoides, a name first used by
Falconer (in Murchison 1868) for remains from caves of the Mendip Hills. Hinton
(in Barrett-Hamilton & Hinton 1g10-21 : 723) stated that recent study tended to
show that this name must be treated as a synonym of S. superciliosus.

I. M. Gromovy of Leningrad studied the specimens from Ightham Fissures, Kent,
and made the following remarks (7 Jitt., translation) :

Two mandibular rami, no. M 11867, are typical remains of Cztellus (Colobotis) superciliosus
Kaup, broadly distributed in the northern part of continental Europe in Late Pleistocene ;
during some part of this time it evidently also inhabited the British Isles. From the
characteristic features of this species, as shown by the above-mentioned material, the following
can be stated: width of P, no more than 11 per cent greater than its length, processus
articularis short and broad below the head, posterior incision of the mandible relatively small,
fovamen mentale situated far from the anterior border of the masseteric plates, etc.

Family GLIRIDAE Thomas 1897
Genus MUSCARDINUS Kaup 1829

Hinton (in Barrett-Hamilton & Hinton Ig10—21 : 351) recorded a single tooth of
Muscardinus from the Forest Bed of Norfolk. This specimen has not been seen by
the present writers. Muscardinus is nevertheless known from some Cromerian
localities on the European continent so that its presence in Britain in Cromerian
times is not improbable.

Muscardinus avellanarius (Linnaeus 1758)
Common or hazel dormouse
Although remains of this species have been found at a number of cave sites — Dog

Holes and Pin Hole Caves (79, 73) : Jackson (1934, 1947) ; Great Doward Cave (17) :
British Museum (Natural History) — all these remains are probably of Holocene age.

78 PLEISTOCENE RODENTS

The hazel dormouse ‘is now distributed in southern, western and central Europe,
including southern Sweden and Britain, and in Asia Minor. On the continent of
Europe it has been found in interglacial deposits as early as Tiglian. It was also
present in France during the last interglacial (Chaline 1972).

Family CASTORIDAE Gray 1821
Genus TROGONTHERIUM Fischer 1809
Trogontherium minus Newton 1890
Giant beaver, extinct
1890b Trogontherium minus Newton : 447-448.
LocaLiTiEs : Red Crag of Felixstowe and *Woodbridge, Suffolk (45, 47) : Newton
18gob, 1891, 1902, Barrett-Hamilton & Hinton 1910-21, Schreuder 1929, 1951.

According to Schreuder (1951), who made a very thorough study of the genus
Trogontherium, this species is more primitive than the other representatives of the
genus, T°. cuviert Fischer and 7. borsvilleti (Laugel).

T. minus is also known from Astian (Pliocene) sediments of Perpignan, France.
The Red Crag specimens (like many other Crag fossils) are probably derived from
earlier deposits and cannot be regarded as part of the contemporary Lower Pleisto-
cene fauna.

Trogontherium boisvilletti (Laugel 1862)
Giant beaver, extinct

1846 Trogontherium Cuviert ; Owen : 184-189, figs 71-73.

1848 Diabroticus Schmerlingi ; Pomel : 167.

1862 ** Conodontes Boisvilletti Laugel : 715-718.

1866 Castor trogontherium Cuvier ; Dawkins & Sanford : xxxvi.
1902 Dipoides Lydekkeri Schlosser : 117.

1951 Trogontherium boisvilletti Laugel ; Schreuder : 403.

1956 Trogontherium sp. ; Spencer : 354.

LocaLiTiges: Sizewell, Suffolk, and Thorpe, Norfolk! (49, 48, Norwich Crag) :
Lydekker 1885, Schreuder 1951.

East Runton (64, Pastonian Forest Bed) : Newton 1882a, 1892, Barrett-Hamilton
& Hinton 1g10—21, Schreuder 1951.

Bacton, Cromer, Kessingland, Mundesley, Overstrand, Paston and West Runton
(58, 63, 52, 60, 62, 59, 65, ‘Cromerian’) : Lyell 1840, Owen 1846, Dawkins & Sanford
1866, Owen 1869, Newton 1881, 1882a, Lydekker 1885, Reid 1890, Hinton 1914,
Schreuder 1929, 1931.

Clacton, Copford, Hoxne and Swanscombe (43, 42, 54, 30, Hoxnian) : Brown 1852,
Newton 1902, Schlosser 1902, Stopes 1904, Newton 1916, Schreuder 1929, 1951,
Spencer 1956, Sutcliffe 1964, Singer et al. 1973.

RANGE: T. boisvilletti was apparently present in Britain during the Tiglian,
Cromerian and Hoxnian. Its remains have been found frequently in deposits of

1 See also footnote on p. 96.

OF BRITISH ISLES 79

Fic. 7. Distribution of remains of Tvogon-
therium boisvilletti (Laugel) in the British
Isles.

the Norfolk Forest Bed ; only a few Hoxnian specimens are known. These are a
solitary incisor from Ingress Vale, Swanscombe (Newton 1902, Stopes 1904, Schreuder
1929, 1951, Sutcliffe 1964), a cheek tooth from Copford (Brown 1852, site dated as
Hoxnian by Turner 1970, BM(NH) 27985), a femur and eight cheek teeth from
Hoxne (Spencer 1956) ; and part of a skull from Clacton (Singer e¢ al. 1973).

GENERAL DISTRIBUTION. TJ. botsvillettt is known from the early Pleistocene in
Britain, Holland, France (type locality Saint-Prest, near Chartres) and western parts
of Germany. Further to the east it is, according to Schreuder (1951), replaced by a
slightly different species, T. cuviert Fischer, first described from the border of the
Azov Sea. T. cuviert was present in Germany during the Holsteinian Interglacial.
According to Lehmann (1953), 7. cwviert is a more advanced species than T. borsvilletts
and may be its descendant. If this is true, the British Hoxnian remains may
represent 7. cuviert, but there are too few suitable specimens to allow a detailed
determination.

Genus CASTOR Linnaeus 1758
Castor fiber Linnaeus 1758
Beaver

1846 Castor europaeus ; Owen : 190-200, figs 74-75.

1864 Castor veterioy Lankester : 355-350.

1889 Castor fibey Linn. (= euvopaeus Owen) ; Whitaker : 336.
1908b Castor plicidens Maj. ; Major : 630-635, figs 132-136.
1964 Castor sp.; Spencer : 338.

LocaLitiEs: Sutton and Woodbridge, Suffolk (47, Red Crag) : Lankester 1864,
Newton 1891, Barrett-Hamilton & Hinton 1910-21, Schreuder 1929.

East Runton (64, Pastonian Forest Bed) : Major 1908b.

Bacton, Kessingland, Mundesley, West Runton (58, 52, 60, 65, Forest Bed) :
Dawkins & Sanford 1866, Owen 1869, Newton 1881, 1882a, 1891, Major 1908b,

7

80 PLEISTOCENE RODENTS

Barrett-Hamilton & Hinton 1910-21, Hinton 1914, Schreuder 1929, 1931, Friant
1962.

Westbury-sub-Mendip Fissure (13, lower middle Pleistocene) : Bishop 1974.

Clacton (43, Hoxnian) : Barrett-Hamilton & Hinton 1910-21, Hinton 1923b,
Sutcliffe 1964.

Swanscombe (Barnfield Pit), Kent (30, Hoxnian) : BM(NH).

Grays Thurrock and Ilford, Essex (33, 35) : Lankester 1864, Dawkins & Sanford
1866, Whitaker 1889, Hinton 1goob, Barrett-Hamilton & Hinton 1g10—21, BM(NH).

Selsey, Sussex and Upnor, Kent (24, 31, Last Interglacial) : BM(NH).

Kent’s Cavern, Devon (6) : BM(NH).

Cambridge Fens, Cambridgeshire ; Hay Wood Rockshelter, Somerset ; Staple
Howe, Yorkshire-; Star Carr, Yorkshire; Yhatcham, Berks (41) 10, 75, 76, 23,
Holocene) : Montagu 1924, Brewster 1963, Fraser & King 1954, Wymer 1962.

Loch of Marlee, Perthshire, and Middlestots Bog, Edrom Parish, Berwickshire
(82, 80) : Barrett-Hamilton & Hinton 1910-21.

iy EA

oa mal Fic. 8. Distribution of fossil remains of

Castor fibey Linn. in the British Isles.

RANGE. Remains of C. fiber have been found in deposits in the British Isles
ranging in age from lowest Pleistocene (Red Crag) to Holocene. The species spread
to Scotland but apparently never reached Ireland. Barrett-Hamilton & Hinton
(Ig10—21) considered that the beaver did not become extinct in Britain before the
thirteenth century A.D.

GENERAL DISTRIBUTION. Although limited to a small number of relict localities
as a result of extermination by man, C. fiber is still widely distributed in Europe and
Asia. Its fossil remains are known from the Pliocene and Pleistocene of Europe and
Asia.

SYSTEMATIC REMARKS. The beaver remains from the Red Crag were described
by Lankester (1864) under the new name C. veterior. Newton (1891) considered that
this species was conspecific with C. fiber and Schreuder (1929) proved it by a detailed
comparison. C. I. F. Major (19g08a) described some remains from the Forest Bed
at East Runton as belonging to C. plicidens, a species described by him from Val

OR SS RIGS ESSE Ss 81

d’Arno in Italy. Schreuder (1929) showed that the folding of tooth-enamel is a
character depending upon the individual age of the animal and C. plicidens is there-
fore a synonym of C. fiber. Later it was found that individuals of ‘C. plicidens’
sometimes appear among postglacial beavers. Viret (1954), however, studying the
skull of a Villafranchian beaver from St Vallier in France, found some morphological
differences from the Recent beaver and was inclined to retain the name C. plicidens
for all the beavers from that period, not only for specimens with folded enamel.
According to Lehmann (1957) these differences are rather of a subspecific character
and he designated the Villafranchian beavers as ‘Castor fiber plicidens Major 1875’.
Until statistical analysis of the variability of the living beaver has been made, it
seems preferable to use only the specific name Castor fiber Linn. for the Astian (Upper
Pliocene) as well as for the Pleistocene beavers.

Family MURIDAE Gray 1821
Genus APODEMUS Kaup 1829
Apodemus sylvaticus (Linnaeus 1758)
Wood mouse

1846 Mus musculus (?) ; Owen : 2009, fig. 79.

1881 Mus sylvaticus ; Newton : 258—259.

1910 Micromys (= Mus) sylvaticus (Linné) ; Jackson : 328.
1910b Mus sp., allied to M. sylvaticus ; Hinton : 489-507.
1915 A podemus sp. ; Hinton : 580.

IQ15 A podemus whitei Hinton : 580-581.

1915 Apodemus sylvaticus L.; Hinton : 581-582.

Localities: West Runton (Upper Freshwater Bed), Norfolk (65) : Newton 1881,
1882a, 1891, Barrett-Hamilton & Hinton 1g10—21, Hinton 1915, BM(NH).

Westbury-sub-Mendip Fissure, Somerset (13) : Bishop 1974.

Hitchin, Hertfordshire (39) : Carreck 1959.

Swanscombe (Ingress Vale), Kent (30): Stopes 1904, Hinton Ig1ob, Barrett-
Hamilton & Hinton rg10—21, Hinton 1915, Sutcliffe 1964, BM(NH).

Hoxne, Suffolk (54) : Wolff (7 litt., Apodemus sp.).

Grays Thurrock, Essex (33) : Hinton, Kennard & Newton 1900, Hinton 1915.

Stutton, Suffolk (44) : Stuart 1974 (A. sylvaticus group).

Bacon Hole, Glamorganshire (16) : det. A. J. Stuart (im litt.).

Kirkdale Cave, Yorkshire (77) : Owen 1846, Dawkins & Sanford 1866.

Swanton Morley, Norfolk (66) : Stuart 1974 (A. sylvaticus group).

Tornewton Cave, Devon (3) : Kowalski 1967, BM(NH).

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924, BM(NH).

Dog Holes Cave, Lancashire (79) : Jackson 1910, BM(NH).

Great Doward Cave, Herefordshire (17): Bristol University Spelaeological
peciery, det. K. K.

Gwaenysgor Cave, Flintshire (68) : Jackson 1947.

Ightham Fissures, Kent (28) : Newton 1894, Barrett-Hamilton & Hinton 1910-21,
Hinton 1915, Carreck 1957, BM(NH).

82 PLEISTOCENE RODENTS

Langwith Cave, Derbyshire (72) : Mullins 1913.

Levaton Cave, Devon (2) : Carreck 1957.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Rowberrow Cavern, Somerset (11) : BM(NH).

Dowel Cave, Derbyshire (71) : Bramwell 1960.

Happaway Cave, Devon (5) : BM(NH).

Lynx Cave, Denbighshire (67) : Blore 1966.

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Joint Mitnor Cave, Devon (1, Layer 10, Holocene) : BM(NH).

Ballynamintra Cave, Co. Waterford (90) : Barrett-Hamilton & Hinton Ig10—21,
Jackson 1929b.

Castlepook Cave, Co. Cork (87) : Scharff, Seymour & Newton 1918, BM(NH).

Edenvale Caves (Alice and Gwendoline, Catacombs, Newhall, Barntick), Co. Clare
(85) : Barrett-Hamilton & Hinton 1910-21, Scharff 1906.

Keshcorran Caves (Coffey, Plunkett), Co. Sligo (84) : Barrett-Hamilton & Hinton
Ig10—21, Jackson 1929b, Scharff e¢ al. 1903.

Kilgreany Cave, Co. Waterford (89) : Jackson 1929b.

Fic. 9g. Distribution of fossil remains of
A podemus sylvaticus (Linn.) in the British
Isles.

DISTRIBUTION IN THE BriTISH ISLES. A. sylvaticus was present in the British
Isles during the Cromerian, Hoxnian and Ipswichian interglacials. It probably
disappeared from the country at the time of the cold stage represented by the
Tornewton Cave Glutton Stratum. A solitary specimen from this deposit may have
been introduced from a higher, interglacial, level. This species was present during
the Last Glaciation and during the Holocene. Most of the localities, especially
caves, cannot be accurately dated, and it is difficult to decide if A. sylvaticus was
present throughout the Holocene without interruption. It may have reached Ireland
during Upper Pleistocene times.

GENERAL DISTRIBUTION. A. sylvaticus is now widely distributed in Europe,
north Africa and western Asia. The fossil remains of this or nearly related species
are known from the Pleistocene of Europe and China. The oldest of them are dated

OF BRITISH ISLES 83

as Tiglian, with further records from deposits of Cromerian, Holsteinian and later
date.

SYSTEMATIC REMARKS. Hinton (1915 : 580) referred the teeth from the Cromerian
Upper Freshwater deposits of West Runton to Apodemus sp. Although he found
that their pattern was indistinguishable from that of A. sylvaticus he considered
that the material was insufficient for fine determination. From Swanscombe
(Ingress Vale) (oc. cit.) he described a new species, which he named A. whitez, from
part of a maxilla differing slightly from recent specimens of A. sylvaticus. However,
the characters concerned lie within the range of individual variability and there does
not seem to be any good reason for regarding the West Runton form as a distinct
species.

Apodemus flavicollis (Melchior 1834)
Yellow-necked mouse

1894 Mus Abbotii Newton : 195, pl. xi, fig. 8 (not M. abbott: Waterhouse).
1899a Mus Lewist Newton : 381.

IQI5 A podemus lewist Newton ; Hinton : 582-584.

1921 A podemus Flavicollis Melchior ; Hinton : 78.

LocaLiTiEs : Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924.
Dowel Cave, Derbyshire (71) : Bramwell 1960.

Etches’ Cave, Derbyshire (71) : Pernetta 1966.

Happaway Cave, Devon (5) : Hinton 1915, BM(NH).

Ightham Fissures, Kent (28) : Newton 1894, 1899a, b, Hinton 1915, BM(NH).
Kent’s Cavern, Devon (6) : Hinton 1915.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

DISTRIBUTION IN THE BriTIsH ISLES. Remains of A. flavicollis have been found
almost exclusively in very late sediments. Many of the records must nevertheless
be considered doubtful in view of the great difficulty in distinguishing remains of
this species from those of A. sylvaticus. The remains recorded from Dowel Cave
were found in a Neolithic layer (Bramwell 1960 : [10]—pages unnumbered) and those
from Etches’ Cave were in early postglacial layers (Pernetta 1966: 12,15). Hinton
(1921) considered that the Aveline’s Hole remains were probably a late introduction.
The same may be true for the Ightham Fissures. The Kent’s Cavern and Happaway
specimens were excavated a century ago. From Pin Hole Cave, Derbyshire
(Jackson 1934), remains of A. flavicollis have been recorded from both the superficial
levels and at a depth of more than 3 m (10 ft), where they were apparently associated
with Upper Pleistocene faunal remains.

GENERAL DISTRIBUTION. At the present time A. flavicollis is widespread in
Europe and east as far as the Urals and Caucasus. It is absent from a large part of
western Europe (France, Belgium, Holland), but is present in England and southern
Sweden. Nothing is known about the fossil remains. It is absent from the
Pleistocene of France (Chaline 1972). It is difficult to decide whether this species
represents a natural migration at the end of Pleistocene or during the Holocene or
whether it was brought by man.

84 PLEISTOCENE RODENTS

SYSTEMATIC REMARKS. According to Hinton (1915), A. lewisz, described from the
Ightham Fissures by Newton (18g99a, b), is closely related to, if not identical with,
A. flavicollis Melchior. Jackson (1934) determined his material as ‘A. flavicollis
(= A. lewist)’. The lack of an anterior accessory cusp on M, 1s, according to Hinton
(1915), of no significance in specific determination, because the cusp is apparently
absent in slightly worn teeth of A. flavicollis. It seems probable that the holotype
of A. lewist is nothing but a Recent specimen of A. flavicollis.

Genus MICROMYS Dehm 1841
Micromys minutus (Pallas 1771)
Harvest mouse

This species was probably introduced into Britain by man in postglacial times.
According to Barrett-Hamilton & Hinton (1910-21) it is unknown there in a fossil
state. Remains found at Tornewton Cave (Kowalski 1967) are from a surface
fissure unconnected with the main sequence of deposits and are unlikely to be of any
antiquity.

M. minutus, now widely distributed in Europe and Asia, is present in England
and perhaps in the southern parts of Scotland. The genus Micromys, and probably
the Recent species, is known in continental Europe from the late Pleistocene.

Genus MUS Linnaeus 1758
Mus musculus Linnaeus 1758
House mouse
The house mouse was introduced into Britain by man. Barrett-Hamilton &
Hinton (1910-21) discussed some early data about its fossil occurrence in Britain

and considered that all are erroneous or refer to Recent remains. There is no
evidence for the occurrence of the house mouse in Britain before the Iron Age.

Genus RATTUS Fischer 1803
Rattus rattus (Linnaeus 1758)
Black rat
Although remains of the black rat have been recorded from Aveline’s Hole

(Hinton 1921) and Langwith Cave (Mullins 1913) these are probably Recent. The
black rat was introduced into the British Isles by man during the historical period.

Rattus norvegicus (Berkenhout 1760)

Brown rat

The brown rat is a recent addition to the British mammalian fauna. Jackson —
(1929b) records its remains from Kilgreany Cave, Ireland, in an upper layer with
remains of domestic animals and prehistoric material, but these are probably of very
recent date.

ORB RI DISH Sil s 85

Family CRICETIDAE Rochebrune 1883
Genus CRICETUS Leske 1779
Cricetus cricetus (Linnaeus 1758)
Common hamster

1909 Cricetus vulgaris Runtonensis Newton : 110—113, fig. (unnumbered).
1967 Cricetus cricetus (Linnaeus 1758) ; Kowalski: 119-120.

LocALITIEs : West Runton (Upper Freshwater Bed), Norfolk (65) : Newton 1909,
Hinton r1910b, Barrett-Hamilton & Hinton 1r9g10—21, Osborn 1922, BM(NH).
Tornewton Cave (Glutton Stratum), Devon (3) : Kowalski 1967, BM(NH).

S aD) av
FAR ree :
y Ape \ Fic. to. Distribution of fossil remains of
le

: i, rm ( : ~ Cricetus cricetus (Linn.) in the British Isles.
oo ss

DISTRIBUTION IN THE BriTiIsH IsLEs. Representatives of the genus Cricetus
appeared twice in Britain: in the Cromerian interglacial and during the cold
stage represented by the Tornewton Cave Glutton Stratum. The Tornewton Cave
remains represent at least two individuals.

GENERAL DISTRIBUTION. C. cricetus, the only living member of its genus, occurs
at the present time in eastern Europe and in the western part of Asia. It is also
present in western and central Europe where its range is insular and seems to be
characterized by gradually increasing expansion. It occurs in steppes, parklands
and meadows. In Europe its occurrence is now associated with arable fields.
Fossil remains of Cricetus are known from the early, middle and late Pleistocene of
western, central and eastern Europe, especially in deposits indicating a cold and arid
environment.

SYSTEMATIC REMARKS. All the fossil remains of the genus Cricetus are very
uniform in their dental morphology and do not differ from the Recent ones. They
do, however, vary greatly in size. On this basis many new forms have been described,
some under specific, others under subspecific names. The specimen from the
Cromer Forest Bed, BM(NH) M18352, an upper jaw with a tooth-row 9-3 mm long,

86 PLEISTOCENE RODENTS

is the holotype of a form known as Cricetus runtonensis or C. cricetus runtonensis.
According to Schaub (1930), it is a synonym of C. c. major Woldiich described from
the late Pleistocene deposits of Czechoslovakia. Other palaeontologists use the
name C. runtonensis to designate larger representatives of Cricetus from the early
Pleistocene, usually without giving sufficient morphological reasons for its distinctness
from later populations. According to Fahlbusch (1969), there are some peculiarities
in the pattern of M, in C. runtonensis, but a similar morphology can also be found
in some Recent specimens. In the opinion of one of us (K. K.) the large early
Pleistocene hamsters are at most subspecifically distinct from those of the late
Pleistocene and present day. The great variability of this species in the Pleistocene
was connected with changing climatic conditions.

Genus ALLOCRICETUS Schaub 1930
cf. Allocricetus bursae Schaub 1930
Hamster, extinct
1870a Cricetus songarus ; Sanford : 51, 56, fig. 6.
1870b Cricetus (Mus) songarus (Pallas) ; Sanford : 128-129; pl. viii, fig. 6.

1913 Phodopus sanfordi Hinton, in Barrett-Hamilton & Hinton 1910-21 : 383, fig. 53.
1967 Phodopus songorus (Pallas) ; Kowalski : 113-114.

LocaLiTIES: Hutton Cave, Somerset (10): Sanford 1870a,b, Newton 1909,
Barrett-Hamilton & Hinton 1910-21, BM(NH).
Tornewton Cave (Glutton Stratum), Devon (3) : Kowalski 1967, BM(NH).

|
Ms
Zan
i i S AREF Fic. 11. Distribution of remains of cf.
$, oe eS D Allocricetus bursae Schaub in the British
C Isles.
SS
ZL Wee
ge

—_

@

DISTRIBUTION IN THE BritisH IstEs. The occurrence of this species in the
Glutton Stratum of Tornewton Cave shows that. it was present during the pre-
hippopotamus cold stage represented by this deposit. Although the age of the
Hutton Cave fauna is unknown, the occurrence of remains of a relatively primitive

OF BRITISH ISLES 87

form of mammoth suggests that this record is at least as early as that from Tornewton
Cave.

GENERAL DISTRIBUTION AND SYSTEMATIC REMARKS. Small representatives of
Cricetidae are common in the early, middle and late Pleistocene of Europe. The
uniform structure of the teeth in Cricetidae makes the systematic determination of
their remains very difficult. According to Schaub (1930), the small hamsters of the
late Pleistocene of Europe should be referred to Phodopus songorus. Janossy (1961)
later stated that Schaub based his determination on incorrectly determined Recent
specimens and claimed that the small hamster described by him is conspecific with
Recent Cricetulus migratorius. The same name was used for English specimens by
Kurtén (1969). Chaline (1972) did not find this species in France, but determined
all late Pleistocene small French hamsters as Allocricetus bursae Schaub 1930. A
more detailed study of English specimens will be necessary before it is possible to
decide whether they belong to one or two species and what species there are.

In the earliest description of British hamster remains (those from Hutton),
Sanford (1870) applied the name Cricetus songarus. Hinton (in Barrett-Hamilton &
Hinton 1910-21 : 383) considered that the remains should be placed in the genus
Phodopus and that they could not be synonymized with P. songorus but needed a
new name, for which he proposed P. sanfordi. No reasons for this decision are
given. If the British specimens are conspecific with A. bursae Schaub then this
name is a junior synonym of P. sanford: Hinton and must be replaced by it.

Small hamsters are now an element of the steppe fauna, and their species are
broadly distributed in south-eastern Europe as well as in western and central Asia.
C. migratorius ranges from south-eastern Europe to China. This species, partly
cited under the name of A. bursae Schaub, is known as a fossil from the late Pleisto-
cene of Germany, Switzerland, Spain, France, Czechoslovakia, Hungary, Poland
and the Soviet Union.

Genus DICROSTONYX Gloger 1841
Dicrostonysx torquatus (Pallas 1779)
Collared lemming

1869 Lemmus torquatus ; Stevens : 110, I13.

1870a Arvicola Gulielmi Sanford: 51, 55, fig. 2.

1870b = Arvicola Gulielmi Sanford: 125, pl. viii, fig. 2.

1874 Myodes torquatus Pall.; Blackmore & Alston : 469-470.
I9g10a Dicrostonyx gulielmit Sanford ; Hinton : 38—39.

Ig1oa Dicrostonyx henseli Hinton : 37-38.

1960 Dicrostonyx sp.; Bramwell: [10].

LocatitiEs : Westbury-sub-Mendip, Somerset (13) (Dicrostonyx sp., lower middle
Pleistocene) : Bishop 1974.

Erith, Kent (27): Newton 1890a, Hinton 1g10b, Barrett-Hamilton & Hinton
Ig10—21, Hinton 1926b, Kennard 1944, Jackson 1947.

Hutton Cave, Somerset (10) : Stevens 1869, Sanford 1870a, b, Blackmore & Alston
1874, Jackson 1909, Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, BM(NH).

88 PLEISTOCENE RODENTS

Tornewton Cave (Glutton Stratum and Reindeer Stratum), Devon (3) : Kowalski
1967, BM(NH).

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924, 1926b, BM(NH).

Chudleigh Fissure, Devon (7) : Hinton 1926b.

Dog Holes Cave, Lancashire (79) : Jackson 1909, 1910, 1929a, Barrett-Hamilton &
Hinton 1910-21, Hinton 1926b, Jackson 1947, BM(NH).

Elder Bush Cave, Staffordshire (69) : Bramwell 1964.

Fisherton, Wiltshire (22) : Stevens 1869, Blackmore & Alston 1874, Jackson 1909,
Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, BM(NH).

Great Doward Cave, Herefordshire (17): Bristol University Spelaeological
Society; det_ kK. Ke

Gwaenysgor Cave, Flintshire (68) : Jackson 1932 (fide Jackson 1947), Jackson 1947.

Ightham Fissures, Kent (28): Newton 1894, Jackson 1909, Hinton rgr0a, b,
Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, Jackson 1929a, 1947, Zimmer-
mann 1959, BM(NH).

Kent’s Cavern, Devon (6) : BM(NH).

King Arthur’s Cave, Herefordshire (17) : BM(NH) (Hinton collection).

Langwith Cave, Derbyshire (72) : Mullins 1913, Hinton 1g10a, Barrett-Hamilton
& Hinton 1910-21, Hinton 1926b, Jackson 1929a, 1947, BM(NH).

Merlin’s Cave, Herefordshire (18) : Hinton 1925, 1926b, BM(NH).

Murston, Kent (32) : Newton 1890a, Barrett-Hamilton & Hinton 1910-21.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Rowberrow Cavern, Somerset (11) : BM(NH) (Hinton collection).

Upton Warren, Worcestershire (20) : Coope et al. 1961.

Brean Down, Somerset (9) : Apsimon e¢ al. 1961.

Fox Hole Cave, Derbyshire (71) : Bramwell 1970.

Dowel Cave, Derbyshire (71) : Bramwell 1960.

Lynx Cave, Denbighshire (67) : Blore 1966.

Etches’ Cave, Derbyshire (71) : Pernetta 1966.

Hackney, London (36) : BM(NH).

Angel Road, Middlesex (36) : Hinton 1912, Barrett-Hamilton & Hinton I910-21,
Hinton 1926b, BM(NH).

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Ponders End, Middlesex (36) : Hinton 1926b.

Corstorphine, nr. Edinburgh, Scotland (81) : Evans 1907, 1913, Barrett-Hamilton
& Hinton 1g10—21, Jackson 1g29b.

Creag nan Uamh Cave, Sutherland (83) : Peach & Horne 1917.

Ballynamintra Cave, Co. Waterford, Ireland (go) : Coleman 1965.

Castlepook Cave, Co. Cork (87) : Ussher et al. 1908, Jackson 1910, Ussher 1910,
Hinton r910a, Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, Jackson 1929b,
1947, Coleman 1965.

Castletownroche Cave, Co. Cork (88) : Coleman 1965.

Edenvale Caves (Alice and Gwendoline, Catacombs, Newhall), Co. Clare (85) :
Scharff 1906, Hinton 1926b, Coleman 1965, BM(NH).

OF BMITISH AISLES 89

Keshcorran Caves (Coffey, Plunkett), Co. Sligo (84) : Jackson 1909, Scharff e¢ al.
1903, Scharff, Seymour & Newton 1918, Barrett-Hamilton & Hinton rg1o-21,
Hinton 1926b, Jackson 1929b, Coleman 1965, BM(NH).

Kilgreany Cave, Co. Waterford (89) : Jackson 1929b, Coleman 1965.

Red Cellar Cave, Co. Limerick (86) : Coleman 1965.

Brandon, Warwickshire, and Penkridge, Staffordshire, are two additional Dicro-
stonyx localities not marked on Fig. 1, which were notified to us by Mr J. Carreck
(in litt.) after the compilation of the above list.

§ : esa tg Fic. 12. Distribution of fossil remains of
ati aS Dicrostonyx torquatus (Pallas) in the
nnaiial A British Isles.
= Urs An ‘
ae
Be ; }
ee ° af
J : a
3 oe,

DISTRIBUTION IN THE BritisH IsLtes. Dicrostonyx is known in the British Isles
from at least three stages of the Pleistocene. The earliest record (Dicrostonyx sp.)
is from the lower middle Pleistocene site of Westbury-sub-Mendip. The second
stage is represented by remains from the Corbicula bed at Erith and the Glutton
Stratum of Tornewton Cave. At the last-mentioned site Dicrostonyx disappeared
in the overlying Last Interglacial deposits and reappeared again in the Last Glaci-
ation Reindeer Stratum. During the Last Glaciation it reached its greatest climax,
spreading throughout England, Wales, Scotland and Ireland. Its remains are
especially abundant in deposits dating from the later part of this stage and it was
clearly a dominant element of the tundra faunal assemblage of that time. At
Nazeing, Essex, it apparently survived until late Glacial pollen Zone III.

GENERAL DISTRIBUTION. The collared lemming is now widely distributed in the
tundra belt of the Holarctic, in both Eurasia and North America. The American
subspecies are sometimes recognized as specifically different from those of Eurasia.
Fossil remains are known from the deposits of four different glaciations in Europe
and they are also present in Asia and North America. Pre-Cromerian collared
lemmings from Czechoslovakia and Poland are morphologically different from later
ones and were named Dicrostonyx simplicior Fejfar 1966. During the Last Glaciation
D. torquatus in Europe reached as far south as Hungary, Switzerland and central
France.

90 PLEISTOCENE RODENTS

SYSTEMATIC REMARKS. Hinton (1g10a) distinguished two species of fossil
collared lemmings in the British Isles, Dicrostonyx gulielmi and D. henselt. Accord-
ing to him the differences between these were as great as those between different
Recent species of the genus Dicrostonyx. It is now generally accepted, however,
that disregarding the Canadian D. hudsonius (which does not concern us here) all
living forms belong to one species. The European fossil forms may, at the most,
be subspecifically different from Recent ones, but this seems not to be the case as the
two forms described by Hinton are present together at nearly all localities. Janossy
(1954) found in Hungarian caves not only all possible intermediate forms between
D. gulielmi and D. henseli, but also specimens with left tooth-row of one type and
right one of the other. Most palaeontologists are of the opinion that there was only
one species of the collared lemming in the Upper Pleistocene of Europe and that it
was conspecific with the living D. torquatus. Like the Recent population, it was
very variable (Mandach 1938). Recently Agadzhanian (1973) has stated that,
among the populations of Recent and fossil collared lemmings, there are different
morphotypes, differing in the grade of complication of the crown pattern of their
molars. He considered that the simpler morphotypes are more numerous among
the populations from the Last Glaciation of Eurasia than among Recent animals from
north Asia. For this reason he favoured preserving the name D. gulielmi as the
oldest available designation for the collared lemmings from the time of the Last
Glaciation of Eurasia.

Genus LEMMUS Link 1795
Lemmus lemmus (Linnaeus 1758)
Norwegian lemming

1870a Lemmus norvegicus (var.) ; Sanford : 51, 56, fig. 4.

1870b Lemmus norvegicus ; Sanford : 125-126, pl. viii, fig. 4.
1874 Mvyodes lemmus (Linn.) ; Blackmore & Alston : 470-471.
1921 Lemmus lemmus Linn. ; Hinton : 75-76.

1950 Lemmus sp. ; Schreuder : 629, 633—634.

LOCALITIES : Westbury-sub-Mendip, Somerset (13) (Lemmus sp., lower middle
Pleistocene) : Bishop 1974.

Hoxne, Suffolk (54) : Wolff (cn litt.).

Swanscombe (Barnfield Pit, Upper Middle Gravel), Kent (30) (Lemmus sp.) :
Schreuder 1950.

Erith, Kent (27): Newton 1890a, Hinton 1g1ob, Barrett-Hamilton & Hinton
1910-21, Hinton 1926b, Kennard 1944, Jackson 1947, BM(NH).

Hutton Cave, Somerset (10): Sanford 1870a,b, Blackmore & Alston 1874,
Jackson 1909, Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, BM(NH).

Tornewton Cave (Glutton Stratum), Devon (3) : Kowalski 1967, BM(NH).

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1926b, BM(NH).

Chudleigh Fissure, Devon (7) : Hinton 1926b.

Dog Holes Cave, Lancashire (79): Jackson 1909, 1910, Barrett-Hamilton &
Hinton 1910-21, Hinton 1926b, BM(NH).

ORSSRITISH: LSVES QI

Elder Bush Cave, Staffordshire (69) : Bramwell 1964.

Great Doward Cave, Herefordshire (17): Bristol University Spelaeological
pociety, det. K. K.

Ightham Fissures, Kent (28) : Newton 1894, Bate 1901, Jackson 1909, Hinton rgtoa,
Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, Zimmermann 1959, BM(NH).

Kent’s Cavern, Devon (6) : Kennard 1945-46.

Langwith Cave, Derbyshire (72): Mullins 1913, Barrett-Hamilton & Hinton
1g10—21, Hinton 1926b.

Merlin’s Cave, Herefordshire (18) : Hinton 1924, 1926b, BM(NH).

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Uphill Cave, Somerset (9) : Barrett-Hamilton & Hinton 1g10~—21, Hinton 1926b,
BM(NH).

Dowel Cave, Derbyshire (71) : Bramwell 1960.

Etches’ Cave, Derbyshire (71) : Pernetta 1966.

Harborough Cave, Derbyshire (70) : Jackson 19292.

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Castlepook Cave, Co. Cork, Ireland (87) : Ussher e¢ al. 1908, Ussher 1910, Scharff,
Seymour & Newton 1918, Barrett-Hamilton & Hinton 1910-21, Jackson 1g29b,
Coleman 1965, BM(NH).

Castletownroche Cave, Co. Cork (88) : Coleman 1965.

After the compilation of the above list we have been informed by Mr J. Carreck
(in litt.) that he has also determined remains of Lemmus lemmus from deposits of
Ipswichian (Zone IIb) age at Wretton, Norfolk (not marked on Fig. 1).

gL.

; 3 Z
Bh &

a

iss
Y ‘Q ; Fic. 13. Distribution of fossil remains of
Te \y \ Lemmus lemmus (Linn.) in the British Isles.

et. cson. 3 4
z ie ¥
ee

Bee

S043
° ee

DISTRIBUTION IN THE BritisH IsLEes. L. lemmus is distributed in Pleistocene
sediments of England and southern Ireland. In many localities it has been found
together with Dicrostonyx torquatus, but it is usually less numerous. Most of the
fossil localities with Lemmus are of Last Glaciation age. Its occurrence at Nazeing,
where it was found in peaty muds lying deeper than the calcareous muds in which

92 PLEISTOCENE RODENTS

Dicrostonyx torquatus appears for the last time, seems to be the latest well-dated
evidence of its presence in Britain. This may indicate that Lemmus disappeared
slightly earlier than Dicrostonyx, though it could also be explained by lack of
sufficient specimens.

Lemmus is known from at least four stages of the British Pleistocene. Earlier
records are the lower middle Pleistocene site of Westbury-sub-Mendip, the Hoxnian
of Hoxne and Swanscombe (Barnfield Pit, Upper Middle Gravel), and the stage or
stages represented by the Corbicula Bed of Crayford and Erith and the Glutton
Stratum of Tornewton Cave.

GENERAL DISTRIBUTION. The genus Lemmus has a circumpolar distribution
ranging in the Old World from Scandinavia to the Bering Strait and including the
arctic regions of North America. The differences between the Eurasiatic and
American forms do not justify their distinction as different species (Sidorowicz
1964) : they can all be treated as subspecies of a single species, L. lemmus. Lemmus
coexists in a great part of its range with Dicrostonyx, but is generally more southerly
in distribution : it does not occur in Greenland or on the Canadian islands north of
Viscount Melville Sound, where Dicrostonyx is abundant.

The genus Lemmus appeared very early in Europe. Its remains are known from
middle Villafranchian deposits of Hungary, Poland and Germany. During the
early Pleistocene it extended further south in eastern Europe than Dicrostonyx,
and was present, for example, in Romania, where Dicrostonyx never penetrated. Its
remains have been found at most early and middle Pleistocene fossil localities of
central and western Europe and the genus is a characteristic element of the arctic
fauna of the last two glaciations.

SYSTEMATIC REMARKS. Fossil remains of Lemmus from the late Pleistocene of the
British Isles and from the continent of Europe are indistinguishable, on the basis of
teeth and preserved skeletal remains, from those of Recent L. lemmus. Villa-
franchian to middle Pleistocene remains are specifically different from the Recent
form.

Genus CLETHRIONOMYS Tilesius 1850
Clethrionomys glareolus (Schreber 1780)
Bank vole

1846 Ayrvicola pratensis ; Owen : 208, fig. 78.

1870a Arvicola glaveolus (= pratensis) ; Sanford : 56.

1870b = Arvicola glareolus (Schreber) = pratensis (Baillon) = riparia (Yarrell) ; Sanford : 124.
1882a Avvicola (Evotomys Coues & Allen) glaveolus Schreber; Newton: 82-83, pl. xiv,

fig. I—Ic.
1g00 Microtus (Evotomys) glaveolus Schreb. ; Hinton, Kennard & Newton : 347-349.
1gOI Microtus glaveolus ; Hinton : 142.
1910b = Evotomys sp. ; Hinton: 492, 494, 497.
1926 Evotomys harrisont Hinton : 216-217, pl. vii, fig. 2.

1926 Evotomys kennardi Hinton : 225-226, pl. vii, fig. 3.
1950 Clethrionomys sp. ; Schreuder : 629, 634-635.

OF ERIGISE ISLES 93

Localities : West Runton (Upper Freshwater Bed), Norfolk (65) : Blackmore &
Alston 1874, Newton 1882a, 1891, Hinton 1g1ob, Osborn 1922, Hinton 1926b,
BM(NH).

Westbury-sub-Mendip, Somerset (13) : Bishop 1974.

Clacton, Essex (43) : Singer e¢ al. 1973.

Hitchin, Hertfordshire (39) : Carreck 1959.

Swanscombe (Barnfield Pit, Upper Middle Gravels and Ingress Vale), Kent (30) :
Stopes 1904, Hinton 1926b, Schreuder 1950, Sutcliffe 1964, Carreck (in litt).

Aveley, Essex (34) : BM(NH).

Grays Thurrock, Essex (33) : Hinton, Kennard & Newton Igoo, Hinton Igor,
1g10b, 1926b, BM(NH).

INorthileet, Kent (29) : det. A. J. Stuart (7m liz.).

Hutton Cave, Somerset (10) : Sanford 1870a, b, Blackmore & Alston 1874.

Tornewton Cave (Glutton Stratum and Reindeer Stratum), Devon (3) : Kowalski
1967, BM(NH).

Alveston Fissure, Gloucestershire (15) : det. G. B. Corbet.

Bacon Hole, Glamorganshire (16) : Stuart (77 Jitt.).

Kirkdale Cave, Yorkshire (77) : BM(NH).

Swanton Morley, Norfolk (66) : Stuart 1974.

West Wittering, Sussex (24) : BM(NH).

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924, BM(NH).

Brixham Cave, Devon (4) : Lydekker 1885, Hinton 1926b, BM(NH).

Chudleigh Fissure, Devon (7) : BM(NH).

Dog Holes Cave, Lancashire (79) : Jackson 1910, BM(NH).

Elder Bush Cave, Staffordshire (69) : Bramwell 1964.

Great Doward Cave, Herefordshire (17) : BM(NH).

Gwaenysgor Cave, Flintshire (68) : Jackson 1947.

Ightham Fissures, Kent (28) : Newton 1894, Hinton 1g10b, 1926b, Zimmermann
1959, BM(NH).

Kent’s Cavern, Devon (6) : Owen 1846, Dawkins & Sanford 1866, Blackmore &
Alston 1874, Lydekker 1885, Kennard 1945-46, BM(NH).

Langwith Cave, Derbyshire (72) : Mullins 1913.

Merlin’s Cave, Herefordshire (18) : Hinton 1925.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Rowberrow Cavern, Somerset (11) : BM(NH).

Dowel Cave, Derbyshire (71) : Bramwell 1960.

Happaway Cave, Devon (5) : BM(NH).

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Joint Mitnor Cave (layer X, Holocene), Devon (1) : BM(NH).

DISTRIBUTION IN THE BriTIsH ISLES. The bank vole appeared in England for
the first time during the Cromerian Interglacial. It was also present during the
Hoxnian (Hitchin, Swanscombe). Although a few remains were found in the pre-
hippopotamus Glutton Stratum of Tornewton Cave, this species is lacking from
typical faunal assemblages of this time, for example from Clevedon Cave. It isa
common element in deposits of Last Interglacial, Last Glaciation and Holocene age.

04 PLEISTOCENE RODENTS

( ~ Fic. 14. Distribution of fossil remains of
us ‘ Clethrionomys glaveolus (Schreber) in the
S British Isles.
ww e% >
aa,
i i : ?
= 8 orion
‘ A one
wleate gh

Clethrionomys glareolus is an element of the forest fauna. It was probably not
present in this country during the maxima of the glaciations, though it was able
to reach it very early after the retreat of the ice and invaded the British Isles many
times in all the interglacials and in post-glacial time.

GENERAL DISTRIBUTION. C. glareolus is now distributed throughout Europe,
northern and central Asia and probably North America. Remains of the genus
Clethrionomys are common in faunas of early middle Pleistocene age from continental
Europe and have been described under different specific names, though they probably
all belong to C. glareolus. C. glareolus is abundant in the interglacial and inter-
stadial deposits of the late Pleistocene of Europe.

SYSTEMATIC REMARKS. Only a few remains of Clethrionomys are known from the
Norfolk Forest Bed. Although they were determined by Hinton (1926b) as ‘“Evo-
tomys sp.’, they cannot be distinguished from Recent C. glareolus. The bank vole
is a primitive, and very conservative, species of vole and its existence in Europe in
the early Pleistocene is not unexpected.

Hinton (1926b) described from the Ightham Fissures two new species of Clethrio-
nomys, ‘Evotomys kennardv’ and ‘E. harrisont’, both from the same faunal assemblage.
The material from Ightham is however a mixture of late Glacial and Recent bones
and the holotype of E. harrison is a well-preserved skull of a typical C. glareolus.
It is the skull of a young individual and its supposed characters (small dimensions,
broad inter-orbital region) are related to its juvenile condition; it is probably a
Recent specimen. ‘E. kennardi’ seems to be of earlier geological age and its dimen-
sions are somewhat larger than those of Recent specimens. According to Hinton
(1926b), it belongs to the ‘FE. nagert’ group. C. g. nageri is now generally recognized
as an Alpine subspecies (or perhaps only an ecotype) of C. glareolus, living under a
more severe climate. It seems probable that during the difficult conditions of the
Last Glaciation individuals of Clethrionomys grew bigger than during interglacial
and postglacial times, but there are no reasons to determine them as a new species.

OF BRITISH ISLES 95
Genus PLIOMYS Méhely 1914
Pliomys episcopalis Méhely 1914
Vole, extinct

1914 ** Pliomys episcopalis Méhely : 195-203, pls 4-5.
1974 Pliomys episcopalis Méhely ; Bishop: 300, 314.

LocaLity : Westbury-sub-Mendip Fissure, Somerset (13) : Bishop 1974, 1975.

CASS
ah
ae Fic. 15. Distribution of remains of Pliomys
a *e D . episcopalis Méhely in the British Isles.
ac
ef Gos
wee!)
PONE ra

DISTRIBUTION IN THE BriTIsH IsLES. Pliomys is known from only one British
locality, Westbury-sub-Mendip, where it occurs in association with a lower middle
Pleistocene fauna.

GENERAL DISTRIBUTION. P. episcopalis (type locality Betfia in Romania) appeared
in the late Villafranchian of Europe and became widespread in the lower middle
Pleistocene. In Italy it probably survived until the Holsteinian. Its remains are
distributed from the European part of the Soviet Union to France and Italy. This
species is generally recognized as an element of the steppe fauna.

Genus MIMOMYS Major 1902
Mimomys pliocaenicus Major 1902
Vole, extinct
1874 Ayvicola amphibius (Linn.) ; Blackmore & Alston : 462-464 (partim).
1882a Arvicola (Evotomys) intermedius Newton : 83, pl. 13 (partim).
1902 Mimomys pliocaenicus Major : 102-107, figs 13-15.

DISTRIBUTION IN THE BriTISH ISLES. Although one rolled microtine tooth from
the Red Crag has provisionally been referred to Mimomys by Spencer (1964) no
remains identifiable at specific level are known from Britain earlier than the Norwich
Crag. M. pliocaenicus has been found in the Norwich Crag and Weybourne Crag

8

96 PLEISTOCENE RODENTS

| aie
nag + We paar : f
as Q ( Fig. 16. Distribution “of “remamis” wf
a oa A 2 Mimomys pliocaenicus Major in the British
\
Ya S eG Isles.

of Covehithe, Easton Bavents and Sizewell, Suffolk, and Thorpe! and Bramerton,
Norfolk (51, 50, 49, 48, 55). Its remains are also known from deposits of uncertain
age at Kyson, Suffolk (47), and from the Weybourne Crag and probably Pastonian
Forest Bed deposits at East Runton, Norfolk (64). All the specimens from this
last locality seen by us in the British Museum (Natural History) are rounded, with
visible traces of transport by water.

BIBLIOGRAPHY. Blackmore & Alston 1874, Newton 1882a, Major 1902, Hinton
1g10b, Barrett-Hamilton & Hinton 1g10-21, Hinton 1926b, Cranbrook 1955Aa, b,
Carreck 1966, West 1968. Specimens have been seen in the BM(NH) and Ipswich
Museum.

GENERAL DISTRIBUTION. M. pliocaenicus was first described by C. I. F. Major
from the Val d’Arno in Italy. It is known from Late Villafranchian (Tiglian)
deposits in France, Holland, Belgium, Germany, Poland, Hungary, Czechoslovakia,
Romania and the Soviet Union (including Siberia). It has not been found in
Cromerian (s.1.) faunal assemblages outside Britain, where it is of Pastonian age.

SYSTEMATIC REMARKS. WM. fliocaenicus is a characteristic species. It developed
from smaller and more primitive ancestors living in the early Villafranchian (M.
polonicus Kowalski) and disappeared or rather evolved into other species at the end
of the Tiglian Interglacial.

1 There are several places in East Anglia known as Thorpe and the location of the M. pliocaenicus
site is not proved beyond doubt. The two most likely localities are Thorpe near Norwich, Norfolk, and
Thorpe near Aldeburgh, Suffolk. C. I. F. Major (1902) described specimens collected by Mr Fitch from
Thorpe, preserved in the Norwich Museum, though he did not indicate the county in which the Thorpe
in question is situated. Carreck (1966) assumed that it was the Thorpe near Aldeburgh. Mr Fitch is
known to have collected in the Norwich Crag of the Norwich area, however, and as the specimens sub-
sequently found their way to Norwich Museum the possibility that they came from the Thorpe near
Norwich must also be considered. Mr H. E. P. Spencer (2m litt.) favours the last-mentioned locality,
since there is a well-documented Crag mammal site there, whereas he points out that Thorpeness in
Suffolk is on Coralline Crag. Mr P. Cambridge informs us of a Norwich Crag pit at Shell Cottages,
Thorpe, Aldringham, near Aldeburgh, Suffolk. He too thinks the Thorpe near Norwich, where there were
formerly several Norwich Crag pits with mammalian remains, is the most likely locality for the remains
of M. pliocaenicus described by Major. We have tentatively accepted the Norfolk alternative, and this
is marked on Fig. 1 as locality 48.

OF BRITISH ISLES 97

Mimomys reidi Hinton 1910
Vole, extinct

1882a Arvicola (Evotomys) intermedius Newton : 83 (partim).
I910b Mimomys veidi Hinton : 491.

DISTRIBUTION IN THE BRITISH IsLEs. This species is known in Britain from the
Weybourne Crag at *Trimingham (61) and from the Norwich Crag at Sizewell (49).
In the collections of the British Museum (Natural History) there are specimens
from the Savin collection labelled “West Runton, Upper Freshwater Bed’, but they
are probably mislabelled or were found in a secondary layer. Hinton (1926b)
stated that this species was not found outside the Norwich and Weybourne Crags.

BIBLIOGRAPHY. Newton 1882a, Barrett-Hamilton & Hinton 1g1o—21, Hinton
1g10b, 1926b. Specimens have been seen at the BM(NH) and Ipswich Museum.

GENERAL DISTRIBUTION. M. reidi (probably identical with M. petenys Méhely
1914) is known from many localities of Tiglian age in Holland, France, Germany,
Italy, Poland, Czechoslovakia, Hungary and the Soviet Union (including Siberia).

SYSTEMATIC REMARKS. The holotype of this species was described from the
Weybourne Crag at Trimingham, Norfolk, but the systematic position of the remains
determined as M. reid: from continental Europe presents a problem: different
forms, probably different stages of the same phyletic line, appear to be represented.

& Fic.17. Distributionof remainsof Mimomys
veidt Hinton (v7) and M. newtoni Major (@)
in the British Isles.

Mimomys newtoni Major 1902
Vole, extinct
1902 Mimomys newtont Major : 102-107, figs 13-15.
DISTRIBUTION IN THE BriTIsH IsLEs. In Britain M. newtont is known from

deposits ranging in age from Norwich Crag to either Weybourne Crag or Forest
Bed. The holotype was described by Major (1902: text-fig. 13, no. 7) as having

98 PLEISTOCENE RODENTS

been found in the ‘East Runton Forest Bed’. It is now known that the deposits
at East Runton are Pastonian and the holotype is most likely to be of this age.
M. newtoni is also known from specimens in the British Museum (Natural History)
collected by the Earl of Cranbrook from the Norwich Crag of Easton Bavents,
Suffolk (50). The specimen in the BM(NH) from the Norwich Crag at Bramerton,
Norfolk, (Hinton 1926b : 376, footnote) ‘which may be referred to M. newton’ is
clearly different from the holotype and probably does not belong to the same
species.

BIBLIOGRAPHY. Major 1902, Hinton 1910b, Barrett-Hamilton & Hinton 1910-21,
Hinton 1926b.

GENERAL DISTRIBUTION. M. newton is known outside Britain from deposits of
Tiglian age in Holland, France, Germany, Poland, Czechoslovakia, Hungary and
the Soviet Union.

SYSTEMATIC REMARKS. Small representatives of the genus Mimomys Major are
so rare in the Lower Pleistocene strata of East Anglia that it is difficult to decide
how many species are represented. Some of the specimens of ‘M. newtoni’ from
East Runton (M6967, numbers 19, 20 and 23) are visibly larger than the holotype
(M, 2-9 and 3-0 mm long, compared with 2-5 mm) and have cement in the re-entrant
angles, absent in the type specimen. There may be two ‘chronospecies’ of different
geological ages. The continental specimens (e.g. from Kadzielnia, Poland) are
identical with these larger specimens of ‘M. newton’ and different from the holotype.
Further research will be necessary to determine whether an additional species should
be established.

Mimomys savini Hinton 1910
Vole, extinct

1874 Ayrvicola amphibius (Linn.) ; Blackmore & Alston : 462-464 (partim).
1881 Arvicola (Evotomys, Coues) intermedia Newton : 258.

1902 Mimomys intermedius (Newt.) ; Major : 102-107.

191I0b) =9©Mimomys savini Hinton : 491.

I910b Mimomys majori Hinton : 491.

1958 Mimomys millert WKretzoi : 55.

DISTRIBUTION IN THE BriTIsH IsLEs. M. savini is a common species in the
Upper Freshwater Bed at West Runton (65) and it is also known from East Runton
(64). According to Newton (1882a), it has also been found at Cromer (63), Geldeston
(53) and Kessingland (52). Mzmomys cf. savini was recorded by Stuart (1974) from
Cromerian deposits at Sugworth, near Oxford (21).

BIBLIOGRAPHY. Blackmore & Alston 1874, Newton 1881, 1882a, 1891, Reid
1890, Major 1902, Hinton 1910b, Barrett-Hamilton & Hinton 1910-21, Hinton
1920, 1926b, Kretzoi 1958, 1965, Pasquier 1972.

GENERAL DISTRIBUTION. Mimomys savini is widely distributed at localities of
‘Cromerian’ age in Holland, France, Germany, Italy, Poland, Czechoslovakia,
Yugoslavia and the Soviet Union (including Siberia).

OF BRITISH ISLES 99

Fic.18. Distributionof remainsof Mimomys
stig 7 savini Hinton in the British Isles.
Q

SYSTEMATIC REMARKS. MM. savini was described by Newton (1881) from the
Upper Freshwater Bed of *West Runton under the name of Avvicola intermedia.
Hinton (1g10b) described two further species of the same genus, WM. majort and M.
savint, from the same deposit. According to him, they differed from M. intermedius
only in slight differences in the pattern of M,, the dimensions of all the molars and
the patterns of all other teeth being identical. Kretzoi (1958) found that A. inter-
media is a junior homonym and must therefore be replaced by the name Mimomys
milleyt Kretzoi; in a later paper (1965) he showed that the three species M. inter-
medius, M. savint and M. major really represent only one variable species. The
name of it must therefore be Mimomys savini Hinton, which is the oldest valid name.
In fact all possible intermediate forms between typical M. intermedius, M. savini
and M. majori are present in the abundant material from the Upper Freshwater
Bed (Pasquier 1972). The presence of three species of one genus with identical
dimensions in one and the same layer is also scarcely imaginable from the ecological
point of view. Specimens determined as ‘M. majom’ and ‘M. savini’ have also been
found in some Czechoslovak and German fossil localities, always associated with
M. intermedius.

Genus ARVICOLA Lacépéde 1799

The systematic study of even the living representatives of this genus presents
many difficulties to zoologists. Miller (1912) divided living populations of water
voles from western Europe into seven species but later investigators (for bibliography
see Reichstein 1963) have demonstrated that nearly all the characters which were
used for the diagnosis of these species lie within the limits of individual variation.
Variations occur principally in the dimensions and in some proportions of the skull,
proportions which may change allometrically with the changes of absolute dimensions.

Cytological studies support the view that there are only two living species in
this genus, A. sapidus Miller 1908, from the south of France and the Iberian Penin-
sula, and A. terrestris (Linnaeus 1758) distributed throughout the rest of Europe

100 PLEISTOCENE RODENTS

as well as the northern part of Asia and the Middle East. The Recent population
in Britain belongs to this last species. Reichstein (1963), after a very thorough
analysis, found distinct but small craniological differences between the two species
of Arvicola and further differences were reported by Corbet e al. (1970).

The systematic study of fossil forms, usually possible only from teeth and frag-
mentary skulls, is still more difficult. Remains from glacial deposits are usually
larger than those from the interglacials and interstadials or from the Holocene, and
most of the characters which have been used for creating new fossil species of water
voles are expressions of individual variability, present also in Recent species. It
seems beyond doubt that Avvicola developed from a late form of the genus Mimomys,
most likely Mimomys savini, through the progressive reduction of the formation of
roots in the molars and acquisition of the continual growth of these teeth. One
character, typical of Mimomys, is nevertheless retained in geologically older popu-
lations of Avvicola: the enamel of the cheek-teeth is thinner on the concave and
thicker on the convex sides of the salient angles, the reverse being true in the late
Pleistocene and Recent populations. The two forms are united by a full series of
intermediate specimens. O. Fejfar (in Koenigswald 1970) re-examined the holotype
of Mimomys cantianus Hinton from the Hoxnian deposits of Ingress Vale, Swans-
combe, and found that there is no trace of root-formation. This form must therefore
be included in the genus Avvicola. Arvicola cantiana (Hinton) is the oldest available
name for the primitive representatives of Avvicola and later names such as A.
bactonensis, A. greent and A. praeceptor must be treated as synonyms.

Koenigswald (1972, 1973) also studied the stratigraphical range of the various
stages of the Mimomys—Arvicola lineage from European localities and defined a
series of faunal groups :

(a) A Mimomys savint fauna, including the Upper Freshwater Bed of West Runton.

(b) and (c) Avvicola faunas groups I and 2, both with A. cantiana (Koenigswald
included Swanscombe in group 2).

(d) Avvicola fauna group 3, with A. ¢erresiris.

Arvicola

cantiana cantiaqna -
terrestris terrestris

Fic.19. Development of the enamel layer in Middle European forms of Avvicola. Arvicola
cantiana (Elster—Riss); Avvicola cantiana-tervestyis transition form (Riss—Eem) ;
Arvicola terrestris (Eem—Holocene). From Koenigswald (1973 : 164).

OF BRITISH ISLES IOI

v

KONE PRUSY

v

UPPER FRESHWATER BED
PRE ZLETICE

VOIGTSTEDT
SUSSENBORN
ERPFINGEN 1 und 3
HUNDSHEIM

SUDMER BERG 2
TARKO

MOSBACH mittl.S.
UPPONY 1
PETERSBUCH
HEPPENLOCH
SWANSCOMBE
TORNEWTON glutton str
ERKENBRECHTSWEILER
BIEDERMANN'SCHER STBR.

MAUER
eer
HUNAS

Sorex runtonensis

Sorex (Orepanosorex) savini
Pliomys lenki
episcopalis
Allocricetus bursae
Paropodemus coron.

Talpa minor

europaea
Mimomys savini
Arvicola cantiona
cont.- terr.

terrestris

Mimomys savini fauna Arvicola fauna, group 1 Arvicola fauna,group 2

gp. 3

Fic. 20. Chronological distribution of Avvicola and other stratigraphically important
small mammal species in European Pleistocene faunas. From Koenigswald (1973 : 161).

He referred the Glutton Stratum of Tornewton Cave to an intermediate stage
between groups 2 and 3 (Fig. 20).

Koenigswald made a further study of British Arvicola remains in 1974 and (pers.
comm.) assigned the phylogenetic position, between A. cantiana and A. terrestris,
of the finds from various sites as follows :

A. cantiana ————————> intermediate forms ———————_———_-> A. terrestris
Clacton Aveley Barrington Ightham
Swanscombe Crayford Joint Mitnor Cave Kent’s Cavern
(Barnfield Pit and (Cave Earth)
Ingress Vale) Grays Swanton Morley
Wye Cave
Ostend Stutton Tornewton Cave (= Merlin’s Cave)
(Otter Stratum)
Westbury-sub-Mendip Tornewton Cave
Tornewton Cave (Reindeer Stratum)

(Glutton Stratum)

These findings are of very far-reaching stratigraphical importance. <A. cantiana
appears for the first time in Britain at the ‘late Cromerian’ site of Westbury-sub-
Mendip, which appears to represent Koenigswald’s group 1, and is also present
in the Hoxnian deposits of Clacton and Swanscombe. The transition form is
represented by remains from the cold stage represented by the Glutton Stratum of

102 PLEISTOCENE RODENTS

Tornewton Cave, in the Otter Stratum of that cave (where it is apparently slightly
more advanced) and in the Last Interglacial hippopotamus faunas of Barrington,
Joint Mitnor Cave and Swanton Morley. A. terrestris is characteristic of Last
Glaciation and Holocene faunas.

It must be noted that Hinton’s and Koenigswald’s divisions between A. cantiana
and A. terrestyis do not exactly coincide. Koenigswald, from a re-examination of
the specimens from the Barrington hippopotamus site, considered that the Avvicola
remains from this locality should be attributed to the A. cantiana-terresiris transition
form rather than to A. praeceptor (= A. cantiana), as determined by Hinton (1926b :
393). Stuart (see p. 69) has suggested that both forms are present. The specimens
from the Last Interglacial site of Hessle, Yorkshire, attributed by Hinton (in Bisat
1940) to A. praeceptor, but which Boylan (1967) considered to lie within the normal
range of variability of A. terrestris, have not been re-examined during the present
investigation.

Surprisingly the remains from Aveley, Grays Thurrock, Harkstead and Stutton
(currently widely regarded as Last Interglacial, though most if not all are sites without
hippopotamus — the undoubted hippopotamus remains from Grays were collected
more than a century ago and their exact provenance within the terrace complex of
that area is unknown) are advanced A. cantiana and appear to be earlier than the
pre-hippopotamus Glutton Stratum of Tornewton Cave.

The Crayford record of A. cantiana must be accepted with some reserve. Although
Arvicola is recorded from this locality by Dawkins & Sanford (1866) and by Whitaker
(1889) it was not mentioned by Hinton (1g00a, b, 1910b, 1926b). The present
determination is based on a single specimen in the British Museum (Natural History).

Arvicola cantiana (Hinton 1910)
Water vole, extinct

1846 Ayvvicola amphibia ; Owen : 201-205 (partim).

1882a Ayrvicola amphibius ? Linnaeus ; Newton: 87-88.
1902 Microtus intermedius ; Hinton & White : 414-415.
I9g10ob Mimomys cantianus Hinton : 491.

1926b Ayrvicola bactonensis Hinton : 386—389.

1926b Arvicola greent Hinton : 389—391, fig. 106, 1-2.
1926b Arvicola praeceptor Hinton : 391-394, fig. 106, 3-17.
1970 Arvicola cantiana (Hinton) ; Koenigswald : 418—420.

Localities: Ostend, Norfolk (57) : Owen 1846, Newton 1882a, Lydekker 1885,
Hinton 1926b, BM(NH).

Westbury-sub-Mendip, Somerset (13) : Bishop 1974, 1975.

Clacton-on-Sea, Essex (43) : Hinton in Warren 1955, Sutcliffe 1964, BM(NH).

*Swanscombe (Ingress Vale and Barnfield Pit, Lower Loam), Kent (30) : Hinton
& White 1902, Stopes 1904, Hinton rg1ob, 1926b, Barrett-Hamilton & Hinton
1910-21, Sutcliffe 1964, Koenigswald 1970, BM(NH).

Aveley, Essex (34) : BM(NH).

OF BRIFISH ISLES 103

Grays Thurrock, Essex (33) : Hinton, Kennard & Newton 1900, Hinton rgrob,
1926b, Barrett-Hamilton & Hinton 1910-21, BM(NH).

Ilford, Essex (35) : Dawkins & Sanford 1866, Whitaker 1889, Newton 18g0a,
Johnson & White 1900, Hinton 1g00a, b, 1g10b, 1926b, BM(NH).

Stutton, Suffolk (44) : Stuart collection and seen in Ipswich Museum.

Harkstead, Suffolk (44) : Ipswich Museum — Stuart (77 litt.).

Northfleet, Kent (29) : det. A. Stuart.

Crayford, Kent (27) : Dawkins & Sanford 1866, Whitaker 1889, BM(NH).

Arvicola cantiana-terrestris transition form
(see pp. 99-102 above)

Loca.iTiEes: Tornewton Cave (Glutton Stratum and Otter Stratum), Devon (3) :
Kowalski 1967, (BM(NH).

Barrington, Cambridgeshire (40) : Hinton 1926b (but see note on p. 69), BM(NH).

Joint Mitnor Cave (layer IV), Devon (1) : BM(NH).

Minchin Hole, Glamorganshire (16) : det. A. J. Stuart.

Stoke Tunnel Beds, Suffolk (46) : Ipswich Museum.

Swanton Morley, Norfolk (66) : Stuart, pers. comm.

ae 3 a Fic. 21. Distribution of fossil remains of
ge Sage. Ayvicola cantiana (Hinton) (f§) and 4A.
ia ae tervestyvis (Linn.) (J) in the British Isles.
we a (Transition form §J, not studied @.)
ed: ?
o Te!
_
(5 e ®ome
ZN AS
Ca

Arvicola terrestris (Linnaeus 1758)

Water vole
1846 Arvicola amphibia ; Owen: 201-205 (partim).
1847 Hypudaeus spelaeus Cuv.; Giebel : 88.
1872 Arvicola amphibius Desm. ; Dawkins & Sanford : 180.

1g10a Ayvicola abbotti Hinton : 34-35.
1960 Ayvicola amphibius terrestvis Bramwell : [10].

104 PLEISTOCENE RODENTS

LocaLiTies: Ightham Fissures, Kent (28): Newton 1894, Hinton r1g1O0a, b,
Barrett-Hamilton & Hinton 1910-21, Hinton 1926b, Zimmermann 1959, BM(NH).

Kent’s Cavern (Cave Earth), Devon (6) : Owen 1846, Dawkins & Sanford 1866,
Blackmore & Alston 1874, Lydekker 1885, Hinton 1926b, Kennard 1945-46,
Carreck 1957, BM(NH).

Tornewton Cave (Reindeer Stratum), Devon (3) : Kowalski 1967.

Wye (= Merlin’s) Cave, Herefordshire (18) : Bate 1901, Hinton 1926b, BM(NH).

Arvicola cantiana | A. terrestris complex
(phylogenetic position not studied)

With the exception of Hoxne (Hoxnian) and Bacon Hole, Hessle, Kirkdale Cave
and the Hyaena Stratum of Tornewton Cave (which are Last Interglacial) the
following sites are all believed to be of Last Glaciation or Holocene age. The
Arvicola remains are probably mostly advanced forms.

Localities : Hoxne, Suffolk (54) : Wolff (vm Jitt.).

Bacon Hole (16) : Stuart (7 Jitt.).

Bobbitshole, Suffolk (46) : Ipswich Museum, det. J. N. Carreck.

Hessle, Yorkshire (74) : Lamplugh 1891, Bisat 1940, Catt & Penny 1966, Boylan
1967, specimen in Institute of Geological Sciences Museum, London, not seen
by K..K.

Kirkdale Cave, Yorkshire (77): Owen 1846, Giebel 1847, Dawkins & Sanford
1866, Lydekker 1885, Hinton 1926b, BM(NH).

Tornewton Cave (Hyaena Stratum), Devon (3) : Kowalski 1967.

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924, 1926b, BM(NH).

Brixham Cave, Devon (4) : Hinton 1926b, Carreck 1957, BM(NH).

Chudleigh Fissure, Devon (7) : BM(NH).

Dog Holes Cave, Lancashire (79) : Jackson 1910, 1912, Hinton 1926b, BM(NH).

Elder Bush Cave, Staffordshire (69) : Bramwell 1964.

Gough’s Cave, Somerset (12) : BM(NH).

Langwith Cave, Derbyshire (72) : Mullins 1913, BM(NH).

Levaton Cave (2) : Carreck 1957.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Happaway Cave, Devon (5) : Hinton 1926b, Carreck 1957, BM(NH).

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Cowside Cave No. 3, Yorkshire (78) : BM(NH).

Thatcham, Berkshire (Holocene, 23) : Wymer 1962.

GENERAL DISTRIBUTION of Avvicola, and systematic remarks. As stated above,
the forms described as A. greeni, A. bactonensis and A. praeceptor must be treated
as synonyms of A. cantiana. The first two of the above-mentioned species were
described by Hinton (1926b) from specimens collected at Ostend, near Bacton,
Norfolk, during the first part of the nineteenth century. A. bactonensis is known

Na IIIT USisl IESIIE'S) 105

only from the holotype and A. greeni is represented only by two isolated teeth.
The stratigraphic position of these remains is uncertain. Hinton (1926b) suggested
that they were contemporary with the Cromer Forest Bed or only slightly younger.
As Arvicola is not represented at the Cromerian type locality of West Runton, where
the ancestral form Mimomys savini is present, it must be supposed that these
specimens are from later deposits. They may be equivalent in age to the faunal
assemblage of Westbury-sub-Mendip, where A. cantiana is present.

A. cantiana is known, under various names, from numerous localities of Holsteinian
(Hoxnian) age in continental Europe and Siberia. Remains of A. terrestris, to
which species it gave rise, are abundant from the Upper Pleistocene of Europe,
including England and Wales. Like other voles, Avvicola never reached Ireland.

A. terrestris is now widely distributed in nearly all Europe and in cold and tem-
perate parts of Asia, as far south as Persia and Palestine. It lives mostly near water,
but under diverse climatic and vegetational conditions, ranging from the taiga belt,
through deciduous forest and steppe, to the deserts in the south. As a species of
high ecological tolerance it was present during the Pleistocene under both glacial
and interglacial conditions.

Genus PITYMYS McMurtrie 1831

Many zoologists are of the opinion that Pitymys is only a subgenus of Mzcrotus.
Pitymys is morphologically similar to Microtus, sensu stricto, and is rather similar in
its ecology, but the living forms are biologically quite different. Chaline (1972),
who studied fossil vole remains from the French Pleistocene, came to the conclusion
that fossil forms of Pitymys are nothing but morphotypes of Allophaiomys plio-
caenicus Kormos 1932. If this is correct, then A. pliocaenicus would be a junior
synonym of Pitymys arvaloides and its name must be changed accordingly. In his
opinion the Recent subgenus Pitymys is polyphyletic and its members are more
closely related to different forms of Microtus than to one another. In the opinion
of one of us (K. K.) Chaline underestimates the ecological similarity of Recent
species of Pitymys and all other zoological evidence, basing his conclusions on the
pattern of M, only. For this reason he is unable to follow Chaline’s suggestions.

Voles of the Pitymys group dominated the European vole faunule during the
Cromerian Interglacial. They were still common during the Hoxnian, after which
they disappeared completely from Britain. Their distribution in Europe and
western Asia is now rather discontinuous. The first evidence of the presence of
Pitymys in Britain was published by Major in 1902. Hinton (1923a) described two
different species from the Upper Freshwater Bed of West Runton, P. gregaloides and
P. arvaloides. They are clearly distinct species, though in continental Europe
intermediate specimens are present. If Pztymys is only a subgenus of Microtus, a
very prevalent opinion among Recent mammalogists, the name Pitymys arvaloides
Hinton is a taxonomic homonym and must be replaced by a new name, Microtus
(Pitymys) arvalidens Kretzoi 1958. In the opinion of K. K. it is more convenient

106 PLEISTOCENE RODENTS

to treat Pitymys as a separate genus and the name given by Hinton is therefore
retained here.

Pitymys arvaloides Hinton 1923
Pine vole, extinct

1882a Avvicola arvalis Pallas; Newton : 88-80, pl. xiv, figs 2-5.
1902 Microtus (Pitymys) sp.; Major: 107, fig. 15 (28).
1923a =©Pitymys arvaloides Hinton : 541-542.

1958 Microtius (Pitymys) arvalidens Kretzoi: 57 (new name for Pitymys arvaloides Hinton
1923, non Arvicola arvaloides Pomel).
1972 Allophaiomys pliocaenicus Kormos ; Chaline: 104.

Localities: West Runton (Upper Freshwater Bed), Norfolk (65) : Newton
1882a, Major 1902, Hinton 1923a, 1926b, Kretzoi 1958, Chaline 1972, BM(NH).

Hitchin, Hertfordshire (39) : Carreck 1959 : 326 (det. as P. cf. arvaloides).

Swanscombe (Barnfield Pit, Lower Gravel and Lower Loam), Kent (30) : Carreck
1959 (: 326), Sutcliffe 1964.

Fic. 22. Distribution of remains of Pitymys
avvaloides Hinton in the British Isles.

DISTRIBUTION IN THE BriTISH ISLES. P. arvaloides is abundant in the Cromerian
Upper Freshwater Bed at West Runton. It is present also in deposits of the
lower part of the Hoxnian Interglacial, but disappears completely before its end.
There is at present no evidence of its presence in Britain during the Anglian cold
phase and it is probable that it invaded this country from the European continent
twice.

GENERAL DISTRIBUTION AND SYSTEMATIC REMARKS. P. arvaloides is present at
numerous localities of ‘Cromerian’ age on the continent (Italy, Yugoslavia, Germany,
Hungary, Poland, Czechoslovakia, Austria, the European part of the Soviet Union
and Georgia). Nearly everywhere it has been found together with P. gregalozdes.
It is not impossible that the two species, different in the morphology of M, but of
identical dimensions, are only two forms of one polytypic species.

OF BRITISH ISLES 107

Pitymys gregaloides Hinton 1923
Pine vole, extinct

1882a Avvicola gregalis Pallas ; Newton : 90-91, pl. xiv, figs 6, 6a.
1891 Microtus (Arvicola) gregalis Pallas; Newton: 53.
1923a ©3=- Pitymys gregaloides Hinton : 541-542.
1972 Allophaiomys pliocaenicus Kormos ; Chaline: 104.
Loca.ities: West Runton (Upper Freshwater Bed), Norfolk (65) : Newton
1882a, 1891, Hinton 1923a, 1926b, BM(NH).
Westbury-sub-Mendip Fissure, Somerset (13) : Bishop 1974, 1975.
Kent’s Cavern, Devon (6) : Hinton 1926b, Kennard 1945-46, BM(NH).

fe i Fic. 23. Distribution of remains of Pitymys
gvegaloides Hinton in the British Isles.

DISTRIBUTION IN THE BriTIsH ISLES. P. gregaloides is present in Cromerian
deposits at West Runton and at the Westbury Fissure, together with P. arvaloides.
Two specimens, one preserved in the British Museum (Natural History), the other
in Bristol Museum, are known from Kent’s Cavern. The stratigraphical position of
these specimens is unknown, but the species has not yet been recorded from any
British Hoxnian locality and it is likely that they are of pre-Hoxnian age.

GENERAL DISTRIBUTION AND SYSTEMATIC REMARKS. P. gvregaloides is present,
together with P. arvaloides, at numerous localities in central and southern Europe
dated as Cromerian (sensu lato).

Genus MICROTUS Schrank 1798

The genus Microtus now holds the dominant position in the subfamily Microtinae,
being both most highly differentiated and most individually numerous. In the
fossil state molar teeth make up most of the finds, mandibles occurring less commonly
and complete skulls only in exceptional cases. Furthermore, it should not be
forgotten that in living populations the variability of the molar pattern is sometimes

108 PLEISTOCENE RODENTS

enormous (Kowalski 1970) and this variability may be discontinuous, representing
genetical polymorphism. At the same time there are species, clearly differentiated
in external characters and in biological requirements, in which the teeth are nearly
identical (e.g. the lower teeth of M. agrestis and M. arvalis cannot be clearly dis-
tinguished, nor can the upper teeth of M. arvalis and M. oeconomus). All these
factors make the correct determination of fossil remains, including the material
from Britain described below, very difficult.

The genus Microtus appeared rather late in the European Pleistocene, being
found first in the Cromerian Interglacial when Pitymys was still numerically domi-
nant. Inthe Hoxnian Interglacial both genera lived together but in later sediments
in Britain Pitymys disappeared completely and on the European continent it now
plays only a subordinate role.

Microtus arvalinus Hinton 1923
Vole, extinct

1902 Microtus sp., recalling M. arvalis ; Major : 107, fig. 15 (27).
1923a Microtus arvalinus Hinton : 541-542.
1972 Microtus (Microtus) avvalis (Pallas) ; Chaline : 106.

LocaLiTies : West Runton (Upper Freshwater Bed), Norfolk (65) : Major 1902,
Hinton 1923a, 1926b, BM(NH).

Westbury-sub-Mendip, Somerset (13) : Bishop (pers. comm.).

Swanscombe (Lower Loam and Upper Middle Gravel), Kent (30) : Schreuder
1950, Sutcliffe 1964, BM(NH).

WP \ Fic. 24. Distribution of remains of Microtus
s! ao ay avvalinus Hinton in the British Isles.

GENERAL DISTRIBUTION. M. arvalinus was present in Britain during the
Cromerian and Hoxnian interglacials and the intervening Westbury stage. A small
form of Microtus, with the M, pattern of the M. arvalis/agrestis group, is present in

ORV RI RiSi Sie S 109

the Cromerian deposits of France, Germany, Austria, Czechoslovakia, Hungary,
Romania and Yugoslavia.

SYSTEMATIC REMARKS. WM. arvalinus is identical in its M, pattern with Recent
representatives of the M. arvalis/agrestis group, and differs from these only in its
slightly smaller dimensions. It is probable that WM. arvalinus is the ancestral form
of M. agrestis. According to Chaline (1972), M. arvalinus is a synonym of M.
arvalis, but there does not seem to be any good reason for putting it into this
synonymy rather than that of M. agrestis.

Microtus agrestis (Linnaeus 1761)

Field vole
1846 Ayvicola agrestis ; Owen : 206—207.
1847 Hypudaeus bucklandii Giebel : 88.
1894 Microtus (= Arvicola) agrestis ; Newton: 197.

1910b Microtus agrestis neglectus ; Hinton : 494.
1910b Microtus agrestoides Hinton : 493.
Microtus arvalis and Microtus avvalis/agrestis group (partim) ; many authors.

GENERAL NOTES ON THE Microtus arvalis/agrestis GROUP. During the nineteenth
century most zoologists did not appreciate the differences between WM. arvalis and
M. agrestis and these two names must be regarded as synonyms in papers published
at that time. Later on, the presence of the M. arvalis group on the Orkney Islands
(M. arvalis does not survive today on the mainland of Britain and M. agrestis does
not occur on Orkney) led to the hypothesis that M. arvalis was the first Microtus
to reach Britain in the postglacial period. Under the influence of this hypothesis
fossil remains of the M. arvalis/agrestis group were usually determined as ‘M.
arvalis’. It must nevertheless be born in mind that only the upper M? provides
diagnostic characters for distinguishing M. agrestis and M. arvalis. Both lower
jaws and other isolated teeth are useless for this purpose (Dienske 1969). An
example of the confusion which can occur in this context is provided by the Mzcrotus
remains from Dog Holes Cave, Lancashire. All the upper dentitions were correctly
determined as ‘M. agrestis’, but all the lower jaws were labelled ‘M. cornera’ (= M.
arvalis). This illustrates well the tendency to decide in each doubtful case in favour
of M. arvalis.

In the site list for M. agrestis given below all the localities of M. agrestis, ‘M.
agrestoides’, ‘M. arvalis/agrestis group’ and ‘M. arvalis’ are enumerated.

Localities: Clacton-on-Sea, Essex (43) : Hinton 1923b, Singer et al. 1973.

Swanscombe (Upper Middle Gravels), Kent (30) : Schreuder 1950 (M. arvalis/
agrestis group).

Ilford (Uphall Estate), Essex (35) : Hinton 1900a, b, Johnson & White 1900.

Grays Thurrock, Essex (33) : Hinton, Kennard & Newton 1900, Hinton Igor,
1g10b, Barrett-Hamilton & Hinton rg10—21, Hinton 1924, 1926b, BM(NH) (‘M.
agrestotdes’).

110 PLEISTOCENE RODENTS

Harkstead, Suffolk (44) : Ipswich Museum.

Stutton, Suffolk (44) : Ipswich Museum.

Northfleet, Kent (29) : Burchell 1935.

Tornewton Cave (Glutton Stratum, Hyaena Stratum, Reindeer Stratum, Dilu-
vium), Devon (3) : Kowalski 1967, BM(NH).

Water Hall Farm Pit, Hertfordshire (38) : BM(NH).

Alveston Fissure, Gloucestershire (15) : det. G. B. Corbet.

Bacon Hole, Glamorganshire (16) : det. A. J. Stuart.

Bobbitshole, Suffolk (46) : Ipswich Museum, det. J. N. Carreck.

Barrington, Cambridgeshire (40) : BM(NH).

Joint Mitnor Cave, Devon (1) : BM(NH).

Kirkdale Cave, Yorkshire (77) : Owen 1846, Giebel 1847, Dawkins & Sanford
1866, Blackmore & Alston 1874, BM(NH).

Minchin Hole, Glamorganshire (16) : det. A. J. Stuart.

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924, BM(NH).

Bleadon Cave, Somerset (10) : BM(NH).

Brixham Cave, Devon (4) : BM(NH).

Chudleigh Fissure, Devon (7) : BM(NH).

Dog Holes Cave, Lancashire (79) : Jackson 1910, 1912, BM(NH).

Elder Bush Cave, Staffordshire (69) : Bramwell 1964.

Gough’s Cave, Somerset (12) : BM(NH).

Great Doward Cave, Herefordshire (17) : BM(NH).

Ightham Fissures, Kent (28) : Newton 1894, 1899b, Hinton rgr1ob, Barrett-Hamil-
ton & Hinton 1910-21, Hinton 1921, 1926b, Jackson 1947, Zimmermann 1959.

Kent’s Cavern, Devon (6) : Owen 1846, Dawkins & Sanford 1866, Blackmore &
Alston 1874, Lydekker 1885-87, Kennard 1944-45, BM(NH).

King Arthur’s Cave, Herefordshire (17) : BM(NH).

Langwith Cave, Derbyshire (72) : Mullins 1913, BM(NH).

Levaton Cave, Devon (2) : Carreck 1957.

Merlin’s Cave, Herefordshire (18) : Hinton 1925.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Brean Down, Somerset (9g) : Apsimon, Donovan & Taylor 1961.

Happaway Cave, Devon (5) : BM(NH).

Lynx Cave, Denbighshire (67) : Blore 1966.

Cowside Cave No. 3, Yorkshire (78) : BM(NH).

Corstorphine, nr. Edinburgh, Scotland (81) : Evans 1913.

Creag nan Uamh Cave, Inchnadamph, Scotland (83) : Peach & Horne 1917.

DISTRIBUTION IN THE BriTIsH IsLes. WM. agrestis is now widely distributed in
Great Britain and many of the remains cited as fossils are probably of relatively
recent age, in any case post-glacial.

The supposed earliest British occurrence of this group or species of rodents is at
the Hoxnian locality of Swanscombe, where it is recorded as having been found in
the Upper Middle Gravel in association with M. arvalinus (Schreuder 1950). The
occurrence together of these two forms, which, it has been suggested (p. 109),
are members of the same phylogenetic lineage, creates a problem. Schreuder

OF BRITISH ISLES east

Fic. 25. Distribution of fossil remains of
the Microtus arvalis/agrestis group in the
British Isles.

distinguished the two forms at Swanscombe on grounds of size and stated that she
would not have hesitated to accept as M. arvalinus the specimen determined by her
as ‘Microtus sp., arvalis/agrestis group’ had it not been so large. It has been pointed
out by Corbet, however (pers. comm.), that there is great variation, which may
reflect the age of the animal concerned, in the crown size of teeth of any microtine
in which the teeth continue to grow during life. In view of the very small size of
the sample studied by Schreuder, the occurrence of both M. arvalinus and M. arvalis/
agrestis together in the same deposit is accepted here with some reserve. The
remains are stratigraphically placed at about the time of transition between these
two forms and a larger collection of remains is needed for study.

M. agrestis was common during the warm stages represented by both Grays
Thurrock and Joint Mitnor Cave, and in the Glutton Stratum of Tornewton Cave,
but is absent from the faunas of Crayford and Clevedon Cave. It was common
during the Last Glaciation. At most localities where fossil remains have been found
only lower jaws were present, but in each case where upper teeth were also found the
second upper molar is of the typical M. agrestis pattern. There can be no doubt
that the field vole was discontinuously present in Britain from at least the time of the
Grays Thurrock deposits.

Hinton (1910b) created the name ‘M. agvestoides’ for the specimens of M. agrestis
from Grays Thurrock. He was surely influenced here by the supposition that Grays
Thurrock has a very ancient fauna. The only distinctive character of his new
species is ‘the constant development of a fourth outer angle in the last upper molar’.
This ‘fourth angle’ is present in a great proportion of specimens of Recent M.
agrestis and its presence is not sufficient for specific determination.

According to Jackson (1929b), a skull of a vole which he regarded as close to M.
arvalis was found in Kilgreany Cave, Co. Waterford, Ireland. This is the only
possible evidence that voles ever reached Ireland; it was a skull with all teeth
missing. It was found in the ‘Lower Stalagmite’ together with Dicrostonyx and
human remains; ‘...a close examination of the sockets suggests that the species

9

II2 PLEISTOCENE RODENTS

is the continental field vole Microtus arvalis or a close ally, and not the common field
vole (M. agrestis)’ (Jackson 1929b : 147).

The specimen has recently been re-examined by Savage (1966), who noted that
it is notably fresh and unaltered and who had no doubt that it should be referred to
M. agrestis. He pointed out that occasional vole remains have been found in north-
east Ireland in pellets dropped by passing owls. On the basis of existing evidence,
it is unlikely that voles ever reached Ireland.

GENERAL DISTRIBUTION. M. agrestis is now widely distributed in Europe, as
well as in the cold and temperate parts of Asia. In Europe it extends north to
include the whole Scandinavian peninsula and Finland. Fossil remains are known
in continental Europe beginning with the Riss (penultimate) Glaciation.

Microtus arvalis (Pallas 1779)
Common vole

19g10a Microtus cornevi Hinton : 35-36.
Microtus arvalis (partim) ; many authors.

DISTRIBUTION IN THE BritisH IsLtEs. Hinton (in Barrett-Hamilton & Hinton
Ig10—21, 2: 467) correctly stated that :

‘On several occasions fossils from various British deposits of Late Pliocene and Pleistocene age
have been determined as belonging to M. arvalis (e.g. from fissures near Bath, Somerset, by
Blackmore and Alston, Proc. Zool. Soc., 1874, 468) ; but in most cases such records imply,
because of the fragmentary nature of the material, on which they are based, nothing more
than the presence of a “‘vole’’ with an arvaloid, i.e. a normal dentition.’

According to Hinton (loc. cit.) the cranial remains from the Ightham fissures are
‘apparently identical’ with M. arvalis ; this locality probably marks the date of the
arrival of the species in Britain.

Hinton (1g10a) described from the Ightham fissures a new species, M. cornert,
very similar to the Orkney vole, M. orcadensis. We now know, however, that M.
orcadensis is nothing other than an isolated population of M. arvalis, possibly
developed within historic times.

M. cornert has been recorded from the following localities :

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924.

Brixham Cave, Devon (4) : BM(NH) (det. Hinton).

Dog Holes, Lancashire (79) : BM(NH) (det. Hinton).

Ightham Fissures, Kent (28): Hinton 1g10a,b, Barrett-Hamilton & Hinton
Ig10—21, Hinton 1921, 1926b, Jackson 1929, Hinton 1952, Zimmermann 1959.

Langwith Cave, Derbyshire (72) : Hinton 1910a, Mullins 1913.

Levaton Cave, Devon (2): Carreck 1957 (‘Muicrotus arvalis group, cf. corners
Hinton’).

The status of M. arvalis, including M. cornert, in the British Pleistocene requires
extensive further study. Although known today on the Orkney Islands, M.
avvalis does not occur on the British mainland. There is geological evidence that

OF BRITISH TSS 113

the Orkney Islands have never been connected to the mainland since the last glacial
phase, when ice covered the British Isles as far south as Norfolk and South Wales,
and the local population of M. arvalis [= orcadensis] must therefore be a recent
introduction from the continent (Corbet 1961). In Europe, M. arvalis is typical of
the zone of mild climate, being absent from the Scandinavian peninsula and from
northern Denmark. Its distribution as well as ecological data make its presence
in the British Isles in the early postglacial period extremely improbable. The
majority of fossil records of ‘M. cornert’ from Britain are of rather doubtful value.
As mentioned above, lower jaws of Mzicrotus (which are specifically undeterminable)
from Dog Holes Cave have been identified as M. corneri whilst all the skulls belong,
without exception, to M. agrestis. The skull from Brixham Cave determined as
belonging to M. cornert is probably a specimen of M. oeconomus.

One of the most potentially important collections for the study of M. arvalis is
that from Ightham Fissures, Kent, where many excellently preserved vole skulls
were collected. It is evident, however, that remains of both Pleistocene (Last
Glaciation) and Holocene age have been mixed together. In such circumstances it
is possible that sometimes a single species is represented in the fossil material by
two different subspecies or by two different populations of different dimensions.
Among the Muicvotus material from this locality, besides the typical skulls of MW.
gregalis and M. agrestis, there are two groups of larger and smaller skulls. Some of
them may belong to M. oeconomus, and the skull selected as the holotype of M.
cornert probably belongs here. Other specimens could be young individuals of this
species, or they may really belong to M. arvalis.

Recently two small further collections of rodents have been studied by Corbet
(pers. comm.) which suggest that M. arvalis may indeed have been present in Eng-
land during the Last Glaciation. These are from Marlow, Buckinghamshire (25,
Treacher Collection), and Beckford, Worcestershire (19, Briggs and Coope Collection),
all in the British Museum (Natural History). The Marlow remains were apparently
predominantly M. arvalis. There was also one first lower molar of M. oeconomus.
Corbet considers that the lack of any second upper molars of M. agrestis, and the
occurrence of only one lower molar of M. oeconomus among the many first lower
molars (the teeth which distinguish these last two closely related species from M.
arvalis), suggest that M. arvalis is the principal species present. The age of the
Marlow rodents is uncertain, but is likely to be Last Glaciation.

Corbet considers that the Beckford specimens, if all one species, must be M.
arvalis. They could be a mixture of M. agrestis and M. oeconomus, but the absence
of the diagnostic second upper molar of M. agrestis or the first lower molar of M.
oeconomus makes this improbable. Beckford is of Last Glaciation age, with a cold
insect-fauna and 14C dates of 27 650+ 250 (Birmingham 293) and 27 300 + 500 years
(Birmingham 595).

At the present state of our knowledge neither the status of M. arvalis in the
British Pleistocene nor the systematic position of ‘M. cornert’ is clear. Craniometric
analysis of the whole collection of Microtus from Ightham would be of great value.
If M. arvalis was present in the British Isles it was an invasion of short duration
during one of the phases of the Last Glaciation.

II4 PLEISTOCENE RODENTS

Microtus nivaloides Major 1902
Vole, extinct

1902 Microtus nivaloides Major : 106, fig. 19.
1923a Microtus nivalinus Hinton : 541-542.

LocaLity: West Runton (Upper Freshwater Bed), Norfolk (65) : Major 1902,
Barrett-Hamilton & Hinton 1910-21, Hinton 1923a, 1926b.

2 fr
eas
P al =e
L Ue Ee pay
“X f /
),
A ep SA
5 HK
Le as
gOS ‘Q ae Bt Fic. 26. Distribution of remains of Microtus
ae 2 YL w ; nivaloides Major in the British Isles.
a ¢
j on
iy ee ah
a wt ) ee
a a ae ge J
a x
De e
ogee oe

DISTRIBUTION IN THE BritTIsH ISLES. M. mivaloides is present in the Cromerian
of Norfolk.

GENERAL DISTRIBUTION. Small representatives of the genus Mzicrotus with the
M, resembling that of M. nivalis, determined as M. mivalinus Hinton, M. nivaloides
Major or M. subnivalis Pasa, have been recorded in ‘Cromerian’ deposits in Germany,
Czechoslovakia, Hungary, Romania, Yugoslavia, Italy and the Ukraine.

SYSTEMATIC REMARKS. In addition to the well-defined form M. nivaloides,
another from the same layer and locality was described as M. nivalinus by Hinton
(1923a). Intermediate specimens between these two forms also occur and it is
probable that only one species is represented. MM. nivaloides, widely distributed in
deposits of “Cromerian’ age in Europe, is probably the ancestor of M. mvalis, a
common species in late Pleistocene localities. In Chaline’s opinion (1972 : 151)
M. nivaloides and M. nivalinus are synonyms of M. nivalis. This is in contradiction
to the picture of evolution in the M. nivalis line as presented by the same author
who considered that M. mivalis developed from M. male: not earlier than the
penultimate glaciation.

Microtus nivalis (Martins 1842)
Snow vole

1907a Microtus nivalis (Martins) ; Hinton : 39-48 ; pl. I, figs 5-23.
1907a Microtus malet Hinton: 49; pl. 1, figs 24-27.

OF BRITISH ISLES 115

Loca.ities : Crayford and Erith (27) : Hinton 1907a, b, rg1ob, Barrett-Hamilton
& Hinton 1910-21, Hinton 1926b, Kennard 1944, Jackson 1947, Chaline 1972,
BM(NH).

Clevedon Cave, Somerset (14): Hinton 1907a, b, Ig9t0b, Barrett-Hamilton &
Hinton 1910-21, Hinton 1926b, Chaline 1972, BM(NH).

Cow Cave, Chudleigh (7) : det. G. B. Corbet.

Gough’s Cave, Somerset (12) : BM(NH).

Tornewton Cave (Glutton Stratum), Devon (3) : Kowalski 1967, BM(NH).

Water Hall Farm Gravel Pit, Hertfordshire (38) : BM(NH).

East Wickham, Plumpstead, London : Hinton 1907b, 1926b, BM(NH).

Some additional records of M. nivalis exist in early papers. This species has
been mentioned from a cave in the Forest of Dean, from Fisherton, Grays Thurrock
and Swanscombe. Hinton (1926b) did not mention these localities in his monograph
as containing M. nivalis and no remains of the snow vole have been seen by the
writers among material collected from these sites.

Fic. 27. Distribution of fossil remains of
Microtus nivalis (Martins) in the British
Isles.

DISTRIBUTION IN THE BritisH ISLES. M. nivalis seems to be a typical element
of the fauna of the period represented by Crayford and the Tornewton Cave Glutton
Stratum. There was a stage when the genus Mucrotus was represented in the
British Isles by two species only, M. nivalis and M. oeconomus. It is entirely
absent from Devensian localities in Britain.

GENERAL DISTRIBUTION. M. nivalis is now distributed in mountain ranges,
mostly in the Mediterranean area, from the Pyrenees in the west to Lebanon, the
Caucasus and Kopet-Dag in the east. It inhabits treeless, mostly rocky or stony
localities, not necessarily cold ones. During the Pleistocene it was more widely
distributed.

SYSTEMATIC REMARKS. M. nivalis has a very variable tooth-pattern. Its
presence in Britain is beyond doubt, but the existence of a second species from this
group, M. male1, does not seem probable. The teeth determined by Hinton (19074)
as belonging to M. male were found exclusively in localities where M. nivalis is

116 PLEISTOCENE RODENTS

numerous, and all intermediate forms between typical nivalis and typical malez can
be found. M. male: must therefore be taken as a synonym of M. mivalis. But
according to Chaline (1972) M. malei is a valid species, and he designates as its
lectotype the specimen M26481 in the British Museum (Natural History) from
Clevedon Cave. In his opinion all the specimens from Clevedon Cave, determined
by Hinton as belonging to M. mivalis, M. malet and M. ratticeps, represent only one
variable species, M. maleiz. M. nivalis developed, according to Chaline, from M.
malet during the late Pleistocene and was not present in Britain. His hypothesis,
as stated above, seems to be contradictory, but it is possible that the representatives
of the group of M. nivalis present in the lowland of Europe, including Britain, were
different from the Recent M. nivalis from the European mountains. In such a case
they must be named Mzcrotus male: Hinton.

Microtus ratticepoides Hinton 1923
Vole, extinct
1923a Microtus rvatticepoides Hinton : 541-542.
Localities: West Runton (Upper Freshwater Bed), Norfolk (65) : Hinton
1923a, 1926b, BM(NH).
Swanscombe (Lower Loam and Upper Middle Gravel), Kent (30) : Schreuder 1950,
Sutcliffe 1964, BM(NH).

3 » (ARP | Fic. 28. Distribution of remains of Microtus
4 << es ‘G : vatticepoides Hinton in the British Isles.
a ¢

)
)
(

DISTRIBUTION IN THE BritTisH Istes, WM. ratticepoides is known from deposits of
Cromerian and Hoxnian age.

GENERAL DISTRIBUTION. M. ratticeboides has been recorded, mostly in
‘Cromerian’ faunal assemblages, in Holland, Germany, Romania, Czechoslovakia,
Poland, Hungary and the European part of the Soviet Union including Georgia.

SYSTEMATIC REMARKS. The tooth-pattern of M. ratticepoides is nearly identical
to that of M. oeconomus, but the dimensions of the fossil form are slightly smaller

OF BRITISH ISLES 117

than those of the Recent one. M. ratticepoides may be the ancestor of the Middle
Pleistocene and Recent M. oeconomus.

Microtus oeconomus (Pallas 1776)
Root vole

1870b Arvicola rvatticeps (Blasius) ; Sanford : 124-125, pl. viii, figs 1, ra—id.
18g0a Microtus (Arvicola) ratticepbs Key. and Bl.; Newton: 453-456, figs 1-2.
IQOI Microtus vatticeps ; Bate : 103-104.

LocaLities: Crayford and Erith (27) : Whitaker 1889, Newton 1890a, Hinton
rg10b, Barrett-Hamilton & Hinton 1910—21, Hinton 1926b, Kennard 1944, Jackson
1947, Hinton 1952, BM(NH).

Harkstead, Suffolk (44) : Ipswich Museum.

Clevedon Cave, Somerset (14): Hinton rgt1ob, Barrett-Hamilton & Hinton
Ig10—21, BM(NH).

Tornewton Cave (Glutton Stratum, Reindeer Stratum, Diluvium), Devon (3):
Kowalski 1967, BM(NH).

Bacon Hole, Glamorganshire (16) : det. A. J. Stuart.

Water Hall Farm Pit, Hertfordshire (38) : BM(NH).

Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924, BM(NH).

Bleadon Cave, Somerset (10) : Palmer 1934, BM(NH).

Chudleigh Fissure, Devon (7) : BM(NH).

Dog Holes Cave, Lancashire (79) : Jackson 1910, Barrett-Hamilton & Hinton
Ig10—21, BM(NH).

Elder Bush Cave, Staffordshire (69) : Bramwell 1964.

Fisherton, Wiltshire (22) : Blackmore & Alston 1874, Bate 1901, Jackson IogI0,
Barrett-Hamilton & Hinton rg10—21, BM(NH).

Gough’s Cave, Somerset (12) : BM(NH).

Great Doward Cave, Herefordshire (17) : BM(NH).

Ightham Fissures, Kent (28): Newton 1894, Bate rgo1, Jackson 1910, Hinton
tg10b, Barrett-Hamilton & Hinton 1g10—21, Hinton 1926b, Zimmermann 1959,
BM(NH).

Isleworth, Middlesex (26) : BM(NH).

Kent’s Cavern, Devon (6) : Sanford 1870b, Kennard 1945-46, BM(NH).

King Arthur’s Cave, Herefordshire (17) : BM(NH).

Langwith Cave, Derbyshire (72) : Barrett-Hamilton & Hinton 1910-21, Mullins
1913.

Levaton Cave, Devon (2) : Carreck 1957.

Marlow, Buckinghamshire (25) : BM(NH).

Merlin’s Cave, Herefordshire (18) : Hinton 1924.

Picken’s Hole, Somerset (10) : Stuart 1974.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Dowel Cave, Derbyshire (71) : Bramwell 1960.

Etches’ Cave, Derbyshire (71) : Pernetta 1966.

118 PLEISTOCENE RODENTS

Lynx Cave, Denbighshire (67) : Blore 1966.

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Cowside Cave No. 3, Yorkshire (78) : BM(NH).

Hay Wood Rockshelter, Somerset (10) : BM(NH).

Somerset levels, Huntspill Cut (8) : Hinton 1952.

Nornour, Isles of Scilly (gt) : Pernetta & Handford 1970.
Creag nan Uamh Cave, Sutherland (83) : Peach & Horne 1917.

Fic. 29. Distribution of fossil remains of
Microtus oeconomus (Pallas) in the British
Isles.

DISTRIBUTION IN THE BriTISH ISLES. M. oeconomus is a widely distributed fossil
species in England and Wales, which also reached northern Scotland. It is a
dominant element of the Crayford fauna and is numerous in the sediments of the
Last Glaciation. It was still present at the end of this period (Nazeing) and dis-
appeared during postglacial times.

GENERAL DISTRIBUTION. MM. oeconomus is now widely distributed in the northern
parts of Europe, Asia and North America. In Europe its westernmost localities
are in Holland and Germany and isolated colonies are known in Hungary and
Austria. Fossil remains are common in the Late Pleistocene and Holocene layers
of central Europe including France, Hungary and Switzerland.

SYSTEMATIC REMARKS. In Chaline’s opinion (1972) M. ratticepoides is a synonym
of M. oeconomus, which developed during the Last Glaciation, its ancestor being
M. malet. This hypothesis is rather contradictory and lacks evidence, at least as
far as material from the British Isles is concerned.

Microtus gregalis (Pallas 1779)
Narrow-skulled vole

1894 Microtus (= Avrvicola) gregalis ; Newton : 197-1098, pl. xi, fig. 12.
1910a Microtus anglicus Hinton : 36-37.

Localities: Aveline’s Hole, Somerset (11) : Davies 1921, Hinton 1921, 1924,
BM(NH).

OF BRITISH ISLES 119

Bleadon Cave, Somerset (10) : BM(NH).

Brixham Cave, Devon (4) : BM(NH).

Chudleigh Fissure, Devon (7) : BM(NH).

Great Doward Cave, Herefordshire (17) : BM(NH).

Ightham Fissures, Kent (28) : Newton 1894, Hinton r1gioa, b, Barrett-Hamilton
& Hinton 1910-21, Hinton 1924, 1926b, 1952, Zimmermann 1959, Chaline 1972,
BM(NH).

Isleworth, Middlesex (26) : BM(NH).

Kent’s Cavern, Devon (6) : Kennard 1945-46, BM(NH).

King Arthur’s Cave, Herefordshire (17) : BM(NH).

Langwith Cave, Derbyshire (72) : Hinton 1924, Mullins 19173.

Levaton Cave, Devon (2) : Carreck 1957 (‘Microtus cf. anglicus’).

Merlin’s Cave, Herefordshire (18) : Hinton 1925.

Picken’s Hole, Somerset (10) : Stuart 1974.

Pin Hole Cave, Derbyshire (73) : Jackson 1934, 1947.

Tornewton Cave (Reindeer Stratum), Devon (3) : Kowalski 1967, BM(NH).

Uphill Cave, Somerset (9) : Hinton 1926b.

Dowel Cave, Derbyshire (71) : Bramwell 1960.

Happaway Cave, Devon (5) : BM(NH).

Nazeing, Essex (37) : Hinton 1952, BM(NH).

Fic. 30. Distribution of fossil remains of
Microtus gregalis (Pallas) in the British
Isles.

Although M. anglicus (= M. gregalis) is recorded from Dog Holes Cave (Jackson
1912) four specimens from this locality preserved in the British Museum (Natural
History) and labelled by Hinton as M. anglicus do not belong to this species but to
M. agrestis.

DISTRIBUTION IN THE BritisH IsLes. WM. gregalis was present in the southern
parts of England during the Last Glaciation, but is absent in all earlier sediments.
It probably reached Britain only once and disappeared before the end of the last
glacial period.

120 PLEISTOCENE RODENTS

GENERAL DISTRIBUTION. WM. gregalis, the unique representative of the subgenus
Stenocranius Kastschenko, is now distributed widely in northern and central Asia
and in arctic Europe from the White Sea in the west to the Bering Strait in the east.
In the western parts of its range it is associated with the tundra belt but in the east,
where the climate is more continental, it is present also in the steppe and desert
zone in eastern and central Asia.

In Europe it was common in the time of the Last Glaciation, reaching France and
Britain in the west and Switzerland and Hungary in the south. According to
Chaline (1972), it was present in France during both the last and penultimate
glaciations.

SYSTEMATIC REMARKS. WM. gregalis was recorded in the Pleistocene of Europe
by Nehring (1875: 7). According to Hinton (1924), fossil remains from Europe
are specifically distinct from M. gregalis and approach M. tuanschamicus Buchner
(now regarded as synonym of M. eversmanni Poljakov). In the modern systematics
of this group M. eversmanni is recognized as a subspecies of M. gregalis. The
morphology and dimensions of the European fossil specimens lie completely within
the range of variability of the polytypic species M. gregalis. According to Chaline
(1972) the name anglicus can be preserved to designate the subspecies of M. gregalis
from the Last Glaciation localities of Europe.

Genus LAGURUS Gloger 1841
Lagurus lagurus (Pallas 1773)
Steppe lemming
1967 Lagurus lagurus (Pallas) ; Kowalski: 115-119, figs 1-24.

SK .
O If } \
reat y \ fie oe :

? S wel \ Fic. 31. Distribution of fossil remains of
ae p pa a Lagurus lagurus (Pallas) in the British
G : ¢ . Isles.

a @

$ ee SS \

aa a } a <~
co eo J
I Se
ae e
Ze a o :

DISTRIBUTION IN THE BritTisH ISLES. Remains of Lagurus lagurus have been
found at only one locality, in the Glutton Stratum of Tornewton Cave, Devon (3)
(Kowalski 1967). Sixty-three first lower molars (3:8 per cent of the total number of

OF BRITISH ISLES 121

rodent remains in this layer) and many other teeth were found in this deposit, which
dates from the cold stage preceding the interglacial hippopotamus-bearing Hyaena
Stratum of this cave.

GENERAL DISTRIBUTION. L. lagurus developed during the Middle Pleistocene
from more primitive forms of the genus Lagurus. It is now distributed in the steppe
zone of Eurasia from the River Dnepr to Mongolia and Chinese Turkestan. Its
fossil remains are relatively common in Upper Pleistocene localities in central and
eastern Europe, but less so in western Europe. In France (Chaline 1973) it has
been found in strata of the penultimate (‘Riss’) and early last (Wtirm) glaciations.

SYSTEMATIC REMARKS. The fossil British Lagwrus remains are identical in
dimensions and molar pattern with recent specimens of this species. The only
character not observed in contemporary material is a tendency among the Tornewton
specimens towards the fusion of the two anterior enamel triangles on the first lower
molar (Kowalski 1967).

Pent otOkY OF THE RODENT FPAUNA OF THE BRITISH ISLES DURING
THE PLEISTOCENE

In the earlier parts of this work rodent localities and rodent species were con-
sidered individually. In conclusion, let us attempt to construct from this infor-
mation a generalized table of the rodent faunas of the various stages of the British
Pleistocene. The construction of such a table must nevertheless be carried out with
reserve. Rodent faunas were not uniform throughout the British Isles. Cold-
adapted species in the north were able to coexist with more temperate species in the
south, and during every glacial and interglacial stage there occurred a shifting of the
vegetational zones which would have caused corresponding movements of the rodent
populations.

Sometimes sea barriers prevented the movement of rodents. The known rodent
faunas of the Lower Pleistocene and of the later glaciations of England are similar
to those of the continent of Europe, suggesting land connections, whereas only a
few rodent species are known from the later interglacials, suggesting separation.
There is no evidence that any rodents crossed to Ireland until some stage during the
Last Glaciation.

Fossil rodent material is seldom rich enough to demonstrate local or temporal
differentiation within particular divisions of the Pleistocene of the British Isles and
there are some stages, especially the earlier glaciations, from which no rodent remains
are yet known. The incompleteness of the record of British rodents was strikingly
demonstrated by the recent discovery of previously unrecognized rodent faunas in
the Westbury Fissure and in Tornewton Cave. Further discoveries can be expected
to provide an ever more detailed sequence for a long time to come.

The most likely stratigraphical relationship of the principal British rodent localities,
at our present incomplete state of knowledge, and the known range of the various
rodent species, is shown in Table 12. A series of faunal stages can be recognized.
In ascending order these are as follows.

122 PLEISTOCENE RODENTS

RED CraG. The study of rodents from the earliest Pleistocene deposit of the
British Isles, the Red Crag, is very difficult because of the mixing in this marine
deposit of contemporary and derived fossils. Castor fiber is apparently a con-
temporary species. Tvogonthertum minus, known from Pliocene sediments on the
continent of Europe, and Hysivix are apparently derived remains of pre-Pleistocene
age.

ICENIAN CRAG AND PASTONIAN ForESsT BED. The next faunal assemblage is that
of the Icenian Crag and the earlier (Pastonian) part of the Forest Bed Series at
East Runton. Mimomys pliocaenicus (from both deposits) and M. rezdi (from the
Icenian Crag) are species of great stratigraphic importance. On the continent of
Europe they are limited to the Upper Villafranchian or Tiglian. There is no evidence
of any difference between the British and European continental rodent faunas at this
stage. Tvogontherium boisvilletti and Castor fiber are two long-ranging rodents also
present in both series of deposits.

CROMERIAN sensu stricto. The rodent fauna from the Cromerian type locality at
West Runton is unusually rich, there being at least eleven species. Tvogontheriwm
boisvilletti and Castor fiber reappear. Mimomys savini replaces M. pliocaenicus ;
Apodemus sylvaticus, Cricetus cricetus, Clethrionomys glareolus, Pitymys arvaloides,
P. gregaloides, Microtus arvalinus, M. nivaloides and M. ratticepoides make an
appearance. Sciurus whiter occurs in the overlying marine ‘Monkey Gravel’.

The West Runton fauna is predominantly of forest and meadow type, testifying
to a mild climate, though Cricetus cricetus could point to colder or more continental
conditions. No arctic elements are present. Allophatomys pliocaemcus and lem-
mings, already present on the continent at the time of the earliest glaciations, are
so far unrecorded from West Runton.

“WESTBURY STAGE’. The rich mammalian fauna from Westbury Fissure, described
by Bishop (1974), does not at present fit into the generally accepted Pleistocene
sequence of the British Isles. Bishop has presented evidence of a previously un-
recognized stage later than the Cromerian sensu stricto of West Runton and possibly
equivalent to Mosbach and other continental sites.

Although the richest rodent-bearing level at Westbury (the ‘Rodent Earth’) is
the latest deposit in the fissure and may be slightly younger than the main bone
deposit (the ‘Calcareous Series’) five rodent species are nevertheless represented in
the latter. Apodemus sylvaticus, Pitymys gregaloides and Mucrotus arvalinus
reappear ; Avrvicola cantiana and Lemmus sp. appear for the first time. All these
forms persist into the ‘Rodent Earth’ which also contained Dicrostonyx sp., the
earliest record in the British Pleistocene, and the only record from the British
Pleistocene of Pliomys episcopalis. Clethrionomys glareolus reappears. Of special
interest is the first appearance in the British Pleistocene of Avvicola cantiana,
which apparently replaces Mimomys savimi, believed to be an earlier stage in the
same phylogenetic lineage, and which is so common at West Runton.

Castor fiber is represented in the lowest water-laid deposits of the cave. The
lowest deposits of Kent’s Cavern and the ‘Forest Bed’ deposit of Ostend, Norfolk,
may also belong to this stage.

TABLE 12

Probable chronological distribution of fossil rodent localities and rodent species from the British mainland and Ireland, according to ava

|
CUMATE

RODENT) ©, | Koenigswald's |
STAGES cals | Mimomys -
IN GiGacet Arvicola
S. BRITAIN eal amips
HOLOCENE FLANDRIAN Tenperate
—_—. | Arvicola fauna
group 3
DEVENSIAN Heel
Joint Mitnor kere |
UPPER Cave interglacial locos
PLEISTOCENE —— a
Tormewton Cave
“Guutton Sratum™ Cod
cold stage
Middle Terrace
of Thames Temperate
interglacial
HOXNIAN Tenpeste Alvicola fauna
¥ group2 |
?
wi T Arvicola fauna
Wes! =
MIDDLE lestbury Stage = Temperate ot
PLEISTOCENE —_—_— == =|
7 7
— =—- —_ +
CROMERIAN Mimomys
¥ Temcerate Savini
sensu stricto fauna |
(Beestonian) ae)

PASTONIAN Forest Bed eure

/ and Teromrate
LOWER patel gree Was
PLEISTOCENE —
FED CRAG —| Terarate

RODENT
Pralhral iecall a fa ] |
IRISH RODENT alae} sie Siete | ello) 2 aie
KOCAUITIES Ea e5) ee ealecelleea iota met eet cca ea ieee
3 3 & & = € 6 = els | = Eps | s | 5]

| eee

Ballynamintra Cave
Castlepook Cave
Castletownroche Cave |
Edenvale Caves
Keshcorran Caves
Kilgreany Cave

Red Cellar Cave

|
BRITISH RODENT LOCALITIES ON THE BRITISH MAINLAND
DEPOSITS |
|
z Cambridge Fens, ?Cowside Cave No.3, Hayward Rockshelter, Huntspill Cut, Staple Howe (Iron Age A, c. 2,400 years ago),
5 Star Carr (Mesolithic, c. 9,500 years ago), Thatcham (Mesolithic c 9,900-9,500 years ago), Nornour, Isle of Scilly (Bronze Age) SCOTLAND
& {Dowel Cave, Etche s Cave, Fox Hole Cave, Nazeing = —— = {itechiatMeree
x
5 uw a f\oInes Hole Great Doward Cave Marlow LATE GLACIAL C14 DATED SITES | Middlestots Bog
alz leadon Cave Gwaenysgor Cave Merlin's Cave Py | Corstorphine
wo] Brixham Cave Harborough Cave ——-Murston eee bead Bere Naren 3/fe00 6250 years age | Creag nan Uamh
<2 Chudleigh Fissure Ightham Fissure Picken's Hole Hackey A OY Cave
Q}3= DogHoles Cave —_King Arthur's Cave Pin Hole Cave Hee Cote ecm cerned veers |
= Ze Elderbush Cave —_Langwith Cave Rowberrow Cave Teen 1280
I5e Fisherton Levaton Cave Uphill Cave Ponders End |
+ | 5 a6 t Minchin Hole, Alveston Fissure, Hessle-— —— —_ — - 7——— -
Fle |
| =9 Barrington, Joint Mitnor Cave, Kirkdale Cave, Swanton Morley, Upnor |
a I" lat
CAVE, -Z| 133 t water Hall Farm Pit— = —_ ——— IPSWICHIAN! sensu/lato
TERRACE 3 Es Bobbitshole
a 5 %® Banwell Cave, Clevedon Cave, Cow Cave (part), 7Goughs Cave (part), ?Hutton Cave Seleeye ine
toke Tunnel Beds
a! +
AuLuvIAL =f 5] 90 =~ Crayford and Erith —— ——_-— — = —— 2West Wittering
wy ? Northfleet
DEPOSITS a Aveley, Grays Thurrock, Ilford, Stutton, Harkstead
|_| —— = = = =
jel =
|
| | Clacton-on-Sea, Copford, Hitchin, Hoxne, Swanscombe (Ingress Vale and Barnfield Pit)
| = = = =
\_ I. IN
| {8
1 | Westbury-sub-Mendip, 7Ostend \ Mundesley Arctic Bed (alternative positions)
CROMER |—*— - oS eee ——— = t E
|
FOREST | _ - — ee ee ee ——_ =
lr ROMERIAN sensu lato
BED | CRO!
Bacton
SERIES | West Runton Gramed
i 7 = = — = Kessingland
Overstrand
| = —— — = Paston
aoa = | Sugworth
East Runton, Happisburgh
ICENIAN
CRAG ; Trimingh
Bramerton, Covehithe, Easton Bavents, Sizewell, Thorpe, Trimingham
RED CRAG Sutton, Woodbridge

=

Altemative.
Positions”

be

o)

lable information
(for Hayward read Hay Wood)

SPECIES ON THE BRITISH MAINLAND a le iin =

IRISH RODENT
— SPECIES
; 3) i | | : : Teel i =|
» = = € | = s | a
s | § 3] . § 5 £3 HI A ll}
r Bacay lege | z| 8 slalele|e See i etc
: Ss 3 ey Pees gi ce = Gi = 3 2 | 2 || RODENT Be ,
: S § onan 8 4 é Ei 2 8 z 2 | 5 arta | Koeniaswald's
3 HN gs | = 3 | $ = @ 3 i & : | 3 a) Bo Sl STAGES Winnie
g : : Rn ieea| z | 3 + | =| 3 8 SS Ele IN ceca | Arvicola
: : E| 3 | 3 2 |] s | : SE ala | SUBRITAIN IN| cil aos ees
Z E 2 | 2 r 5 3 3 2 5 8 ih b
. § Pee | = | 2 BI) 38) ||) 08 8 a sl alle |
4 | a | = = = =. z =I a 3 & | g | 5
| } -- f a =
| | | | HOLOCENE FLANORIAN Temperate
| | SS — | IL ae s
| | | Arvicola fauna
| group 3
| 2 |
| my | | DEVENSIAN ea
! ! |
|
| | \|
} — el ei |
1 Joint Mitnor
I | | UPPER Cavelntergiacial [=e
| = > ie PLEISTOCENE }
1 | Tomewton Cave
| | | "Glutton Stratum” Cold
el ||| ea el |__| |__ cold stage
| | | i] | Middle Terrace
| | / | | of Thames | Temperate
| | interglacial
4 Ly t the ee ay |
| i)
Ne : 2 7 |
! | | | | | P| | Arvicola fauna
l | | 1 HOXNIAN Tenpeate| AMCOR TS
[TNs Le ae oils
| | | ee |
} } | | | } | | } | eet 4 fl f
1 1 lel | | i] | | eal h ;
| Westb t - rvicola fauna
| | i] | | MIDDLE lestbury Stage _| Temperate aan
Pas eee Se SSS | PLEISTOCENE - - |
4} | 1 | aed AW :
| ti | mn rv Pq |e fl q Mimomys
| | CROMERIA Temperate savini
| sensu stricto rata
| |
¢ | | (Beestonian) cat
| rus | | = =
! | | | = | a PASTONIAN Forest Bed | Temperaie
mie ahd | and ae
| |CENIAN CRAG Cold
LOWER
| | = |} — = ; PLEISTOCENE
| t f
| RED CRAG _|Tempenle
1 = | H = L = i

OF BRITISH ISLES 123

MuNDESLEY ArcTIC BED. The Sfermophilus described by Newton (1882b) from
a deposit with remains of arctic plants at Mundesley, Norfolk, provides the only
known instance of a fossil rodent from a British Lower or Middle Pleistocene non-
interglacial deposit. According to Newton the deposit was immediately overlain
by thick glacial deposits which West & Wilson (1966) attributed to the Lowestoftian
(Anglian) Glaciation. They referred underlying interglacial deposits to the
Cromerian sensu stricto.

For the time being the exact relationship between the Mundesley Arctic Bed and
the Westbury deposits must remain uncertain. If, as might be expected, Mundesley
is later than Westbury then it is necessary to explain why the Westbury stage has
not been recognized, except possibly at Ostend, in the Forest Bed sequence of East
Anglia. If, on the other hand, it is earlier then it is necessary to postulate a hitherto
unrecognized cold stage between the Cromerian sensu stricto and the ‘Westbury’ stage.

HOXNIAN DEposiTs. The best sequence of deposits with rodent remains of
Hoxnian age is that of Barnfield Pit, Swanscombe. A considerable time interval is
believed to have occurred within this sequence, between the Lower Loam and the
Middle Gravel (see p. 52). From the lowest deposit, the Lower Gravel, only
Pitymys arvaloides has been recorded. In the overlying Lower Loam this species
is accompanied by Castor fiber, Arvicola cantiana, Muicrotus arvalinus and M.
vatticepordes.

These last two species continue into the Upper Middle Gravel, where they are
joined by Clethrionomys glareolus and Lemmus sp. Although Tvogontherium
boisvillettt and Apodemus sylvaticus have not been recorded from Barnfield Pit, they
are known from a nearby pit in the same terrace deposits at Ingress Vale. Tvogon-
therium borsvillettt appears to have been common during the Hoxnian, being recorded
also from Clacton, Copford and Hoxne. It is unknown from later deposits.

COMPARISON OF THE RODENT FAUNAS of West Runton, Westbury-sub-Mendip and
Swanscombe. Although the geological evidence indicates at least one major glaci-
ation between the time of accumulation of the deposits at West Runton and at
Swanscombe, and although there are very great differences between the large
mammals of the three sites mentioned above, changes in the rodent fauna were
remarkably gradual. The distribution of the various rodent species is shown in
fable 13.

Whilst incomplete collecting probably accounts for many of the gaps in the table
a gradual replacement of species is nevertheless apparent. Mimomys savini,
abundant at West Runton, appears to be replaced at Westbury by Arvicola cantiana.
This species persists at Swanscombe and in slightly later deposits. Mucrotus
arvalinus and M. ratticepoides are present at both West Runton and Swanscombe
(the former occurs also at Westbury), but apparently disappear from the British
Pleistocene after the Upper Middle Gravel. Tvogontheriuwm boisvilletit and Pitymys
arvvaloides are of similar distribution, last seen at Ingress Vale and in the Lower
Loam of Barnfield Pit respectively. Pitymys gregaloides occurs at West Runton
and Westbury but does not reappear at Swanscombe. Mucrotus nivaloides is not at
present known from deposits later than West Runton. Pliomys is at present known

124 PLEISTOCENE RODENTS

TABLE I3

Comparison of rodent species represented at West Runton,
Westbury-sub-Mendip (Bishop 1974 and pers. comm.) and Swanscombe

§
= Westbury-sub-
a Mendip Swanscombe
a a
= Barnfield Pit
FSS
Fs |
: sor 5
8 Toy = LI
Sa ge ame = a ee ri
GO ee ee: we ee
e ns 5 ca g oe Ss
oy =) o 2 O 4 = n
ge BURR Se | tee
a Ne OS Boe. se) Gee
5 | Oo O°. os oon ae 5
Sciurus whiter x
Trogontherium boisvilletti x x
Castor fiber x x 4
Apodemus sylvaticus x x x x
Cricetus cricetus x
Dicrostonyx sp. x
Lemmus sp. x x x
Clethrionomys glareolus x x X
Pliomys episcopalis x
Mimomys savini x

Ayrvicola cantiana x x x x

Pitymys arvaloides x x x

P. gregaloides x x x

Microtus arvalinus x x x x x
M. nivaloides x

M. ratticepoides x x x

only from the youngest deposit at Westbury. Dzicrostonyx and Lemmus, so common
in British Upper Pleistocene deposits, appear for the first time at Westbury. Lemmus
reappears in the Upper Middle Gravel of Swanscombe.

No major break is discernible in the above sequence of rodent faunas. The
Westbury fauna has some characters in common with both West Runton and
Swanscombe, confirming Bishop’s supposition that it is of intermediate age. The
rodent faunas of the three localities are similar to Koenigswald’s Mimomys savint
fauna, Arvvicola fauna group I and Arvicola fauna group 2 of the continent of Europe
(see p. 101) and suggest land connections during the periods concerned. A similarity
of the large mammals of Swanscombe and those from the Holsteinian site of Steinheim
an der Murr in Germany has previously been observed by Sutcliffe (1964).

NOTES ON THE WOLSTONIAN—IPSWICHIAN PART OF THE SUCCESSION. Although
palaeobotanical evidence provides support for only one further interglacial in the

OF BRITISH ISLES 125

British Isles after the Hoxnian it is difficult to fit all the known Upper Pleistocene
rodent faunas into such a simple sequence of events. The generally accepted
Wolstonian—Ipswichian part of the succession appears, from the rodent and other
evidence previously described, to be composite. The so-called ‘Ipswichian’ inter-
glacial is here interpreted as double, with the Middle Terrace of the Thames rep-
resenting a warm stage earlier than the Joint Mitnor Cave warm stage.

MIDDLE TERRACE OF THE THAMES INTERGLACIAL. Reasons for regarding the
deposits of the Middle Terrace of the Thames (type locality Grays Thurrock) as
representing the earlier of these two warm periods have been given on pp. 55-58.
Ilford, Aveley, Harkstead, and Stutton in Suffolk are also included here. The
known rodent fauna of this stage is very sparse. Avvicola cantiana reappears ;
other species include Castor fiber, Apodemus sylvaticus, Clethrionomys glareolus and
Microtus agrestis. Consideration needs to be given to the possibility that this
stage is Hoxnian (Zeuner (1945) believed that Grays was of Penultimate Interglacial
= Hoxnian age) but the occurrence of Avvicola cantiana in other Middle Terrace
localities such as Aveley and Stutton, both of which have been claimed on palaeo-
botanical grounds to be Ipswichian, makes it unnecessary to resort to such an early
date.

TORNEWTON CAVE GLUTTON STRATUM COLD STAGE. The Avvicola from the
Glutton Stratum of Tornewton Cave is the transition form between A. cantiana and
A. terrestyis, suggesting that this deposit is later than the Middle Terrace of the
Thames where the species is A. cantiana. We know that this stage was followed by
a further interglacial, since these deposits were immediately overlain by the Hyaena
Stratum, with remains of hippopotamus. Other species from the Glutton Stratum
are Microtus oeconomus and M. nivalis, which make up the bulk of the rodent fauna,
and also M. agrestis, Lagurus lagurus (the only British record of this species), cf.
Allocricetus bursae and Cricetus cricetus. Dicrostonyx torquatus and Lemmus lemmus
are present, but not abundant.

This assemblage closely resembles that of the Late Middle Terrace of Crayford
and Erith, regarded by Hinton (1926b) as later than that of the Middle Terrace of
Grays Thurrock. Rodents common to both Crayford and the Tornewton Cave
Glutton Stratum are Muicrotus oeconomus, M. nivalis, M. agrestis, Arvicola sp.,
Lemmus lemmus and Dicrostonyx torquatus. Most of the species that are not com-
mon to both deposits are so rare in the British Pleistocene that stratigraphic com-
parison with other localities cannot be made. Spermophilus primigenius from
Crayford and Cricetus cricetus from Tornewton Cave are unknown from any other
post-Hoxnian British sites ; cf. Allocricetus bursae (Tornewton Cave) is known from
only one other British locality, Hutton Cave, which is of uncertain age. Lagurus
lagurus is known only from Tornewton Cave.

The mollusc evidence suggests that the Corbicula Bed at Crayford, in which most
of the rodent remains were found, is still interglacial, though absence of forest
species of land mollusca indicates open grassland. The living representatives of the
subgenus Urvocitellus, to which Spermophilus primigemius belongs, are also open-
country species. On the continent S. primigenius is unknown later than the
‘penultimate’ glaciation.

126 PLEISTOCENE RODENTS

From the available evidence the most likely stratigraphic position for Crayford
would seem to be the end of the Grays—Ilford—Aveley interglacial. S. primigenius
prevents us from attributing a date late in the Pleistocene to the Crayford deposits.
Crayford apparently heralds the arrival of the Tornewton Cave Glutton Stratum
rodent fauna and marks the beginning of the arrival of a great wave of eastern
species of rodents from the European continent. Characteristic of this particular
invasion of continental forms are Spermophilus primigenius, Cricetus cricetus,
Allocricetus bursae, Lagurus lagurus and Microtus nivalis. In addition, there
arrived three other boreal species, Dicrostonyx torquatus, Lemmus lemmus and
Microtus oeconomus. During the coldest phase of this period all rodents connected
with forest environment, including such adaptable species as Clethrionomys glareolus
and Apodemus sylvaticus, apparently disappeared from the British Isles. The
absence of Microtus gregalis, so common in Last Glaciation deposits, is a characteristic
feature of the fauna of this stage.

The apparent occurrence of this cold stage between two interglacial stages with
‘Ipswichian’ floras is not in accordance with current palaeobotanical opinion, which
identifies only one Ipswichian. The severity of this cold stage requires further
assessment. The occurrence of both reindeer and wolverine, in addition to the
above-mentioned rodent species, suggests more than a minor cool phase within an
interglacial. Intense disturbance of the Glutton Stratum and the mixing of it
with a vast quantity of fragments of broken stalagmite formations could be inter-
preted as evidence of very intense frost disturbance later than the Glutton Stratum
but earlier than the deposition of the overlying layers. The relationship of this
apparent cold stage to known glacial stages of the Upper Pleistocene (ice of the
penultimate glaciation is believed to have reached north Devon) must remain, for
the time being, unknown.

Other localities which Hinton (1926b) considered, from the rodent evidence, to be
contemporaneous with Crayford are Clevedon and Banwell Caves. Cow Cave (part),
Gough’s Cave (part), Northfleet and Water Hall Farm (each with Microtus nivalis),
and Hutton Cave (with cf. Allocricetus bursae) may be further examples of deposits
approximately contemporaneous with the Glutton Stratum of Tornewton Cave.

Microtus nivalis apparently disappears from the British Pleistocene after this
stage.

JOINT MITNOR INTERGLACIAL. The occurrence of a further interglacial stage with
hippopotamus is reliably demonstrated at Tornewton Cave where the interglacial
Hyaena Stratum overlies the Glutton Stratum. The occurrence of remains of M.
nivalis and M. oeconomus, apparently underlying hippopotamus-bearing deposits,
at Water Hall Farm gravel pit provides a further example of this relationship.

Important hippopotamus sites dating from this interglacial probably include
Barrington, Joint Mitnor Cave and Swanton Morley. The rodent fauna of this
stage is very sparse. The Avrvicola cantiana-terrestris transition form and Mzcrotus
agrestis are the most common species. Clethrionomys glareolus, Apodemus sylvaticus
and Castor fiber also occur. The Otter Stratum of Tornewton Cave (with the clawless
otter, Cyrnaonyx, and white-toothed shrew, Crocidura, unknown elsewhere in the
British Pleistocene) may represent the beginning of this stage, before the arrival of

Of BRITISH TSLES 127

hippopotamus ; Minchin Hole appears to represent the end, after it had disappeared.
The sparseness of this fauna suggests isolation of the British Isles from the European
continent at that time.

Before leaving discussion of this interglacial, represented also by other sites such
as Trafalgar Square with plant remains, we must refer once more to the Middle
Terrace of the Thames interglacial, indistinguishable from it on palaeobotonical
evidence. Evidence from the rodent remains nevertheless makes it difficult to
explain the sequence other than with two post-Hoxnian interglacial stages separated
by the Tornewton Glutton Stratum cold stage. The sequence becomes even more
complicated if we consider the significance of the Baker’s Hole cold stage. Was this
the same as the Tornewton Cave Glutton Stratum cold stage or (if, as Carreck (in
litt.) has suggested, it occurred between the Ilford and Crayford stages) does it
represent yet another earlier cold stage within the ‘Ipswichian’, sensu lato?

There is at present no means of knowing to which of these warm stages the
Ipswichian type locality of Bobbitshole belongs, although recent studies by Coope
(1974) have showed very close similarities between the insects of this site and of
Trafalgar Square, suggesting that it could be the later one. This is in agreement
with the view of Evans (1971) who assigned the Ipswichian sensu stricto to his half-
cycle 3W (see pp. 40-42).

Last GraciaTIon. The Last (Devensian) Glaciation opened the way for a new
immigration. This time the immigrants included not only both species of lemmings
but also two new arrivals, the ground squirrel Spermophilus superciliosus and the
vole Microtus gregalis. Cricetus, Lagurus and Allocricetus, though present at this
time in France, did not find their way back to Britain. Avvicola had by this time
evolved to A. terrestris (Koenigswald’s Arvicola fauna group 3, see pp. 100-101).
Apodemus sylvaticus, Clethrionomys glareolus and Microtus agrestis are other species
recorded and M. arvalis may have been present for a short time. Castor fiber
apparently disappeared.

This fauna differs substantially from that of the preceding cold phase represented
by the Glutton Stratum of Tornewton Cave, from which it is readily distinguishable.
Diagnostic criteria include the absence of Spermophilus primigenius (found at
Crayford), Cricetus, Allocricetus, Microtus mvalis and Lagurus and the appearance of
Spermophilus superciliosus and Microtus gregalis. Arvicola had evolved from the
A. cantiana-terrestris transition form to true A. terrestris. Dicrostonyx, Lemmus and
Microtus oeconomus were present during both stages.

It is impossible at present to reconstruct in detail all the changes of the rodent
fauna of the Last Glaciation. We have nevertheless now reached the most recent
stage within the Quaternary, except for the Holocene, and the later part of it lies
within the range of radiocarbon dating. Some changes of the rodent fauna are
apparent, which offer a basis for more detailed study in the future, when more and
better-dated material becomes available for study.

During the coldest part of the Last Glaciation Apodemus sylvaticus and perhaps
also Clethrionomys glareolus disappeared from a large part or perhaps all of Britain.
During the warmest interstadial the two lemmings were apparently absent, at least
in the south of England.

Io

128 PLEISTOCENE RODENTS

Three sites with 14C dates are of special importance. These are:

1. Willment’s Pit, Isleworth, west London. 43 140+1520 or —1280 years B.P.
Microtus oeconomus and M. gregalis (possibly the earliest record of this species in
Britain) occur ; no lemmings.

2. Upton Warren, Worcestershire. 41000 years B.P. Diucrostonyx present.

3. Beckford, Worcestershire. 27650+250 years B.P. Probably with Microtus
arvalis.

Coope, who is making a detailed study of the insect faunas of the Last Glaciation,
has named the period of time covered by the above sites the Upton Warren Inter-
stadial complex. He found evidence (pers. comm.) of an intense and short amelior-
ation of climate at about 43 000 years ago, followed by climatic deterioration with
increase of continentality at about 41 000 years. The lack of lemmings at Isleworth
is in accordance with the evidence provided by the insects. The Beckford insect
fauna (pers. comm.) indicates cold conditions.

Other important Last Glaciation sites in the south of England, which lack
lemmings, are Levaton Cave and Picken’s Hole. The deposits of this last-mentioned
site are interpreted by Stuart (1974) as possibly Early Devensian (layer 5, with
Microtus cf. oeconomus and M. cf. gregalis) and Middle Devensian (layer 3, with
Citellus sp. (= Spermophilus sp.) and Microtus cf. gregalis).

The best evidence of the rodent faunas of the end of the Last Glaciation and
early Holocene is provided by the deposits of the Lea Valley and by the Peakland
Archaeological Society’s excavations in Dowel Cave, Etches’ Cave and Fox Hole
Cave. The end of the Pleistocene is marked by a special abundance of lemmings.
Late Glacial deposits in the Lea Valley (dated on palaeobotanical evidence) contained
both Lemmus lemmus and Dicrostonyx torquatus and also Arvicola terrestris, M.
oeconomus and M. anglicus (= M. gregalis). Of these Dicrostonyx torquatus was
still present in pollen Zone III (the last stage of the Pleistocene, ending about
I0 000 years ago) and M. oeconomus survived into the Holocene. The cave sites
mentioned above also indicate an abundance of lemmings in the Late Pleistocene.
Other species present were Avrvicola terrestris, M. gregalis and M. oeconomus, the
last-mentioned species surviving there too into the Holocene.

At the end of the Last Glaciation arctic species of rodents began to disappear,
though both Lemmus and Dicrostonyx seem to have survived at Nazeing until
near the end of this period, the latter species still being present in pollen Zone III.
M. gregalis does not appear to have persisted into post-Pleistocene times.

HoLocEeneE. The postglacial rodent fauna is poorer than that of the neighbouring
continent of Europe, but two species of rodents, now extinct in the British Isles,
persisted. These are Microtus oeconomus and Castor fiber. M.oeconomus apparently
continued in the Lea Valley until pollen Zone V, Mesolithic, about 9000 years ago.
It also survived until Mesolithic times at Dowel Cave and was apparently present on
the Isles of Scilly during the Bronze Age.

Castor fiber, apparently absent during the Last Glaciation, had already reappeared
by about 9500 years ago (pollen Zone IV-—V transition, Mesolithic, Star Carr),
surviving until possibly the thirteenth century. Although the date of the Scottish
finds is uncertain, Castor fiber may have reached Scotland during this stage.

On BRITISH ISLES 129

Arvicola terrestris persisted from the Pleistocene and still occurs in Britain at the
present day. In the Lea Valley Apodemus sylvaticus, Clethrionomys glareolus and
Microtus agrestis reappeared in pollen Zone V-VI (about gooo years ago) after an
apparent interval at the end of the Pleistocene.

Sciurus vulgaris was probably a post-Pleistocene arrival. Others were A podemus
flavicollis, Muscardinus avellanarius and Micromys minutus. It is difficult to decide
whether these last three species reached Britain with or without participation of
human activity. Geological and ecological data suggest rather the second pos-
sibility. Glis glis, Mus musculus, Rattus rattus and Rattus norvegicus are evidently
recent human introductions.

IRELAND. At some stage during the Last Glaciation the two species of lemmings,
but no voles nor any steppe elements, managed to cross to Ireland. A !4C date of
33 000 years for an associated mammoth bone from Castlepook Cave suggests that
Lemmus and Dicrostonyx were probably established in Ireland at that time. Both
are northern species which might be expected to reach Ireland in advance of other
rodents. On the basis of existing evidence, the availability of lemmings for immigra-
tion from the British mainland appears entirely plausible.

Apodemus sylvaticus, recorded from the Pleistocene levels of Castlepook Cave,
may also have arrived in Ireland during the Pleistocene, although it has been
observed that at some cave sites this species was most abundant in the upper levels,
suggesting a later date. Other rodent species which occur in Ireland today are
probably post-Pleistocene arrivals.

V. ACKNOWLEDGEMENTS

The writers gratefully acknowledge help with the compilation of this Bulletin
from many sources.

Dr W. von Koenigswald, of Tubingen University, advised on British representa-
tives of the Mimomys—Arvicola lineage, and commented on the typescript.

Mr M. J. Bishop, of University College, London, provided us with full details of
his determinations of the rodent remains from the Middle Pleistocene fissure deposit
at Westbury-sub-Mendip and allowed us to see the typescript of his report on this
site before publication.

Mr J. N. Carreck (Queen Mary College) read the typescript and provided many
valuable suggestions. He also allowed us to mention important findings from his
own unpublished studies in the Thames estuary.

Dr G. B. Corbet, of the British Museum (Natural History), provided determinations
of rodent remains from Alveston Fissure, Beckford and Marlow and commented on
the typescript.

Dr A. J. Stuart, of the University of Cambridge, provided determinations of
rodent remains excavated in Minchin Hole in 1973 and much other help. He
commented on our typescript and allowed us to see that of his review of British
Pleistocene vertebrates before publication.

130 PLEISTOCENE RODENTS

Dr J. d’A. Waechter provided an important collection of rodent remains excavated
from the Lower Loam of Barnfield Pit, Swanscombe, during 1968-72.

Dr G. R. Coope, of Birmingham University, provided rodent remains from Beckford
and advice on the events of the Last Glaciation in Britain.

Dr R. G. Wolff, of the University of Florida, provided details of rodent remains
collected from Hoxne, Norfolk, by the University of Chicago under grants GS41435
and GS2907 of the National Science Foundation to Dr R. Singer.

Stratigraphic and other information was received from Mr D. Bramwell (Peakland
Archaeological Society), Dr A. S. Clarke (Royal Scottish Museum, Edinburgh),
Dr J. Jewell (British Museum (Natural History)), Mr D. Mayhew (Cambridge),
Professor G. F. Mitchell (Trinity College, Dublin), Mr H. E. P. Spencer (Ipswich),
Dr R. G. West and Mrs G. Wilson (University of Cambridge).

Mr J. J. Hooker sieved Pleistocene sediment from Aveley, Essex, for rodent
remains, Mr J. Simons sediment from Willments Gravel Pit, Isleworth, London, and
from Water Hall Farm Pit, Hertfordshire, all with important results.

Mr A. P. Currant (British Museum (Natural History)) prepared Figs 1, 5-18 and
21-31 and Mrs A. J. Sutcliffe drew Figs 2-4.

Mr D. L. F. Sealy provided much editorial help, including preparing the index,
and Mrs M. N. M. Bhukhureea prepared the typescript with great patience.

VI. REFERENCES

An asterisk (*) distinguishes those papers cited in the text which are not directly concerned
with Quaternary rodents of the British Isles.

ApgBoT, W. J. L. 1917. The ossiferous fissures of the valley of the Shode. Appendix I in:
BENNETT, F. J. Ightham, The Story of a Kentish Village and its Surroundings : 101-1109,
I pl., 2 tables, 1 fig. London.

* AGADZHANIAN, A. K. 1973. (Collared lemmings of the Pleistocene). Nov. tekt. nov.
otlozhi 1 chelovek, Moscow, 5: 320-355. [In Russian.]

Apsimon, A. M., Donovan, D. T. & TayLor, H. 1961. The stratigraphy and archaeology
of the late-glacial and post-glacial deposits at Brean Down, Somerset. Proc. speleol. Soc.,
Bristol, 9 : 67-136.

* AZZAROLI, A. 1953. The deer of the Weybourne Crag and Forest Bed of Norfolk. Bull. Br.
Mus. nat. Hist. (Geol.), London, 2 : 1-96.

BaARRETT-HAMILTON, G. E. H. & Hinton, M. A.C. 1910-21. A History of British Mammals
(unfinished). 1 (1910) : xvi+263 pp., 20 pls. 2 (1911-21): 748 pp., 37 pls. London.

Bate, D.M.A. t1go1. A short account of a bone cave in the Carboniferous Limestone of the
Wye Valley. Geol. Mag., London, (n.s. 4) 8: 101-106.

BEIRNE, B. P. 1947. The history of the British Land Mammals. Ann. Mag. nat. Hist.,
London, (11) 14: 501-514.

BisaT, W. S. 1940. Older and Newer Drift in East Yorkshire. Proc. Yorks. geol. Soc.,
Leeds, 24: 137-151.

Bisuop,M.J. 1974. A preliminary report on the Middle Pleistocene mammal bearing deposit
of Westbury-sub-Mendip, Somerset. Proc. speleol. Soc., Bristol, 13 : 301-318.

1975. Earliest record of Man’s presence in Britain. Nature, Lond. 253 : 95-97.
Biackmorge, H. P. & Aston, E. R. 1874. On Fossil Arvicolidae. Proc. zool. Soc. Lond.
1874 : 460-471, 3 figs.

OPOERELISH ESLES 131

BLorE, J. D. 1966. The excavation of Lynx Cave, 1962-1964. A preliminary report.
Newsl. Peakland archaeol. Soc., Stockport, 21: 17-19, 35.
Boyan, P. J. 1967. The Pleistocene mammalia of the Sewerby—Hessle buried cliff, East
Yorkshire. Proc. Yorks. geol. Soc., Leeds, 36: 115-125.
BRAMWELL, D. 1960. The vertebrate fauna of Dowel Cave — final report. Newsl. Peakland
archaeol. Soc., Stockport, 17 : [9—12, 23] (pages unnumbered).
1964. The excavations at Elder Bush Cave, Wetton, Staffs. North Staffs. J. Fld Studs,
Keele, 4: 46-60.
1970. The tenth report on Fox Hole Cave, High Wheeldon. Bull. Peakland archaeol.
Soc., Stockport, 25 : 8-10.
BREWSTER, T.C. M. 1963. The Excavation of Staple Howe. Malton, E. Riding Arch. Res.
Cttee.
Brown, J. 1852. On the Upper Tertiaries at Copford, Essex. Q. Jl geol. Soc. Lond. 8 : 184-
193.
BuRCHELL, J. P. T. 1935. Evidence of a further glacial episode within the valley of the
Lower Thames. Geol. Mag., London, 72: 90-91.
* BURLEIGH, R. 1972. Carbon 14 dating with application to dating of remains from caves.
Stud. Speleol., London, 2 : 176-190.
CAMPBELL, J. B. & Sampson, C. G. 1971. A new analysis of Kent’s Cavern, Devonshire,
England. Univ. Ore. anthrop. Pap., Eugene, Oregon, 3: I-40.
CaRRECK, J. N. 1957. A late Pleistocene rodent and amphibian fauna from Levaton,
near Newton Abbot, South Devon. Proc. Geol. Ass., London, 68 : 304-308.
1959. In: KerNey, M. P. An interglacial tufa near Hitchin, Herts. Proc. Geol.
Ass., London, 70 : 322-337.
1966. Microtine remains from the Norwich Crag (Lower Pleistocene) of Easton Bavents,
Suffolk. Pvoc. Geol. Ass., London, 77 : 491-496.
Catt, J. A. & Penny, L. F. 1966. The Pleistocene deposits of Holderness, East Yorkshire.
Proc. Yorks. geol. Soc., Leeds, 35 : 375-420.
CHALINE, J. 1972. Les Rongeurs du Pléistocéne Moyen et Supérieur de France. 410 pp.,
17 pls. Paris.
CHEADLE, R. W. 1876. On mammalian bones recently found at Crayford and Erith.
Proc. W. Lond. scient. Ass. Fld Club 1: 70-71.
CoLEMaN, J.C. 10965. The Caves of Iveland. 88 pp., 21 figs, 8 pls. Tralee.
*Coopr, G. R. 1974. Interglacial Coleoptera from Bobbitshole, Ipswich, Suffolk. J. geol.
Soc. Lond. 130 : 333-340.
1975. Anecological study of a temperate interlude in the middle of the Last Glaciation,
based on Coleoptera from Isleworth, Middlesex. J. anim. Ecol., London. 44 : 365-391.
, SHOTTON, F. W. & STRACHAN, I. 1961. A late Pleistocene fauna and flora from Upton
Warren, Worcestershire. Phil. Trans. R. Soc., London, B 244: 379-421.
* CorBET, G. B. 1961. Origin of the British insular races of small mammals and of the
‘Lusitanian’ fauna. Nature, Lond. 191 : 1037-1040.
1966. The Terrestrial Mammals of Western Europe. xi+264 pp. London.
, Cummins, J., Hepces, S. R. & Krzanowskt, W. 1970. The taxonomic status of
British water voles, genus Arvicola. J. Zool., Lond. 161 : 301-316.
CRANBROOK, LoRD 1955a. Fossil Voles. Tvans. Suffolk Nat. Soc., Ipswich, 9: 48-49,
3 figs [published anonymously].
1955b. Records of Mimomys. Trans. Suffolk Nat. Soc., Ipswich, 9:58 [published
anonymously].
Davies, J.A. 1921. Aveline’s Hole, Burrington Combe. Proc. speleol. Soc., Bristol, 1: 61-68.
Dawkins, W. B. & SANFoRD, W. A. 1866-72. British Pleistocene Mammalia, 1 (Felidae).
1+194 pp., 26 pls. Palaeontogr. Soc. (Monogr.), London.
* DIENSKE, H. 1969. Notes on difference between some external and skull characters of
Microtus arvalis (Pallas, 1778) and of Microtus agrestis (Linnaeus, 1761) from the Nether-
lands. Zéol. Meded. Leiden 44 (6) : 83-108.

*

*
*

132 PLEISTOCENE RODENTS

* ELLERMAN, J. R. & Morrison-Scott, T. C. S. 1966. Checklist of Palaearctic and Indian
Mammals. 2znded. 810 pp., map. London, British Museum (Natural History).
* Evans, P. 1971. Towards a Pleistocene time-scale. Inu: The Phanerozoic Time-scale —-a
Supplement. Spec. Publs geol. Soc. Lond. 5: 123-351.
Evans, W. 1907. Mammals of the Edinburgh or Forth area. Pvoc. R. phys. Soc. Edinb.
16 : 387-405.
1913. Lemming remains in south-east Scotland. Scott. Nat., Edinburgh, 17 : 97-100,
1 fig.
* FAHLBUSCH, V. 1969. Pliozdne und Pleistozane Cricetinae (Rodentia, Mammalia) aus
Polen. Acta zool. cracov. 14 (5) : 99-137, pls 8-18.
FALCONER, H. See Murchison, C. 1868.
* FRANKS, J. W. 1960. Interglacial deposits at Trafalgar Square, London. New Phytol.,
Oxford, 59 : 145-152.
FRASER, F.C. & Kine, J. E. 1954. Faunal remains. Ch. 3 iv: Ciarx, J.G.D. Excava-
tions at Star Cary : 70-95. Cambridge.
FRIANT, M. 1962. Sur le Castor plicidens F. Maj. de l’époque des Palafittes. Jb. naturh.
Mus. Bern 1960-62 : 117-125, pls 1-2.
* GIEBEL, C.G. A. 1847. Fauna der Vorwelt mit steter Beriichsichtigung der lebenden Thieve 1
(1), Die Saugethiere. 467 pp. Leipzig.
* Gromov, I. M. 1965. In: Gromov, I. M., Brpixov, D. I., KALABUKHov, N. I. & MEIER,
M.N. Nezemnye belichi (Marmotinae). Fauna SSSR, Moscow & Leningrad, 3 (2) : 1-466.
Hinton, M. A.C. tIgooa. On the Pleistocene deposits of the Ilford and Wanstead district,
Essex. Essex Nat., Buckhurst Hill, 11: 161-165.
1900b. The Pleistocene deposits of the Ilford and Wanstead district. Proc. Geol. Ass.,
London, 16: 271-281.
1901. Excursion to Grays Thurrock. Proc. Geol. Ass., London, 17: 141-144.
1907a. On the existence of the alpine vole (Microtus nivalis Martins) in Britain during
Pleistocene times. Pvoc. Geol. Ass., London, 20: 39-58.
1907b. Note on the occurrence of the alpine vole (Micvotus nivalis) in the Clevedon
Cave deposit. Pyvoc. Bristol Nat. Soc. (4) 1: 190-1091.
I910a. Some new late Pleistocene voles and lemmings. Ann. Mag. nat. Hist., London,
(8) 6 : 34-39.
1910b. A preliminary account of the British fossil voles and lemmings; with some
remarks on the Pleistocene climate and geography. Proc. Geol. Ass., London, 21 : 489-507.
1912. Noteonthe lemming remains from the Arctic Bed at Angel Road. Im: WARREN,
S.H. A Late Glacial stage in the valley of the River Lea. Q. Jl geol. Soc. Lond. 68 : 249.
1914. Onsome remains of rodents from the Red Crag of Suffolk. Ann. Mag. nat. Hist.,
London, (8) 13: 186—195, 1 pl.
1915. Note on British fossil species of Apodemus. Ann. Mag. nat. Hist., London, (8)
15 : 580-584.
1920. Description of the rodent teeth. Ju: TRECHMANN,C.T. Ona deposit of inter-
glacial loess and some transported pre-glacial clays on the Durham Coast. Q. J/ geol. Soc.
Lond. 75 : 201.
1921. Note on the remains of small mammals obtained from Aveline’s Hole, Burrington
Combe, Somerset. Proc. speleol. Soc., Bristol, 1: 74-78.
1923a. Diagnoses of species of Pitymys and Microtus occurring in the Upper Freshwater
Bed of West Runton, Norfolk. Ann. Mag. nat. Hist., London, (9) 12: 541-542.
1923b. Noteon the rodent remains from Clacton-on-Sea. Q. Jl geol. Soc. Lond. 79 : 626.
1924. Notes on the rodent remains from Aveline’s Hole. Pyvoc. speleol. Soc., Bristol,
2: 34-37.
1925. Preliminary note upon the mammalian remains from Merlin’s Cave. Proc.
speleol. Soc., Bristol, 2 : 156-158.
1926a. The Pleistocene mammalia of the British Isles and their bearing upon the date
of the glacial period. Proc. Yorks. geol. Soc., Leeds, (n.s.) 20 : 325-348.

OF BRITISH ISLES 133

Hinton, M. A.C. 1926b. Monograph of the Voles & Lemmings (Microtinae) Living and
Extinct. 1: xvi+488 pp., 15 pls, 110 figs. London, British Museum (Natural History).
1952. Jn: ALLISON, J., GODWIN, H. & WaRREN,S.H. Late-glacial deposits at Nazeing

in the Lea Valley, North London. Phil. Trans. R. Soc., London, B 236: 169-240.
, KENNARD, A. S. & Newton, E. T. 1900. Contributions to the Pleistocene geology of
the Thames Valley. I. The Grays Thurrock area, part 1. Essex Nat., Buckhurst Hill,
11 : 336-370.
& WHITE, G. 1902. Note on the occurrence of Microtus intermedius in the Pleistocene
deposits of the Thames Valley. Proc. Geol. Ass., London, 17: 414-415.
Jackson, J. W. Ig09. On the discovery of the remains of lemmings in Dog Holes, Warton
Crag. Lancs. Nat., Darwen, 2: 227-229.
1910. On the vertebrate fauna found in the cave-earth at Dog Holes, Warton Crag.
Lancs. Nat., Darwen, 2 : 323-332.
1912. Additional notes on the Pleistocene fauna of Dog Holes, Warton Crag. Lanes.
Nat., Darwen, 4: 420-422.
1929a. Remains of lemmings in Derbyshire caves. Naturalist, Hull 1929 : 105-107.
1929b. Report on the animal remains found in Kilgreany Cave, Co. Waterford. Pyroc.
speleol. Soc., Bristol, 3: 137-152.
1932. [Dicrostonyx in Welsh caves.] Proc. Dyserth Distr. Fld Club, Lianfairfechan,
54 [not seen].
1934. Therodent remains from the Pin HoleCave. Rep. Br. Ass. Advmt Sci., Aberdeen :
250-257.
1947. The Upper Pleistocene fauna and its relation to the Ice Age. Proc. Lpool geol.
Soc. 19 : 165-183.
* Janossy, D. 1954. Fossile Microtinen aus dem Karpathenbecken. I. Lemminge. Aumnls
hist.-nat. Mus. nain. hung., Budapest, n.s. 5: 39-48, I pl.
1961. Die Entwicklung der Kleinsaugerfauna im Pleistocén (Insectivora, Rodentia,
Lagomorpha). Z. Sdugetierk., Berlin, 26: 1-11.
1962. Vorldufige Mitteilung tiber die Mittelpleistozine Vertebratenfauna der Tark6-
Felsnische (NO-Ungarn, Biikk-Gebirge). Annis hist.-nat. Mus. natn. hung., Budapest,
54: 155-170.
Jounson, J. P. & WHITE, G. 1900. Some new sections in, and contributions to, the fauna of
the river drift of the Uphall estate, Ilford, Essex. Essex Nat., Buckhurst Hill, 11 : 157-160.
* Kaup, J.J. 1839. Description d’Ossements Fossiles de Mammiferes (&c.)5:91-119. Darmstadt.
KENNARD, A.S. 1944. The Crayford brickearths. Proc. Geol. Ass., London, 55: 121-169.
1945-46. The early digs in Kent’s Hole, Torquay, and Mrs Cazalet. Proc. Geol. Ass.,
London, 56 : 156-213.
KERNEY, M. P. 1971. . Interglacial deposits in Barnfield Pit, Swanscombe, and their
molluscan fauna. J. geol. Soc., London, 127 : 69-93.
* Kine, W. B. R. & OAKLEy, K. P. 1936. The Pleistocene succession in the lower parts of
the Thames Valley. Proc. prehist. Soc., Cambridge, (n.s.) 2: 52-76.

* KOENIGSWALD, W. VON 1970. Mittelpleistozane Kleinsdugerfauna aus der Spaltenfiillung
Petersbuch bei Eichstatt. Mitt. bayer. St. Paldont. Hist. Geol., Munich, 10: 407-432.
1972. Sudmer-Berg-2, eine Fauna des friihen Mittelpleistozans aus dem Harz. Neues

Jb. Geol. Paléont. Abh., Stuttgart, 141 : 194-221.

1973. Veradnderungen in der Kleinsdéugerfauna von Mitteleuropa zwischen Cromer und
Eem (Pleistocin). Eiszeitalter Gegenw., Ohringen, 23-24 : 159-167.

* Kormos, T. 1934. Neue Insectenfresser, Fledermause und Nager aus dem Oberpliocadn
der Villanyer Gegend. Fdldt. Kézl., Budapest, 64 : 296-321.

Kowatski, K. 1966. Stratigraphic importance of rodents in the studies on European

Quaternary. Folia quatern., Krakow, 22: 1-16.

1967. Lagurus lagurus (Pallas 1773) and Cricetus cricetus (Linnaeus 1758) (Rodentia,

Mammalia) in the Pleistocene of England. Acta zool. cracov. 12: 111-122, 24 figs.

1970. Variation and speciation in fossil voles. Symp. zool. Soc. Lond. 26 : 149-160.

*

134 PLEISTOCENE RODENTS

Kretzo1, M. 1958. New names for Arvicolid homonyms. Anmnls hist.-nat. Mus. natn.
hung., Budapest, 50: 55-58.

1965. Die Nager und Lagomorphen von Voigstedt in Thiringen und ihre chronologische

Aussage. Paldont. Abh. Berl. A 11: 585-660, 2 pls, 2 figs.

* KuRTEN, B. 1969. Pleistocene Mammals of Europe. viiit+317 pp. London.

LampLuGH, G. W. 1891. Final report of the Committee... appointed for the purpose of
investigating an ancient sea-beach near Bridlington Quay. Rep. Br. Ass. Advmt Sci.,
Leeds : 375-377.

LANKESTER, E. R. 1864. On new mammalia from the Red Crag. Ann. Mag. nat. Hist.,
London, 14: 353-360, pl. 8.

* LauGEL, A. 1862. La faune de Saint-Prest pres Chartres (Eure et Loire). Bull. Soc. géol.

Fr., Paris, (2) 19: 709-718, 3 figs.
* LEHMANN, U. 1953. Eine Villafranchiano-Fauna von der Erpfingen Hoéhle (Schwab. Alb).
Neues Jb. Geol. Paldont. Mh., Stuttgart, B 1953 : 437-464.
1957. Weitere Fossilfunde aus dem 4ltesten Pleistocan der Erpfingen Héhle (Schwa-
bische Alb.). Mitt. geol. StInst. Hamb. 26 : 60-99.
LYDEKKER, R. 1885-87. Catalogue of the Fossil Mammalia in the British Museum (Natural
History). 1 (Primates, Chiroptera, Insectivora, Carnivora and Rodentia), 1885, xxx+
268 pp. 5 (Tillodontia, Sirenia, Cetacea, Edentata, Marsupialia, Monotremata and Supple-
ment), 1887, xxxv+345 pp. London.
LYELL, C. 1840. On the Boulder Formation or drift and associated freshwater deposits
composing the mud cliffs of eastern Norfolk. Lond. Edinb. phil. Mag., London, (3) 16: 345-
380, 18 figs.
Major, C. I. F. 1902. [Exhibition of, and remarks upon, some jaws and teeth of Pliocene
voles (Mimomys, gen. nov.).]_ Proc. zool. Soc. Lond. 1902 (1) : 102-107.
1908a. The mammalian fauna of the Forest Bed. Geol. Mag., London, (5) 5: 329-330.
1908b. On species of Castor, fossil and recent. Proc. zool. Soc. Lond. 1908 : 630-635.
* MANDACH, E. V. 1938. Skelettreste von Dicrostonyx groenlandicus Trail als Inhalt von
Raubvogelgewéllen. Meddr Gronland, Copenhagen, 112 (4) : 1-54, 22 pls.

* MaTTHEWS, L.H. 1952. British Mammals. xii+ 410 pp., 16+ 48 pls. London.

* MEHELY, L. v. 1914. Fibrinae Hungariae. Die ternaren und quartaren Wurzelzahnigen
Wiihlmause Ungarns. Anmls hist.-nat. Mus. natn. hung., Budapest, 12 : 155-243.

* MILLER,G.S. 1912. Catalogue of the Mammals of Western Europe. xv+1019 pp. London,
British Museum.

MITCHELL, G. F. 1969. Pleistocene mammals in Ireland. Bull. Mammal Soc. Br. Isil.,
Birmingham, 31: 21-25.

, Penny, L. F., SHotton, F. W. & West, R.G. 1973. A correlation of Quaternary
deposits in the British Isles. Spec. Rep. geol. Soc. Lond. 4: 99 pp.

Montacu, I. 1924. On the remains of Fen Beaver in the Sedgwick Museum. Pyvoc. zool.
Soc. Lond. 1924 : 1081-1086.

Muttiins, E. H. 1913. The ossiferous cave at Langwith. J. Derbysh. archaeol. nat. Hist.
Soc., Derby, 35 : 137-153, 6 pls.

Murcuison, C. (ed.) 1868. Palaeontological memoirs and notes of the late Hugh Falconer.
1: lvi+590 pp., 34 pls. 2: xili+675 pp., 38 pls. London.

NEHRING, A. 1875. Fossile Lemminge und Arvicolen aus den Diluviallehm von Thiede
bei Wolfenbiittel. Z. ges. Naturw., Berlin, 45: 1-28, pl. I.

Newton, E. T. 1881. Notes on the Vertebrata of the pre-glacial Forest Bed Series of the
East of England. Geol. Mag., London, (2) 8: 256-259.

1882a. The Vertebrata of the Forest Bed Series of Norfolk and Suffolk. viii+143 pp.,

19 pls. Mem. geol. Surv. U.K., London.

1882b. On the occurrence of Spermophilus beneath the glacial till of Norfolk. Geol.

Mag., London, (n.s. 2) 9: 51-54, pl. 2.

1890a. On the occurrence of lemmings and other rodents in the brick-earth of the

Thames Valley. Geol. Mag., London, (n.s. 3) 7: 452-455, 8 figs.

*

OF BRITISH ISLES 135

Newton, E. T. 1890b. On some new mammals from the Red and Norwich Crags. Q. J/
geol. Soc. Lond. 46: 444-453, pl. 18.

1891. The Vertebrata of the Pliocene deposits of Britain. xi+137 pp., Io pls. Mem.

geol. Surv. U.K., London.

1892. Ona skull of Tvogontherium cuvieri from the Forest Bed of East Runton, near

Cromer. Tvans. zool. Soc. Lond. 13: 165-175, pl. 19.

1894. The vertebrate fauna collected by Mr Lewis Abbot from the fissure near Ightham,

Kent. Q. Jl geol. Soc. Lond. 50: 188-210, pls 10-12.

1899a. [Exhibition of remains of Mus Lewisi from Ightham.] Proc. zool. Soc. Lond.

1899 : 381.

1899b. Additional notes on the vertebrate fauna of the rock-fissure at Ightham (Kent).

Q. Jl geol. Soc. Lond. 55 : 419-429, pl. 28.

1902. Tvogontherium from a Pleistocene deposit in the Thames Valley. Geol. Mag.,

London, (n.s. 4) 9: 385-388.

1909. Hamster remains from the Norfolk Forest Bed. Geol. Mag., London, (5)

6: I10—113.

1916. Tvogontherium from the Pleistocene of Copford, Essex. Geol. Mag., London, (6)

3 : 322-323, 4 figs.
* Norton, P. E. P. 1967. Marine Mollusca in the Early Pleistocene of Sidestrand and the
Royal Society borehole at Ludham, Norfolk. Phil. Trans. R.Soc., London, B 253 : 161-200.
*OKEN, L. 1815-16. Okens Lehrbuch der Naturgeschichte. III Zoology 1: xxviii+ 842+
XVili+iv pp., 40 pls. 2: xvi+1270+(2) pp., 1 tab.; pp. 843-50 (suppl. to 1).
OsBorN, H. F. 1922. Pliocene (Tertiary) and early Pleistocene (Quaternary) mammalia of
East Anglia, Great Britain, in relation to the appearance of Man. Geol. Mag., London,
59 : 433-441.
* Ovey, C. D. (ed.) 1964. The Swanscombe Skull. Occ. Pap. R. anthrop. Inst., London,
20 : xii+215 pp., 25 pls.
Owen, R. 1846. A History of British Fossil Mammals, and Birds. xlvi+560 pp., 236 figs.
London.

1869. On the distinction between Castor and Tvogontherium. Geol. Mag., London,
6 : 49-56, pl. 3, 4 figs.

PaLMER, L.S. 1934. Some Pleistocene breccias near the Severn Estuary. Proc. Geol. Ass.,
London, 45 : 145-161.

PasguierR, L. 1972. Etude d’une population de Mimomys savini Hinton, 1910 (Arvicolidae,
Rodentia) provenant de l’Upper Freshwater Bed (Quaternaire Ancien d’Angleterre).
Mammalia, Paris, 36 : 214-225.

Peacu, B. N. & Horne, J. 1917. The bone cave in the valley of Allt nan Uamh (Burn of
the Caves) near Inchnadamph, Assynt, Scotland. Pyvoc. Roy. Soc. Edinb. 37 : 327-349.
PERNETTIA, J. C. 1966. Report on the findings at Etches’ Cave, Dowel Dale, Derbyshire.

Newsl. Peakland archaeol. Soc., Stockport, 21 : 11-16, 35.
& HaNDFoRD, P. T. 1970. Mammalian and avian remains from possible Bronze Age
deposits on Nornour, Isles of Scilly. J. Zool., Lond. 162 : 534-540.
* PIKE, K. & Gopwin, H. 1952. The interglacial at Clacton-on-Sea. Q. Jl geol. Soc. Lond.
108 : 261-272.
* PoMEL, A. 1848. Description de la téte du Castoroides Ohioensis Forster par Wyman.
Archs Sct. phys. nat., Geneva, 9 : 165-167.
* REICHSTEIN, H. 1963. Beitrag zur systematischen Gliederung des Genus Arvicola Lacépéde
1799. Z. Zool. Syst. EvolForsch., Frankfurt a.M., 1: 155-204.
Rep, C. 1890. The Pliocene deposits of Britain. viii+326 pp., 5 pls, 48 figs. Mem. geol.
Surv. U.K., London.
* RZEBIK, B. 1968. Cvociduva Wagler and other Insectivora (Mammalia) from the Quaternary
deposits of Tornewton Cave in England. Acta zool. cracov. 13 : 251-263.
SANFORD, W. A. 1870a. On the rodentia of the Somerset caves. Proc. Somerset archaeol.
nat. Hist. Soc., Taunton, 1868-69 : 51-57, 1 pl.

Il

136 PLEISTOCENE RODENTS

SANFORD, W. A. 1870b. On the rodentia of the Somerset caves. Q. Ji geol. Soc. Lond. 26:
124-131.
SAVAGE, R. J.G. 1963. Martin Alister Campbell Hinton, 1883-1961. Buiogr. Mem. Fellows
R. Soc., London, 9: 155-170, 1 pl.
1966. Irish Pleistocene mammals. Jy. Nat. J., Belfast, 15: 117-130.
SCHARFF, R. F. 1906. The exploration of the caves of County Clare. Tvans. R. Iv. Acad.,
Dublin, B 33: 1-76.
——, CoFFEy, G., CoLE, G. A. J. & UssHER, R. J. 1903. The exploration of the Caves of
Kesh, County Sligo. Tvans. R. Iv. Acad., Dublin, B 32: 171-214, pls 9-11.
, SEYMour, H. J. & Newton, E.T. 1918. The exploration of Castlepook Cave, County
Cork. Proc. R. Iv. Acad., Dublin, B 34: 33-72+map.
* SCHAUB, S. 1930. Quartare und jungtertiare Hamster. Abh. schweiz paldont. Ges., Basel,
49 (6) : I-49, 2 pls.

SCHLOSSER, M. 1902. Beitrage zur Kenntniss der Saugetierreste aus den siiddeutschen
Bohnerzen. Geol. palaeont. Abh., Berlin, 9: 117-258, 5 pls.

SCHREUDER, A. 1929. Conodontes (Tvogontherium) and Castor from the Tegelen Clay,
compared with Castoridae from other localities. Avchs Mus. Teyler, Haarlem, (3) 6: 99—
320, 16 pls, 16 figs.

1931. Conodontes, Trogontherium and other Castoridae. Palaeont. Z., Berlin, 13 : 148-

176, 6 figs.

1950. Microtinae from the Middle Gravels of Swanscombe. Ann. Mag. nat. Hist.,

London, (12) 3: 629-635.

1951. The three species of Tvogontherium with a remark on Anchitheriomys. Ayrchs
néerl. Zool., Leiden, 8 : 400—433, pls 20-23.

* SIDOROWICZ, J. 1964. Comparison of the morphology of representatives of the genus Lemmus
Link, 1795, from Alaska and the Palaearctic. Acta theriol., Biatowieza, 8 : 217—226.

SINGER, R., WyMER, J., GLADFELTER, B. G. & Wo LFF, R. G. 1973. Excavation of the
Clactonian industry at the golf course, Clacton-on-Sea, Essex. Proc. prehist. Soc., Cam-
bridge, 39 : 6-74.

* SOUTHERN, H. N. (ed.) 1964. The Handbook of British Mammals. xxi+ 465 pp., 60 pls.
Oxford.

* SparkKS, B. W. & WEsT, R.G. 1972. The Ice Age in Britain. xvii+ 302 pp., 44 pls, figs.
London.

SPENCER, H. E. P. 1956. The Hoxne mammalian remains, Gn: West, IkesGa ihe

Quaternary deposits at Hoxne, Suffolk. Phil. Trans. R. Soc., London, B 239 : 265-365.
1964. The contemporary mammalian fossils of the Crags. Tvans. Suffolk Nat. Soc.,
Ipswich, 12 : 333-344.

1966. New mammalian fossils from the Red Crag. Tvans. Suffolk Nat. Soc., Ipswich,

13 : 154-156.

STEVENS, E. T. 1869. Drift series— mammalian remains. In: Some Account of the
Blackmore Museum 2: 109-115. Devizes.

STOPES, C. 1904. Palaeolithic implements from the Shelly Gravel Pit at Swanscombe,
Kent. Rep. Br. Ass. Advmt Sci., Southport : 803-804.

STRINGER, C. B. 1975. A preliminary report on new excavations at Bacon Hole Cave.
Gower, Swansea, 26 : 32-37.

Stuart, A. J. 1974. Pleistocene history of the British vertebrate fauna. Biol. Rev.,
Cambridge, 49 : 225-266.

* SUTCLIFFE, A. J. 1960. Joint Mitnor Cave, Buckfastleigh. Tvans. Torquay nat. Hist. Soc.

13 (1) : 1-26.

1964. The mammalian fauna. In: Ovey, C. D. (ed.), The Swanscombe Skull. Occ.
Pap. R. anthrop. Inst., London, 20: 85-111.

& Bowen, D. Q. 1973. Preliminary report on excavations in Minchin Hole, April—
May 1973. Newsl. Pengelly Cave Stud. Trust 21 : 12-25.

OR BRATS Ee LSS me

SUTCLIFFE, A. J. & CoLtitincs, H. D. 1972. Gnawed bones from the Crag and Forest
Bed of East Anglia. Suffolk nat. Hist., Ipswich, 15 : 497-498.
& ZEUNER, F. E. 1962. Excavations in the Torbryan Caves, Devonshire. I.
Tornewton Cave. Proc. Devon archaeol. Explor. Soc., Torquay, 5: 127-145, pls 26-27.
* TURNER, C. 1970. The Middle Pleistocene deposits at Marks Tey, Essex. Phil. Tvans. R.
Soc., London, B 257 : 373-440.
UssHER, R. J. 1910. Cave hunting. Jy. Nat., Dublin, 19: 37-43.
, SEYMOUR, H. J., Newton, E. T. & ScuarFr, R. F. 1908. On the Cave of Castlepook,
near Doneraile, Co. Cork. Rep. By. Ass. Advmt Sci., Dublin : 697 ; Geol. Mag., London (5)
5 : 462-463; Irv. Nat., Dublin, 17 : 231-232.
* VIRET, J. 1954. Le loess a bancs durcis de Saint-Vallier. Nouv. Arvchs Mus. Hist. nat.
Lyon 4: 1-200, 33 pls.
WarREN, S. H. 1916. Further observations on the Late Glacial, or Ponders End, stage of
the Lea Valley. Q. Jl geol. Soc. Lond. 71 : 164-182.
1955. Lhe Clacton (Essex) Channel deposits. Q. Jl geol. Soc. Lond. 111 : 283-307.
* West, R. G. 1961. The Glacial and Interglacial deposits of Norfolk. Tvans. Norfolk
Norwich Nat. Soc. 19 : 365-375.
1968. Pleistocene Geology and Biology (&c.). xili+377 pp., 16 pls, figs. London.

* 1969. Pollen analyses from interglacial deposits at Aveley and Grays, Essex. Proc.
Geol. Ass., London, 80: 271-282.

* 1972. Land-sea-level changes during the Pleistocene. Ju: A discussion on problems
associated with the subsidence of southeastern England. Phil. Tvans. R. Soc., London,
A 272 : 87-98.

* —__. LAMBERT, A. & SPARKS, R. W. 1964. Interglacial deposits at Ilford, Essex. Phil.
Tvans. R. Soc., London, B 247 : 185-212.

= & Witson, D.G. 1966. Cromer Forest Bed Series. Nature, Lond. 209 : 497-498.

WHITAKER, W. 1889. The geology of London and of part of the Thames Valley 1. ix+
556 pp. Mem. geol. Surv. U.K., London.
*WrymMe_ErR, J. 1957. A Clactonian flint industry at Little Thurrock, Grays, Essex. Proc.
Geol. Ass., London, 68 : 159-177.
—— 1962. Excavations at the Maglemosian sites at Thatcham, Berkshire, England. Proc.
prehist. Soc., Cambridge, 28 : 329-361.
* 1974. Clactonian and Acheulean industries in Britain — their chronology and signifi-
cance. Proc. Geol. Ass., London, 85: 391-421.
ZEUNER, F. E. 1945-59. The Pleistocene Period, its Climate, Chronology and Faunal Suc-
cessions. ist edition (1945), 322 pp. London (Ray Society). 2nd edition (1959), 447 pp.,
80 figs. London.
x 1954. Rissor Wiirm? FEvszeitalter Gegenw., Ohringen, 4—5 : 98-105.
ZIMMERMANN, K. 1959. Uber eine Kreuzung von Unterarten der Feldmaus Microtus
avvalis. Zool. Jb., Jena, (Abt. Syst.) 87: 1-12.

Vil. INDEX

The page numbers of the principal references are printed in bold type. Anasterisk (*) denotes
a figure. Tab. 12 faces p. 122.

Acheulean industry 50-1 alluvial deposits 41, 53, 74, tab. 12; see
Africa 82 localities
alder 52 Alopex see fox (arctic)
Alice and Gwendoline Caves 68; see Eden- Alston, E. R. 37
vale Alveston Fissure 45, 67, 93, 110, tab. 12, 129
Allocricetus 66, 86-7, 127 amphibia 59
cf. bursae 63-4, 66, 86-7, 101, tab. 12, Angel Road 45, 59, 88, tab. 12
125-6 Anglian 41-2, 47-8, 76, 106, 123

Allophaiomys pliocaenicus 105-7, 122 Antian 41, 43

138 INDEX

A podemus 81-4
fiavicollis 62, 67, 70-1, 83-4, tab. 12, 129
lewist 83-4
sylvaticus 48, 52-3, 59, 61, 63-72, 81-3,
tab. 12, 122-7; 129
whitei 52, 81-3
sp. 69, 81, 83
Arctic Freshwater Bed 48
Arvicola 37-9, 52-3, 64, 69, 99-102, 103-5;
see Mimomys—Avrvicola lineage
faunas 61, 100-1, tab. I2, 124, 127
abbotti 103
agrestis 109
amphibia 37, 102-3
amphibius 95, 98, 102-3
terrestris 103
arvvalis 37, 106
arvaloides 106
bactonensis 48, 100, 102, 104
cantiana 48, 51-3, 57, 60-2, 69, 72, 100%,
tor, 102-3, 104-5, tab. 12, 122-5
cantiana-terrestris transition form 60, 64-5,
67, 69, 72, 100*, 101-2, 103, 125-7
cantiana|terrestris complex 104-5
glareolus 92
gveeni 48, 62, 100, 102, 104-5
gregalis 107, 118
gulielmi 37, 87
intermedia 37, 95, 98-9
praeceptor 53, 100, 102, 104
pratensis 92
vatticeps 117
viparia 92
sapidus 99
tervestvis 37, 59, 62, 64-5, 69-71, 99, I00*,
1o1-2, 103-4, 105, tab. 12, 127-9
Sp. 54, 63-7, 69, 72, 125
Asia 76, 80, 82, 84-5, 87, 89, 94, 100, 105,
I12, 118, 120-1
Asia Minor 78
Astian 41, 43, 78, 81
Austria 106, 109, 118
Aveley 45, 49-50, 53, 55-7, 69, 93, IoI-2,
tab. 12, 125-6, 130
Aveline’s Hole 45, 70-1, 81, 83-4, 88, 90, 93,
104, 110, Li2, 1E7—S, tab. 12
Avon, Warwickshire 72
Azov Sea 79

Bacon Hole 45, 67, 81, 93; 104, 11Oj117
Bacton 45, 46*, 48, 78-9, 104, tab. 12
Baker’s Hole 49-50, 59-60

cold stage 127

Ballymote 45

Ballynamintra Cave 45, 68, 82, 88, tab. 12

bank vole see Clethrionomys

Banwell Cave 45, 66, tab. 12, 126

Barnfield Pit 51-3, 80, 90, 92-3, 101-2, tab.
12, 123-4, 130; see Swanscombe

Barntick Cave 68, 82 ; see Edenvale

Barrington 45, 69, 72, 101-3, IIo, tab. 12,
126

Bath 112

Baventian 41, 43

Beaker Age 71

bear 60, 62-3, 65

brown 63-4, 68

Bear Stratum (Tornewton Cave) 63-4, 66

beaver 37, 39, 81; see Castor, Trogontherium

Beccles 45

Beckford 45, 72, 113, tab. 12, 128-30

Beestonian 41, 47-8, tab. 12

Belgium 83, 96

Bering Strait 92, 120

Berkshire 45 ; see Thatcham

Berwickshire 45 ; see Middlestots Bog

Betfia, Romania 95

Bhukhureea, Mrs M. N. M. 130

Biedermann’scher Stbr. Io1

birch 56

birds of prey 60, 69 ; see owls

Bishop, M. J. 108, 129

bison 56

Black Mould (Kent’s Cavern) 62

Blackmore, H. P. 37

Bleadon Cave 45, 70-1, 76-7, I10, 117, II9Q,
tab. 12

Bobbitshole 45, 69, 72, 104, 110, tab. 12, 127

Bos _ see ox, giant

bovids 62-3

Bramerton 43, 45, 46*, 95, 98, tab. 12

Bramwell, D. 130

Brandon 89

Brassington 45

Brean Down 45, 72, 88, 110, tab. 12

Breccia (Kent’s Cavern) 62, tab. 12

brickearth 48, 50, 53-4, 56-8

Briggs, D. J. 72, 113

Bristol Museum 107

Bristol University Spelaeological Society 81,
88, OI

British Museum (Natural History) 36, 58,
60=1,, 66, -72;-97-—8, 102, 107, 113, £16,

119
Brixham Cave 45, 70-1, 93, 104, 110, 112-3,
119, tab. 12

Bronze Age 128 ; see Nornour

INDEX 139

Buckfastleigh 45
Buckinghamshire 45 ; see Marlow
Burchell, J. P: ‘T..60

Burrington 45

Burwell 45

Buxton 45

Calcareous Group (Westbury) 60-1, 122, 124
Cambridge Fens 45, 72, 80, tab. 12
Cambridge, P. 96 (footnote)
Cambridgeshire 45 ; see localities
Canada 92; see North America
Canis see wolf
lupus mosbachensis 61
Cappagh 45
carbon-14 (14C) dated sites 36, 58, 68-9, 72,
113, tab. 12, 127-9
Carnforth 45
Carrant Brook 72
Carreck, J. N. 50, 54, 56, 59-60, 69, 89, QI,
Bs, 104, T1Oj127, 120
Castlepook Cave 45, 68-9, 82, 88, 91, tab. 12,
129
Castletownroche Cave 45, 68-9, 88, 9gI,
tab. 12
Castor 39, 54, 69, 79-81
euvopaeus 79
fiber 37, 43, 47-8, 52-3, 61-2, 70-3, 79-81,
tab. 12, 122-8
plicidens 81
plicidens 79-81
tvogontherium 78
veterioy 79-80
Sp. 53, 79
Castoridae 78-81
Catacombs 68 ; see Edenvale
Caucasus 83, 115
cave deposits 41-2, 56, 58, 60-9, tab. 12;
see localities
disturbed 67
Cave Earth (Kent’s Cavern) 62, Io1, 104,
tab. 12
Cervus see red deer
Chaline, J. 39
channels, buried 48-9
Cheddar 45
China 82, 87
Chinese Turkestan 121
Chudleigh (Fissure) 45, 70-1, 88, 90, 93, 104,
LO; £17, 119, tab. 12
Citellus 75, 128; see Spermophilus
evythrogenoides 75
evythrogenys 75

eversmanni 75
nogaict 75
(Colobotis) superciliosus 75, 77
(Urocitellus) polonicus 75-6
primigenius 75
sp. 76
Clacton-on-sea 45, 49*, 52-3, 78-80, 93,
IOI-2, 109, tab. 12, 123
Clactonian industry 50-1, 53
Clare, Co. 45 ; see Edenvale Caves
Clarke, Dr A. S: 130
classification 73-121
‘clay pebbles’ 47
Clethrionomys 92-4
glarveolus 37, 48, 52-3, 59-67, 70-2, 92-4,
tab. 12, 122-7, 129
nagevt 94
Sp. 53, 67, 92
Clevedon Cave 45, 66, 93, III, 115-7, tab. 12,
126
Coelodonta see rhinoceros, woolly
Coffey Cave see Keshcorran
Colchester 45
Colobotis see Spermophilus superciliosus
Connaberry 45
Conodontes Boisvilletti 78
coombe rock 50, 59-60; see _ solifluxion
deposits
Coope,“G.R.«72,, 183;°127=8, 130
Copford 45, 69, 78-9, tab. 12, 123
Coralline Crag 41, 96 (footnote)
Corbet, Dr Gr b.58-72,93, LIO=1, PIS prrg,
129
Corbicula Bed 54, 89, 92, 125
Cork, Co. 45; see localities
Corstorphine 45, 73, 88, 110, tab. 12
Covehithe 43, 45, 46*, 96, tab. 12
Cow Cave 45, 66, 115, tab. 12, 126
Cowside Cave No.3 45, 70-1, 104, IIo, 118,
tab. 12
Cranbrook, Earl of 98
Crags 37, 41-2, 60; see Coralline, Icenian,
Norwich, Red, Shelly and Weybourne
Crags
Crayford 38, 45, 49*, 50, 54, 60, 64, 66, 75,
O2¢ TOL 3, oITL, LES; aak7— 6 sta ba pra,
125-7
Gravel 50
Creag nan Uamh Cave 45, 68, 88, 110, 118,
tab. 12
Cresswell 45 ; see Pin Hole Cave
Cricetidae 85-121
Cricetulus migratorius 87
Cricetus 48, 85-6, 127

140 INDEX

cricetus 48, 63-6, 85-6, tab. 12, 122, 124-6
major 86
vuntonensis 86
vuntonensis 86
songavus 37, 80-7
vulgaris Runtonensis 37, 85
Crocidura 64, 126
Crocuta see hyaena
Cromer 45, 46*, 78, 98, tab. 12
Cromer Forest Bed Series (Norfolk) 36-7,
40-1, 43, 46-8, 74, 77, 79, 85, 94, 97-8,
105, tab. 12
Cromerian stage 40-1, 47, 60-2, 69, 78, 82-3,
85; 93). 106, 98; 201, 105=8) 1984, 116;
tab. 12
sensu stricto 47-8, 52, 61-2, 73, tab. 12,
122-3
Currant, Ay Py 130
Cyrnaonyx 64, 126
Czechoslovakia 76, 86-7, 89, 96-9, 106, 109,
£4, FLO

Dama _ see fallow deer
deep-sea cores 40
deer see fallow, giant, red deer, reindeer, &c.
Denbighshire 45 ; see Lynx Cave
Denmark 113
Derbyshire 45 ; see localities
Devensian 41-2, 115, tab. 12, 127-8
Devon 45, 126; see localities
Diabroticus Schmerlingi 78
Dicerorvhinus see rhinoceros, narrow-nosed
etyvuscus 61, 69
hemitoechus 56
kirchbergensis 56
Dicrostonyx 67, 69, 72-3, 87-90, 92, III,
122-3, 127-9
gulielmi 87, 90
henseli 87, 90
hudsonius 90
simplicior 89
torquatus 37, 54, 59, 62-8, 70-2, 87-90,
g1-2, tab. 12, 125-6, 128
sp. 61, 87, 89, 124
Dierden’s Pit 51; see Swanscombe
‘Diluvium’ (Tornewton Cave) 62-4, 66, IIo,
117
Dipoides Lydekkeri 78
Diss 45 ; see Hoxne
distribution 73-121
Dnepr, river I21
Dog Holes Cave 45, 70-1, 77, 81, 88, 90, 93,
104, 109-10, I12-3, 117, 119, tab. 12
domestic animals 84

Domnitz 42

Doneraile 45

dormouse see Muscardinus, Glis

Dove Holes 60

Dowel Cave 45, 70-1, 74, 82-3, 88, 91, 93,
ki7 ~§10o, tabs 12) 128

Dowel Dale 45; see Dowel Cave, Etches’
Cave

Drepanosorex savini 101

Droitwich 45

Earl Sterndale 45
East Anglia 42, 46*, 98; see localities
deposits in 42-8, 69, 72
East Runton 38, 45, 46*, 47-8, 78-80, 96,
OS, UO, U2 122
East Wickham 115
Easton Bavents 43, 45, 46*, 95, 98, tab. 12
Ebbsfleet 49-50, 59-60
Eburonian 41, 43
Edenvale Caves 45, 68, 82, 88, tab. 12
Edinburgh 45; see Corstorphine
Edmonton 48, 49*, 59
Edrom Parish 45 ; see Middlestots Bog
Eemian 40-2, 100
Elder Bush Cave 45, 70-1, 88, 91, 93, 104,
ELO, Li7, tab. 42
elephant, straight-tusked 56-7
Elsterian 41, 61, 100
Ennis 45
Eocene 43
Eppelsheim 76
Equus see horse
mosbachensis 61
Erith 38, 45, 49*, 50, 54, 75, 87, 89-90, 92,
Ly, LE7, tab, 12,925
Erkenbrechtsweiler 101
Erpfingen 1 and 2 Io1
Essex 45; see localities
Etches’ Cave 45, 70-1, 83, 88, 91, 117, tab.
12, We
Europe, continent of 36, 39, 43, 54, 74, 76-8,
80, 82-5, 87, 89, 92, 94,97, 99, 105-8,
112-4, 116, 118, 120-2, 124, 126-8
Evans, P., half-cycles 40, 42, 127
Evotomys glareolus 92
harrisont 92, 94
intermedius 95, 97-8
kennardi 92, 94
nagert 94
Sp. 92, 94

Falconer, H. 37
fallow deer 56, 63

INDEX 141

Felis see lynx
gombaszoegensis O1

Felixstowe 45, 46*, 78

field-vole see Arvicola

Finland 112

fir cones, gnawed 74

Fisherton 45, 72, 76, 88, 115, 117, tab. 12

Fitch, Mr 96 (footnote)

Flandrian 41, 59, tab. 12

Flint bed 47

Flintshire 45 ; see Gwaenysgor Cave

Floodplain Terrace complex (Thames) 54,
55-8

footprints 51

Forest Bed 42, 46*, 48, 53 ; see Cromer

Forest of Dean 115

fox, arctic 72

Fox Hole Cave 45, 70-1, 88, tab. 12, 128

France 74, 76, 78-9, 83, 87, 89, 95-9, 105,
HOG; TLS, E26=1,11827

Geldeston 45, 46*, 98

geological background 40-2

Geological Society of London 40, 50, 55

Georgia 106, I16

Germany 76, 79, 87, 92, 96-9, 106, Io9, I14,
iO; 118, 124

giant beaver see Tvogontherium

giant deer 68

Gippingian, Gipping cold stage 41, 56

glacial till 48

Glamorganshire 45 ; see Penard

Gliridae, glirids 43, 77-8

Glis glis 129

Gloucestershire see Alveston Fissure

glutton 63, 126

Glutton Stratum (Tornewton Cave) 63-6,
82, 85-6, 88-90, 92-3, I01-3, I10-I,
I15, 117, 120, tab. 12, 125-6

cold stage tab. 12, 125, 127

Gough’s Cave 45, 66, I04, IIo, II5, I17,
tab. 12, 126

Grays Thurrock 38, 45, 49*, 50, 53-4, 57, 69,
80-1, 93, IOI-3, 109, III, 115, tab. 12,
125-6

Great Doward Cave 45, 70-1, 77, 81, 88, 91,
Os, 1L0,, 117, 119, tab. 12

Green, Rev. C. 48

Greenland 92

Gromoy, I. M. 75, 77

ground squirrels see Spermophilus

Gulo_ see glutton, wolverine

Giinz, Giinz-Mindel 76

Gwaenysgor Cave 45, 70-1, 81, 88, 93, tab. 12

Hackney (Marshes) 45, 48, 49*, 59, 88, tab. 12

hamster see Cricetus, Allocricetus

Happaway Cave 45, 70-1, 82-3, 93, 104, IIo,
LO, tabs 12

Happisburgh 45, 46*, 47-8, tab. 12

Harborough Cave 45, 70-1, 91, tab. 12

Harkstead( 45, 60,72, 102—3, 1 10).117, tab;
22) 125

harvest mouse see Micromys

Hay Wood Rockshelter 45, 70-1, 80, 118,
tab; 12

Helmsley 45

Heppenloch ro1

herbivores, large, trampling by 54

Herefordshire 45 ; see localities

Hertfordshire 45; see localities

Hessle 45, 69, 102, 104, tab. 12

High Wheeldon Hill 45 ; see Fox Hole Cave

hillwash 49 ; see Northfleet

Hinton, M. A. C. 35-40

rodent faunas 40-1

hippopotamus, Hippopotamus 50, 56-9, 63,
67, 69, 86, 93, 102, 121, 125-7

history of rodent faunas 121-9

Hitchin 45, 69, 81, 93, tab. 12

Holland 43, 79, 83, 96-8, 116, 118

Holocene 41, 59-60, 62-4, 67, 70-3, 77, 80,
82-3, 93, TOO, T02, 104, 113, .tab. 12;
128-9

Holsteinian Interglacial 40-2, 79, 83, 95, 105,
124

Homo, hominids 51 ; see industries, man

Homotherium 62

latidens 61

Hooker, J. J. 130

hornbeam 56

horse 56, 63, 72

Hoxne 42, 45, 69, 78-9, 81, 90, 92, 104, tab.
12, 123, 130

Hoxnian Interglacial 40-2, 52-3, 55-6, 65,
69, 78-80, 82, 92-3, 100-1, 104-8, II0,
116, tab. 12, 123, 125

Hull see Kingston-upon-Hull

Hunas Iol

Hundsheim 61, 101

Hungary 74, 76, 87, 89-90, 92, 95-8, 106,
LOO, 114; 116, £18,.120

Humntspill Gut 45, 73; 11S) Gtabs 125) See
Somerset

Hutton Cave 45, 58, 66-7, 86-7, 90, 93, tab.
I2, 125-6

hyaena 62-3, 67, 69

Hyaena Stratum (Tornewton Cave) 63-6,
LO4, 110; F24, tab,, 12; 1125-6

142 INDEX

Hypudaeus bucklandit 109
spelaeus 103
Hystrix 43, 122

Iberian peninsula 99 ; see Spain
ice covering Britain 113
ice wedges 58
Icenian Crag 41, 43, tab. 12, 122
Ightham Fissures 37-8, 41, 45, 67, 76-7, 81,
83-4, 88, 91, 93-4, IOI, 104, I10, 112-3,
LIZ. cro, tab: 12
Ilford 45, 49-50, 53, 55-8, 60, 69, 80, 103,
109, tab. 12, 125-7
Inchnadamph 45, 67; see Creag nan Uamh
Cave
Ingress Vale 51-2, 79, 81, 93, 100-2, tab. 12,
123-4}; see Swanscombe
insects 36, 58, 72, 113, 127-8
Institute of Geological Sciences Museum 104
Ipswich 45, 72 ‘
Ipswich Museum 72, 97, 103-4, I10, II7
‘Ipswichian’ 57, 63, 67, 69, 72, tab. 12, 125-7
Ipswichian stage (interglacial) 40-2, 50,
53-7, 02, OL, 12455))227),
Ireland 42, 91, 129; see counties
caves 68-9
rodents in 121
voles in 111-2
Iron Age 84; see Staple Howe
Isleworth 45, 48, 49*, 58, 117, 119, tab. 12,
128
Italy 95, 97-8, 106, 114; see Val d’Arno

Jewell, Dr J2 2130
Joint Mitnor Cave 45, 56, 67, 82, 93, 101-3,
I1o-I, tab. 12, 126
interglacial tab. 12, 125-6

Kadzielnia 98

Kalkberg, Nagyharsanyberg 76

Kent 45, 50; see localities

Kent’s Cavern 45, 60, 61-2, 80, 83, 88, 91, 93,
LOL, 104, 107,11, 117, 119, tab i222

Kerney, M. P. 51, 60, 63

Keshcorran Caves 45, 68, 82, 89, tab. 12

Kessingland 45, 46*, 78-9, 98, tab. 12

Kilgreany Cave 45, 68-9, 82, 84, 89, III,

tab. 12
King Arthur’s Cave 45, 70-1, 88, I10, 117,
10, tab; 'T2

Kingston-upon-Hull 45 ; see Hessle

Kinlock 45

Kirkdale Cave 45, 56, 67, 81, 93, 104, IIo,
tab. 12

Koenigswald, W. von 36, 39, 64, 67, 69, Io1,
129

Mimomys-Arvicolafaunal groups of 100-1,
tabok2) 24a 27
Konéprusy 101
Kopet-Dag 115
Kyson 45, 96

Lagurus 120-1, 127
lagurus 63-6, 120-1, tab. 12, 125-6
Lancashire 45 ; see Dog Holes Cave
Langwith Cave 45, 70-1, 74, 76, 82, 84, 88,
QI, 63,. 104, TIO; 112, 1L7, PIO Maw. Ie
Last Glaciation 50, 54, 56, 58-60, 62-4,
67-72, 76, 82, 89-94, 102, 104, III, 113,
118-21, 126, 127-8, 129
Last Interglacial 40, 42, 50, 54-5, 50*, 59,
80, 89, 93, 102, 104
last interglacial 65, 78
Lea Valley 38, 41, 48-60, 49*, 58-9, 128-9;
see Nazeing, Ponders End, Water Hall
Farm, &c.
Lebanon 115
lemming 54, 64, 67, 122, 127-9; see Lemmus,
Dicrostonyx, Lagurus
Lemmus 37, 67-8, 90-2, 123-4, 127-9
lemmus 54, 59, 62-6, 68, 70-1, 90-2, tab.
12, 125-6, 128
norvegicus 37, 9O
torquatus 87
SP. 52,101, (09,090) E2254
Levallois industries 49-50, 54
Levaton Cave 45, 67, 70-1, 82, 104, I10, 112,
117, 119, tab. 12, 128
Limerick, Co. 45 ; see Red Cellar Cave
lion 56, 63
Little Thurrock 50
localities with fossil rodents 42-73, 44-5*
Loch of Marlee 45, 73, 80, tab. 12
London 48, 58; see Hackney, Angel Road,
CEC
Lough Gur 45
Lower Crayford Brickearth 50
Lower Loam (Swanscombe) 42, 52, 102, 108,
EEG, 123—4,. 130
Lower Thames Valley province 48
Lowestoftian 41, 123
Ludhamian 41
lynx 68
Lynx Cave 45, 70-1, 82, 88, 110, 118, tab. 12

INDEX 143

MacEnery, J. 62

Macoma balthica 43

Main Chamber see Tornewton Cave

Major, C. I. Forsyth 75

Male, H. C. 66

mammals other than rodents 56, 60, 72; see

names

mammoth 50, 56-60, 62, 67-9, 72, 87, 129

Mammuthus 59; see mammoth
primigenius 50
tvogontherit 50, 56

man, introductions to Britain by 83-4, 129
remains, human IItI

Marlow 45, 48, 58, 113, 117, tab. 12, 129

mastodon 43

Mauer 61, Iol

Mayhew, D. 130

Mediterranean I15

Medway, Upper Floodplain Terrace of 54

Megaceros see giant deer

melt-water deposits 59-60

Menapian 41

Mendips_ see Bleadon Cave, &c.

Merlin’s Cave (= Wye Cave) 45, 70-1, 88, 91,
@3, 1@f, 104,110; 117, 110, tab. 12
Mesolithic 62, 128; see Star Carr, Thatcham

Micromys 66, 84
minutus 65, 84, tab. 12, 129
sylvaticus 81
Microtinae 37-8, 43, 107, III
Microtius (Pitymys) arvalidens 106
Microtus 69, 105, 107-8, 108-20
agrestis 37, 53-4, 58-9, 62-9, 72, 108,
109-12, 113, 119, tab. 12, 125-7, 129
neglectus 109
of agrestis group 53
agvestoides 109, I11
anglicus 60, 118-20, 128
avvalinus 48, 52, 61, 108-9, 110-1, tab. 12,

122-4

ayvvalis 58, 60, 72, 108-9, 112-3, tab. 12,
127-8

sp. (avvalis/agrestis group) 52, 60, 70-1,
108, 109, 111

cornervt 109, 112-3

everysmanni 120

glaveolus 92

gregalis 58-60, 62-4, 66, 70-I, 107, 113,
118-20, tab. 12, 126-8

intermedius 102

malet 114-6, 118

nivalinus 114

nivalis 37, 54, 59-60, 63-6, 72, 114-6,
tab. 12, 125-7

nivaloides 48, 114, tab. 12, 122-4
oeconomus 37, 54, 58-9, 62-3, 65-8, 70-3,
108, 113, 115-6, 117-8, tab. 12, 125-8
ovcadensis 112-3
vatticepoides 48, 52, 116-7, 118, tab. 12,
122-4
vatticeps 68, 116-7
subnivalis 114
tiamschanicus 120
sp. 66, 69
(Pitymys) arvalidens 105-6
sp. 106
Middle East 100
Middle Terrace of Thames interglacial tab.
P2125) 127
Middlesex 45; see Isleworth, Angel Road,
Ponders End
Middlestots Bog 45, 73, 80, tab. 12
Midlothian 45 ; see Corstorphine
Mimomys 60, 95-100
cantiana, cantianus 100, 102
intermedius 98-9
majori 98-9
millert 98-9
newtont 43, 47, 97-8, tab. 12
petenyt 97
pliocaenicus 43, 47, 95-6, tab. 12, 122
polonicus 96
Veidt. 43, 97, tab; 12, 122
savini 47-8, 52, 69, 98-9, 100-1, 105, tab.

12, 122—4
fauna tab. 12, 124; see Arvicola
SP: 43
Mimomys—Arvicola lineage 36, 100, tab. 12,
129
Minchin Hole 45, 67, 103, 110, tab. 12,127, 129
Mindel 76
Miocene 43

Mitchell, Professor G. F. 130
mollusca 43, 51-2, 54, 56, 58, 63, 125
Monastirian, Late 40, 42, 50, 55
Main 42, 50, 55
Mongolia 121
‘monkey gravel’ 48, 73, 122
Mosbach 61, IoI, 122
multituberculate origin of rodents 38
Mundesley 45, 46*, 48, 75-6, 78-9, tab. 12,
122
Arctic Bed. tab. 12, 123
Muridae, murids 43, 81-4
Murston 45, 88, tab. 12
Mus 84
Abbotti 83
Lewisi 83

144 INDEX

musculus 81, 84, tab. 12, 129

songarus 86

sylvaticus 81

sp. 81
Muscardinus 48, 77-8

avellanarius 67, 70-1, 77-8, tab. 12, 129
Myodes lemmus 90

torquatus 87

narrow-skulled vole see Microtus gregalis

Nazeing 45, 48, 49*, 59, 82, 88-9, 91, 93, 104,
118-09, tab. 12, 128

Neolithic 62, 83

Netherlands see Holland

Newhall Caves 68; see Edenvale

Newton; E13 37

Norfolk 38, 42, 45-6, 113 ; see localities

Nornour 45, 73, 118, tab. 12

North America 89-90, 92, 94, 118

Northfleet 45, 49*, 50, 59-60, 93, 103, I10,
tab. 12, 126

Norwich Crag 38, 41, 43, 46*, 47, 78, 95-8;
see Icenian Crag

Norwich Museum 96 (footnote)

oak 56

Orkney Islands 109, 112-3

Orkney vole 112; see Microtus arvalis

Orsett Road 50, 53

Ostend (Norfolk) 45, 46*, 48, 53, 74, 101-2,
104, tab: 12, 123

Forest Bed 47-8, 53, 74-5, 92, 122-3

otter, clawless 64, 126; see Cyrnaonyx

Otter Stratum (Tornewton Cave) 36, 64-5,
101-3, tab. 12, 126

Overstrand 45, 46*, 78, tab. 12

Owen, R. 37

owls 69, I1I2

ox, giant 56

Oxfordshire 45 ; see Sugworth

palaeobotany 39-40, 46-8, 55, 57, 59, 124-7;
see pollen, plants

Palaeolithic 62-3, 66, 71

Palaeoloxodon antiquus 59; see elephant,
straight-tusked

Palestine 105

Panthera see lion

Pavapodemus coron. 101

Paston 45, 46*, 78, tab. 12

Pastonian (Forest Bed) 41, 43, 47-8, 78-9,
96, 98, tab. 12, 122
Peak District 36
Peakland Archaeological Society 67, 128, 130
Penard 45
Penkridge 89
Penultimate Glaciation 42
‘penultimate’ Glaciation 54, 56, 59, 64-5,
LE2, Lisson) a25
Perpignan 78
Persia 105
Perthshire 45 ; see Loch of Marlee
Petersbuch 101
Phodopus 87
sanford 86
songorus 86-7
Picken’s Hole 45) 70-1, 76, 117, 119,tabaae,
128
Pin Hole Cave 45, 70-1, 76-7, 82-3, 88, 91,
93, LO4, [HO Ig EHO, tab: ire
pine 52, 56
pine vole see Pitymys
Pitymys 105-8
avvalidens 105
avvaloides 48, 52, 105, 106, 107, tab. 12,
122-4
gregaloides 48, 61-2, 105-6, 107, tab. 12,
122-4
sp. 69
plants 53, 58, 122, 127; see palaeobotany
Pleistocene 67, 74, '82=5, “87, ‘o1=—2 eam)
I21I-9, tab. 12
generalized British sequence 41
Lower 43, 60, 98, 121
lower middle 80, 87, 89, 92, 94-5
Upper 83
Phocene 41; 43, 112, 122 ;- “Pliocenes37,.46
Pliomys 95, 123
episcopalis 61, 95, Io1, tab. 12, 122, 124
lenkt 101
Plumpstead 115
Plunkett Cave 68; see Keshcorran
Poland 76, 87, 89, 92, 96-8, 106, 116
pollen 36, 52, 55-6, 89, 128-9; see palaeo-
botany
zonation of “Ipswichian’ sites 56*, 57-8
Ponders End 45, 48, 49*, 50, 59, 88, tab. 12
post-Ipswichian 67
post-Pleistocene 68, 73 ; see Holocene
‘preglacial’ deposits 37
pre-‘Ipswichian’ 65-7
Prestatyn 45
Prezletice 101
Pyrenees 115

INDEX 145

radiocarbon dating see carbon-14
Rangifer see reindeer
rat see Rattus
Rattus 84
norvegicus 84, tab. 12, 129
vattus 67, 70-1, 74, 84, tab. 12, 129
Red Cellar Cave 45, 68, 89, tab. 12
Red Crag 41, 43, 46*, 78-80, 95, tab. 12, 122
basal nodule bed 43
red deer 56, 63, 67
reindeer 58, 63, 67-9, 72, 126
Reindeer Stratum (Tornewton Cave) 63-6,
88-9, 93, 101, 104, 110, 117, 119, tab. 12
Reuverian 41
rhinoceros 56, 59, 63
narrow-nosed 63
woolly 62, 67, 69, 72
iiss 42, 76, 100, TT2, 121
Rodent Earth (Westbury) 60-1, 122, 124
rodent localities on British mainland tab. 12
Romania 92, 95-6, 109, I14, 116
Romano-British 71
root vole see Muicvotus oeconomus
Rowberrow Cavern 45, 70-1, 82, 88, 93,
tab. 12
Russia 76, 87, 95-8, 106, 116

Saale 42, 50

Saalian 41

sabre-toothed cat 62; see Homotherium

Saint-Prest, Chartres 79

St Vallier, France 81

Salisbury 45

Salwarpe river 72

Sanford, W. A. 37

Savin collection 97

Scandinavia 92, 112-3 ; see Sweden, &c.

Scarborough 45

Scarcliffe 45 ; see Langwith Cave

Scilly Isles 45, 73, 118, tab. 12, 128; see

Nornour

Sciuridae 73-7

Sciurus 73-4; see squirrel
vulgaris 67, 70-1, 74, tab. 12, 129
whiter 48, 73-4, tab. 12, 122, 124

hungaricus 74

Scotland 73, 80, 128; see localities
caves in 67-8, 118

Sealy, D.L. F. 130

Selsey 45, 69, 80, tab. 12

Settle 45

Seven Kings Station 55-6, 58

Severn river 72

Shell Cottages, Thorpe, Aldringham 96 (foot-
note)
Shelly Crag (East Runton) 38, 41, 47
shrew, white-toothed 64-5, 126; see Croci-
dura
Siberia 96-8, 105
Sieveking, G. de G. 60
Siliceous Group (Westbury) 60-1, 124
silver fir 56
Simons, J. 58, 66, 130
singer Dr Rk: 140
sizewell 43° 45, 40%, 78, 95,97, tab: ¥2
Sligo, Co. 45 ; see Keshcorran Caves
snow vole see Microtus nivalis
solifluxion deposits 59-60
Somerset 45; see localities
Caves 37
Levels see Huntspill Cut
Sorex runtonensis 101
savint 101
Soviet Union see Russia
Spain 87
Spencer, H. E. P. 96 (footnote), 130
Spermophilus 48, 54, 75, 76
altaicus 75-6
citillus 76
eversmanni 75-6
evythrogenoides 37, 75-7
evythrogenys 75
Sp. 75, 128
(Colobotis) superciliosus 70-2, 75, 76-7,
tab. £2, 127
(Urocitellus) parryi 76
primigenius 54, 75-6, 77, tab. 12, 125-7
undulatus 76
squirrels 48, 74; see Sciurus
Staffordshire 45 ; see Elder Bush Cave
Stalagmite (Kent’s Cavern) 62
Staple Howe 45, 72, 80, tab. 12
Star Carr 45, 73, 80, tab. 12, 128
Steinheim an der Murr 124
Stenocranius 120; see Microtus gregalis
steppe lemming see Lemmus lemmus
Stoke Tunnel Beds 45, 72, 103, tab. 12
Stuart, A; jj: 30, 60; 69) 72; 8%, 935 “102-4,
LLG, Li, 126
studies, history of 37-40
stutton 45, 60, 72, SI, 1o1=—3, 110, tab: 12,
125
Sudmer Berg 2 Io1
Suffolk 42-3, 45; see localities
Sugworth 45, 69, 98, tab. 12
Siissenborn 101
Sussex 45; see West Wittering, Selsey

146 INDEX

Sutcliffe, Mrs A. J. 130

Sutherland 45; see Creag nan Uamh Cave

Sutton 43, 45, 46*, 79, tab. 12

Swaffley 45

Swanscombe 36, 38-9, 42, 45, 49*, 50, 51-2,
53,59) 78-82, 83, 19078 02—3,) noe—z,
108-11, 115-6, tab. 12, 123-4

Swanton Morley 45, 69, 72, 81, 93, I0I-3,
tab. 12, 126

Sweden 78, 83

Switzerland 87, 89, 118, 120

Symond’s Yat 45

Talpa europaea 101
minor 101
Taplow terrace 55
Tarko 61, 74, IOI
terrace deposits 41-2, 48-60, tab. 12; see
Thames
Thames 48-60, 49*, 62
High” Terrace 38; 45, 49), 51=35 see

Swanscombe

Lower 51-8

Middle Terraces 36, 38, 41, 49, 55, 57, 125
early 53
interglacial see Middle Terrace of

Thames interglacial

late 54, 125 ;

Middle Terrace/Floodplain Terrace prob-
lem 55-8

Third Terrace 38, 41, 54, 55

Upper 58

Upper Floodplain Terrace 55-8

Thatcham 45, 73, 80, 104, tab. 12

Thorpe, Aldeburgh, Suffolk 96 (footnote)

Thorpe, Norfolk 43, 45, 46*, 78, 96 (incl.
footnote), tab. 12

Thorpeness 96 (footnote)

Thurnian 41, 43

Tiglian 41, 43, 78, 83, 96-8, 122

tooth-enamel 81, 100

Torbryan 45

Tornewton Cave 36, 45, 62-5, 66-7, 81,
84-90, 92-3, IOI-4, IIO-I, 115, I17,
119-21, tab. 12, 125-6

Torquay 45

Trafalgar Square 55-8, 127

Treacher collection 58, 113

Trimingham 43, 45, 46*, 97, tab. 12

Trogontherium boisvilletti 43, 47-8, 52-3, 69,
78-9, tab. 12, 122-4

cuviert 37, 53, 78-9

minus 37, 43, 78, 122

sp. 78
Turner, C. 54

Ukraine 74, 76, 114

Uphall Estate 109

Uphill Cave 45, 70-1, 91, 119, tab. 12

Upnor 45, 49*, 54, 80, tab. 12

Upper Freshwater Bed (West Runton) 38,
41, 48, 52, 73, 81, 83, 85, 93, 97-101,
105, 107-8, I14, 116, 124

Upper Langwith 45

Upper Rodent Stratum (Tornewton Cave)
64-6

Uppony Io1

Upton Warren 45, 69, 72, 88, tab. 12, 128

interstadial complex 128

Urals 83

Uvocitellus 125; see Spermophilus primi-
genius

Ursus see bear
deningeri 61

Val d’Arno, Italy 80-1, 96

Vejlby I and II mild stages (Denmark) 40, 42

Villafranchian 41, 43, 47, 76, 81, 92, 95-6,
n22

Viscount Melville Sound 92

Voigtstedt Io1

voles 38-9, 47, 51, 6©9;~>-see Avvutcola,
Mimomys, Microtus

in Ireland ? 69, 129

Waalian 41, 43

Waechter, J. d:A. 51; 130

Wales, south 113

Waltonian 41

Warthe 42

Warton Crag 45; see Dog Holes Cave

Water Hall Farm (Pit) 45, 48, 49*, 58-9, 110,
TTS) gh17, tabs 125 126,130

Waterford, Co. 45; see localities

water-rat, water-vole see Avvicola

Weichsel 42

Weichselian 41

West, R. G. 47, 130

West Runton 38, 41, 45, 46*, 47-8, 52, 60-1,
73, 78-9, 81, 83, 85, 93, 97-100, 105-8,
I14, 116, tab. 12, 122-4

West Thurrock 50

INDEX 147

West Wittering 45, 93, tab. 12

Westbury-sub-Mendip (fissure) 36, 45, 60-1,
80-1, 87, 89-90, 92-3, 95, IOI-2, 105,
107-8, 121, tab. 12, 122-4, 129

Westbury stage 108, tab. 12, 122, 123

Wetton 45

Weybourne Crag 38, 41, 43, 46*, 47, 95-7;
see Icenian Crag

Whitchurch 45

White Sea 120

Whitechurch 45

Willment’s Gravel Pit 58, 128, 130

Wilson, Mrs G. 130

Wiltshire 45 ; see Fisherton

Wintringham 45

wolf 62-3

Wolff, R. G. 69, 81, 90, 104, 130

AS. Surcuirre, Pu.D.

Department of Palaeontology

BritisH Museum (NaTuraL History)
CROMWELL RoapD

Lonpon SW7 5BD

Wolstonian stage 40-2, 50, 52, 54-5, 124-5
wolverine 126; see glutton

wood mouse see Apodemus sylvaticus
Woodbridge 43, 45, 46*, 78-9, tab. 12
Worcestershire 45 ; see localities

Wretton 91

Wiirm 42, 76, 121

Wye Cave _ see Merlin’s Cave

Xenocyon lycaonoides 61

yellow-necked mouse see Apodemus flavi-
collis

Yorkshire 45 ; see localities

Yugoslavia 98, 106, 109, 114

K. KOWALSKI

Institute of Systematic and
Experimental Zoology

PoLisH ACADEMY OF SCIENCES

KRAKOW

POLAND

Accepted for publication 28 October 1975

‘
® ie
nid
>
7a
\ x uae
=f 4 S

t
=== od
an
,
at
oF i
7
f ;
i
be
an
7 agit, a
‘ rim oy aN
pee!
: ic :
if i"
7
f .
~
ij
i
‘ f :
Me
4 4
‘ i

ae
Paling)

. ae

Io.

TE.

A LIST OF SUPPLEMENTS
TO THE GEOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

~ Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya.

Pp..213; 30 Plates; 2 Text-figures. 1965. OUT OF PRINT.

. ELt-NaGcGar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper

Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley,
Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. 11.

. DAvEy, R. J., Downie, C., SARJEANT, W. A. S. & WitiiaMs, G. L. Studies on

Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 Plates; 64 Text-
figures. 1966. £8.20.

. APPENDIX. DAVEY, R. J., DOWNIE, C., SARJEANT, W. A. S. & WILLIAMS, G. L.

Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 24.
1969. 95P.

. E..iotr, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of

the Middle East. Pp. 111; 24 Plates; 16 Text-figures. 1968. OUT OF PRINT.

. Rwopes, F. H. T., Austin, R. L. & Druce, E. C. British Avonian (Carboni-

ferous) Conodont faunas, and their value in local and continental correlation.
Pp. 313; 31 Plates; 92 Text-figures. 1969. £13.10.

. Cuitps, A. Upper Jurassic Rhynchonellid Brachiopods from Northwestern

Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £5.25.

. Goopy, P. C. The relationships of certain Upper Cretaceous Teleosts with

special reference to the Myctophoids. Pp. 255; 102 Text-figures. 1969.
£7.70.

. OwEN, H.G. Middle Albian Stratigraphy in the Anglo-Paris Basin. Pp. 164;

3 Plates; 52 Text-figures. 1971. £7.20.

. Sippigul, Q. A. Early Tertiary Ostracoda of the family Trachyleberididae

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. £9.60.
Forey, P.L. A revision of the elopiform fishes, fossil and Recent. Pp. 222;
92 Text-figures. 1973. £11.35.

Witiiams, A. Ordovician Brachiopoda from the Shelve District, Shropshire.
Pp. 163; 28 Plates; 11 Text-figures; 110 Tables. 1974. £12.80.

Printed in Great Britain by fohn Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 5NU

SEN), &f

A MONOGRAPH ON FOSSIL BEES‘. ws”
(HYMENOPTERA: APOIDEA)

Pr EE ZEUNER
AND

F. J. MANNING

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
GEOLOGY Vol. 27 No. 3
LONDON: 1076

A MONOGRAPH ON FOSSIL BEES
(HYMENOPTERA : APOIDEA)

BY
THE LATE FREDERICK EVERARD ZEUNER
AND
THE LATE FRANCIS JOSEPH MANNING

EDITED WITH AN APPENDIX
BY

SAMUEL FRANCIS MORRIS

British Museum (Natural History)

Pp. 149-268 ; 4 Plates

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
GEOLOGY Vol. 27 No. 3
LONDON : 1976

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

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

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

This paper is Vol. 27, No. 3, of the Geological
(Palaeontological) series. The abbreviated titles of
periodicals cited follow those of the World List of
Scientific Periodicals.

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

ISSN 0007-1471

© Trustees of the British Museum (Natural History), 1976

BRITISH MUSEUM (NATURAL HISTORY)

Issued 23 September, 1976 | Price £8.80

A MONOGRAPH ON FOSSIL BEES
(HYMENOPTERA : APOIDEA)

By THE LATE F. E. ZEUNER AND THE LATE F. J. MANNING

CONTENTS

Page
SYNOPSIS : : ‘ ; : : : , : ‘ 155
I. INTRODUCTION i ; : : : ; : 155
II. TERMINOLOGY AND CLASSIFICATION : : : ; ‘ ‘ 156
III. MATERIAL STUDIED 3 : 158

IV. SYSTEMATIC PALAEONTOLOGY: SyNoNomIc LIST OF Fossit BEES
(EXCLUDING APIDAE) . ‘ 2 : ‘ : ‘ 161
Superfamily APOIDEA Seeeead ‘ : : ; ‘ : 161
Family CoLLETIDAE Bingham , : 3 : : : 161
Family HaLicTipAE Ashmead ; : a : ; : 162
Subfamily HarictinaE Ashmead : , ‘ 2 ‘ 162
Genus Cyrtapis Cockerell . : ; ; : : 163
Cyrtapis anomalus Cockerell : : : : 163
Genus Halictus Latreille . : ; ; ‘ 163
Halictus ruissatelensis Timon- David : : : 164
Halictus florissantellus Cockerell . : ; : 164
Halictus miocenicus Cockerell , : ; . 164
Halictus scudderiellus Cockerell . : 5 : 164
Halictus schemppi (Armbruster) . : : ; 164
Family ANDRENIDAE Latreille é : : : : ; 165
Subfamily ANDRENINAE Latreille . ; ; ‘ : 165
Genus Andrena Fabricius . : : : : é 165
Andrena wrisleyi Salt . : F , ; ‘ 166
Andrena clavula Cockerell . ‘ : : : 166
Andvena grandipes Cockerell : : : : 166
Andrena hypolitha Cockerell . : : 5 166
Andrena lagopus Latreille . : ‘ : : 167
Andrena percontusa Cockerell 3 ‘ : : 167
Andrena sepulta Cockerell . : : : : 107
Andvena primaeva Cockerell . é : ; ‘ 107
Andrena spp. indet. . : : ; . : 168
Genus Lithandrena Cockerell ‘ : : ; : 169
Lithandrena saxorum Cockerell d ; : : 169
Genus Pelandrena Cockerell : : F F : 169
Pelandrena reducta Cockerell : ‘ : ; 170
Subfamily PANURGINAE Leach . : : ‘ : ; 170
Genus Libellulapis Cockerell : : : : : 170
Libellulapis antiquorum Cockerell . : : , 170
Libellulapis wilmattae Cockerell . : z A gt
Family Andrenidae incertae sedis : ; ; : : 171
Andrenidae gen. et sp. indet. . : ; : 171
Family MELITTipAE Dumeril : : : - : : 171
Subfamily CTENOPLECTRINAE Cockerell : : : 5 172
Genus Ctenoplectrella Cockerell . ; : : ; E72
Ctenoplectrella dentata Salt . ; : : ; E72

Ctenoplectrella vividiceps Cockerell . : : ‘ £73

T5

MONOGRAPH ON

Genus Glypiapis Cockerell .
Glyptapis fuscula Cockerell .
Glyptapis mirabilis Cockerell
Glyptapis neglecta Salt
Glyptapis reducta Cockerell .
Glyptapis reticulata Cockerell

Subfamily MELITTINAE Dumeril
Genus Melitta Kirby . ,
Melitta willardi Cockerell
Family Melittidae incertae sedis.
Dasypoda (s.1.) sp. indet.
Family MEGACHILIDAE Latreille
Subfamily LITHURGINAE Newman

Genus Lithurge Latreille

Lithurge adamitica (Heer)
Subfamily MEGACHILINAE Latreille
Tribe ANTHIDIINI Michener

Genus Anthidium Fabricius
Anthidium mortuum (Meunier)
Anthidium exhumatum Cockerell
Anthidium scudderi Cockerell

Genus Dianthidium Cockerell

Dianthidium tertiavium Cockerell .

Genus Lithanthidium Cockerell

Lithanthidium pertriste Cockerell
Tribe MEGACHILINI Latreille .

Genus Heriades Spinola. :
Heriades bowditchi Cockerell .
Heriades halictinus Cockerell
Heriades laminarum Cockerell
Heriades mersatus Cockerell .
Heriades mildvedae Cockerell
Heriades priscus Cockerell
Heriades saxosus Cockerell

Genus Megachile Latreille .
Megachile amaguensis Cockerell
Megachile praedicta Cockerell
Megachile sp. indet.

Genus Osmia Panzer .

Osmia carbonum Heyden
Osmia antiqua Heer
Osmia nigra sp. nov.
Osmia sp. indet. .

Family Megachilidae incertae sedis
‘Apiaria’ dubia Germar
Megachilidae gen. et sp. indet.

Family ANTHOPHORIDAE Dahlbom

Subfamily XyLOcoPINAE Latreille

Tribe CERATININI Latreille

Genus Cevatina Latreille

Ceratina disrvupta Cockerell .
Tribe XyLocopPini Latreille
Genus Xylocopa Latreille

Page
173
173
173
174
174
174
174
LD
175
175
175
175
176
176
176
178
178
178
178
179
179
179
180
180
180
180
180
181
181
181
181
I8I
182
182
182
182
183
183
183
184
184
185
186
186
186
186
187
187
188
188
188
188
189

FOSSIL BEES

Xylocopa friesei Statz .
Xylocopa hydrobiae Zeuner
Xylocopa jurinei (Heer)
Xylocopa senilis Heer .
Subfamily ANTHOPHORINAE Dahlbom
Tribe EucERINI Latreille
Genus Tetvalonia Spinola P
Tetralonia berlandi Théobald
Tribe ANTHOPHORINI Dahlbom
Genus Anthophora Latreille
Anthophora melfordi Cockerell
Genus Anthophorites Heer .
Anthophorites gaudryi Oustalet
Anthophorites thovacica Heer
Anthophorites longaeva Heer .
Anthophorites mellona Heer .
Anthophorites titania Heer
A nthophorites tonsa Heer
Anthophorites veterana Heer .
Tribe MELECTINI Westwood .
Genus Promelecta Cockerell.
Promelecta brevipennis Cockerell
Family Anthophoridae incertae sedis
‘Anthophora’ effossa Heyden.
Anthophoridae gen. et sp. indet.

Superfamily Apoidea (except Apidae) incertae sedis.

Apoidea gen. et sp. indet.

Fossils erroneously classified as Apoidea ore Apidae)

‘Formicinae’
Polistes kirbyanus Cockrel.
Larval Chambers : :
‘Genus’ Celliforma Brown :
Celliforma favosites Brown
Celliforma spirifer Brown
Celliforma germanica Brown .
Celliforma nuda Brown
Celliforma bedfordi sp. nov. .
Celliforma septata sp. nov.
Celliforma sp. indet.
‘Genus’ Uruguay Roselli ‘
Uruguay auroranormai Roselli
SYSTEMATIC DESCRIPTIONS OF FOSSIL APIDAE
Family APIDAE Latreille
Chalcobombus group
Genus Chalcobombus Gockel
Chalcobombus hirsutus Cockerell
Chalcobombus humilis Cockerell

Chalcobombus martialis Cockerell .

Tribe BomBInI Latreille
Genus Bombus Latreille
Bombus crassipes Novak
Bombus abavus Heer

Bombus florissantensis (Cockerell) .

Page
189
189
189
190
192
192
193
193
193
194
194
194
195
195
195
196
197
198
198
199
199
199
200
200
200
201
201
202
202
202
203
203
203
203
204
204
204
205
205
205
205
206
206
206
200
207
208
208
209
209
209
210
212

153

154 MONOGRAPH ON

Bombus proavus Cockerell

Tribe MELIPONINI Handlirsch
Genus Sophrobombus Cockerell
Sophrobombus fatalis Cockerell
Genus Trigona Jurine ;
Subgenus Trigona Jurine
Trigona (Tvigona) erythra Eenicttenes
Subgenus Hypotrigona Cockerell
Trigona (Hypotrigona) gribodot Magretti
Subgenus Tetvagona Lepeletier & Serville .
Trigona (Tetragona) succini (Tosi)
Trigona (Tetragona) ivridipennis Smith
Subgenus indet.
Trigona sicula (Tosi)

Tribe APINI Latreille :
Genus Electvapis Cockerell .

Subgenus Electvapis Cockerell . 5
Electrapis (Electvapis) apoides enn Tine
Electrapis (Electvapis) meliponoides (accel eepen)
Electvapis (Electvapis) tornquisti Cockerell ,

Subgenus Protobombus Cockerell ;
Electvapis (Protobombus) indecisus (Cockerell) .
Electrapis (Protobombus) tristellus (Cockerell) .

Subgenus Roussyana Manning . ;
Electrvapis (Roussyana) palmnickenensis (Roussy)
Electrapis (Roussyana) proava (Menge)

Genus Apis Linnaeus
Subgenus Synapis Cockerell

A pis (Synapis) cuenoti Théobald

A pis (Synapis) henshawi Cockerell
Apis (Synapis) henshawi dormiens subsp. nov.
Apis (Synapis) henshawi henshawi Cockerell
Apis (Synapis) henshawi kaschket (Statz)

Subgenus Apis Linnaeus.

Apis (Apis) armbrustert Zeuner :
Apis (Apis) avmbrusterit avmbrustert Zeuner.
Apis (Apis) armbrusteri scharmanni (Armbruster)
Apis (Apis) armbrusteri scheert (Armbruster)
Apis (Apis) armbrusteri scheuthlet (Armbruster)

Apis (Apis) melisuga (Handlirsch) 5

Apis (Apis) mellifera Linnaeus

Family Apidae incertae sedis :

‘Apis’ aquitaniensis de Rilly

‘Apis’ dormitans Heyden

‘Apis’ styriaca Pongracz

‘Bombus’ carbonarius Menge.

‘Bombus’ muscorum Roussy .

‘Bombus’ pusillus Menge

‘Bombus’ antiquus Heyden

‘Bombus’ grandaevus Heer ‘

‘Bombusotdes’ menget Motschulsky

Apidae gen. et sp. indet.

FOSSIL BEES 155

Page

Not APIDAE , : ‘ ‘ : 255
Lithoblatta Beeeiae (Germar) : : 5 : 255

VI. APPENDIX, by S. F. Morris : : : j : 255
Colletidae gen. et sp. iciee : ‘ : 255

Ctenoplectrella splendens ioe seule’ ‘ z 255

Anthiidini gen. et sp. indet. : : : 5 256

Osmia sp. . ‘ ‘ , ; 256

Apoidea (? Anehophertnae) Sp. ; A : : 256

Probombus hirsutus Piton (ms) . : 256

Trigona (Hypotrigona) eocenica Kelner Pilate ‘ 256

Trigona (Hypotrigona) dominicana Wille & Chandler 250

Trigona (Nogueivapis) silacea Wille ; ‘ F 257

Electrapis minuta Kelner-Pillault 5 : ; 257

Apis catanensis Roussy ‘ : : Z : 257

? Apoidea sp... ; § : : ; : 257

Not Apoidea : , : ; 5 : ‘ 257

VII. ACKNOWLEDGEMENTS. é i : : : : : 257
VIII. REFERENCES. : ; ; : ; : 5 : ‘ 258
xX. INDEX... ; : é : : 5 5 : ‘ : 204

SYNOPSIS

The status of 128 species of fossil bees and nests are reviewed, 19 of which are left under open
nomenclature. Four species are removed from the Apoidea. A new species Osmia nigra
sp. nov. and subspecies Apis (Synapis) henshawi dormiens subsp. nov. of bees, and two new
‘species’ of bees’ nests Celliforma bedfordi sp. nov. and Celliforma septata sp. nov., are described.
An appendix covering the years 1960—74 has been added.

i. INTRODUCTION
by S. F. Morris

UNTIL 1949, Professor Zeuner and Dr Manning had each been independently working
on monographs of fossil bees, and until that time they were unaware of each other’s
work. But in 1949 they met and agreed to combine their efforts into a joint
monograph. This collaboration continued until the death of Professor Zeuner in
1963, followed by the death of Dr Manning in 1966.

The major part of the present work was already finished by 1959, but the authors
seemed to have had problems which prevented them completing it. The monograph
lacks a projected part which was to have dealt with the evolution of the bees, but
there are no extant manuscript notes for this. Certainly in 1964 Dr Manning was
investigating a sphecid wasp from the Jurassic of Lerida Province, Spain, which he
thought might be (or be closely related to) the ancestor of the bees. No written
material of this work is extant either.

Entries to the main body of the monograph ceased about 1959, so the editor has
added an Appendix (p. 255) containing fossil bee references since that date, and a
few earlier ones that the authors had missed which have since come to his notice.
A few minor alterations to the body of the text have been made in order to bring it
into line with modern taxonomic work.

156 MONOGRAPH ON

Although the monograph is intended to be a world-wide catalogue of fossil bees,
the authors did not have the opportunity to re-examine the American species, so
that the taxonomic aspects of these are not treated in as great a depth as are the
European ones. In the main the authors have had to rely on the original authors’
descriptions and figures for the American species except where the specimens were
deposited in a European institution.

Since this work was carried out some years ago, references to various institutions
may be badly out-of-date, especially the war-damaged European museums. It has
not, however, been the principle of the editor to undertake any significant alteration
to the typescript as it stood. The taxonomic work as published is in two parts,
the Apoidea excluding the Apidae (Section IV), and the Apidae (Section V). The
division appears to have been made in this way because it most nearly represents
the division between the New World and Old World fossil bees, and allows for
different treatments of the two parts.

The textfigures and plates had been made but have disappeared and were not
amongst the effects of either of the authors. Fortunately the British Museum
(Natural History) still retained negatives of many of the photographs and these
have been reprinted for the plates. No attempt has been made to redraw the missing
textfigures, since it is not known what they were or what they were intended to show.
New photographs have been taken of specimens that are deposited in the British
Museum (Natural History).

Professor Zeuner was a Research Associate of the British Museum (Natural
History), which position he had held since 1934 until his death. His obituary was
published in 1963 in Nature, Lond. 200: 1263. An obituary for Dr Manning
appeared in 1966 in Proc. R. ent. Soc. Lond. (C) 31: 62.

Il. TERMINOLOGY AND CLASSIBPICATION

The study of fossil bees, like that of any other group, requires a reasonably satis-
factory system of classification. It may not always express fundamental relation-
ships, but it must be consistent with the morphological evidence available. Michener
(1944, 1965) attempted a revision of the families, subfamilies and tribes of bees of
the world and of the North American genera in particular — a formidable task. From
its very magnitude it was clear that the result would not completely satisfy the
author himself, nor could it entirely agree with all the considered opinions of all bee
taxonomists. His groups, however, well express relationships that have for long
been obscure, and the inclusion of the bee families in the Sphecoidea has much to
recommend itself. For these reasons, Michener’s classification of the bees has been
adopted in the present monograph.

In the past, the lack of uniformity of the use of a terminology for their morphology
has been a serious difficulty in the study of bees. Fortunately Snodgrass (1935)
and more recently Michener (particularly 1944) have devoted considerable time and
effort to clarifying this unhappy state of affairs, and we have taken advantage of
the results of the labours of these authors. Moreover, the often excellent preser-
vation of the fossil forms makes it necessary to treat them as if they were Recent

POSSliE BEES 157

specimens, and to make detailed reference to morphological characters. This is
particularly true, for instance, of specimens from Rott which show such intimate
features aS wax mirrors and parts of the alimentary canal. A modern terminology,
therefore, is imperative. The terms used in this monograph for the structural
features of the hind legs are defined later.

With regard to wing venation the system of Ross (1936) has been adopted, together
with the use of the following terms: submarginal cells, submarginal cross veins
(i.e. 2nd abscissa Rs, 1st r-m, 2nd r-m) and recurrent veins (Ist m-cu and 2nd m-cu).
The term ‘intercubitus’ has become obsolete with the rejection of the use of the
term ‘cubital cells’ for the submarginal cells. The very convenient term ‘basal
vein’, however, is retained for the free (i.e. not fused) sections of both M and Rs,
which run from opposite directions towards each other before they unite to form a
vein Rs+M.

In addition, the term ‘bomboid’ is frequently used in the discussions that follow.
This is not synonymous with ‘Bombus-like’ but is intended to suggest that the
characters in question are merely reminiscent of those of the tribe Bombini. Bom-
boid forms need not be closely related to Recent Bombini, though perhaps derived
from either the ancestral form of the Bombini or, just as probably, from the ancestral
form from which both the Bombini and the other Apidae arose. The last-mentioned
possibility might imply that the Apini and Meliponini are not descended from the
Bombini. The term ‘apoid’ is used in a somewhat similar sense in its appropriate
context.

Other terms, important because of their evolutionary significance, are those
defining certain structures of the hind legs. The most essential are the following :
(x) tibial comb, (2) tibial rake, or pollen rake, (3) tibial spur, (4) auricle, (5) depstum,
and (6) basitarsal brush.

The TIBIAL COMB! is a structure much like a comb, situated on the outer side of
the apex of the tibia, above the tibio-basitarsal joint. It consists of strong hairs,
fused at the base, which usually project backwards and downwards.

The TIBIAL RAKE is the. dense row of spines or spine-like hairs along the apical
edge of the tuner side of the tibia.

The TIBIAL SPURS are the two large and sharp spines which are jointed to the
tibia. They protrude from the lower apical margin of the hind tibia. In the normal
walking position, with the hind legs slightly turned outwards, they have to be looked
for underneath in a seemingly forward position. If two spurs are present they are
distinguished as the inner and outer hind tibial spurs. The inner one is often serrate.
If only one is present, it may be either the outer or the inner, and one can be dis-
tinguished from the other by its position relative to the apical rim of the tibia and
by the presence or absence of serration, sometimes expanded into a comb-like
structure. Thus a single spur with one edge serrate, occupying a relatively deep,

1 The honey bee and the bumble bee lack this comb, but nevertheless have one or two hairs situated
on the outer apical face of the corbicula. The derivation of these hairs is not known. Their position,
however, precludes them from being direct derivatives of the tibial comb. They have not been noticed
in fossil Apinae but might easily have been overlooked. They are used for stabilizing the load of
pollen.

\

158 MONOGRAPH ON

inner position, is assumed to be the inner tibial spur, and a straight spur without
serration, occupying a more marginal position, is assumed to be the outer.

The AURICLE is the more or less broadened upper tip of the basitarsus which lies
below the distal end of the tibia. It is like a projecting lip neatly fitting the com-
plementary shape of the tibial apex, and in Recent Afyvs it is set with minute stud-
like eminences. Often both the lip and tibial apex have been jointly termed the
auricle, a practice not followed here.

DEPSTUM is a term used for an angularity of the basitarsus which is sometimes
present where no true auricle is developed. This is a primitive condition, although
some bees (e.g. Meliponini) do not even have a depstum in the modern species.

Lastly, the term BASITARSAL BRUSH is reserved for the neatly aligned rows of
bristles present on the inner surface of the basitarsus of the Apini. It is not used for
the haphazard arrangement of hairs found in other groups.

Lit. MATERIAL SiuUDIED

For the study of the fossil Apidae a fair amount of material was available in the
British Museum (Natural History), the total number of specimens being 52. The
Meliponini, for instance, lacked only the specimens from the Sicilian amber which
had been well described and illustrated by Tosi as long ago as 1896. The Apini
lacked only representatives of the bees from Randecker Maar, mainly in the Arm-
bruster Collection, but these were examined by one of the authors. It was only
in the Bombini that the Museum collection really lacked material, due to the
extreme paucity of the fossils of this tribe. Fortunately the most important
representative, Bombus proavus Cockerell, has been well photographed and described
(1931).

Furthermore, during 1950 and again in 1951, both authors travelled independently
on the European continent studying fossil Apidae. Collections in Holland, Germany
and Switzerland were examined. One of us had carried on this study of continental
material for a number of years before it was decided to continue the work jointly.

Whilst fossil Apoidea are plentiful in Europe, the absence of material belonging to
families of bees other than the Apidae is disappointing. This material is chiefly in
American museums and must await discussion by someone from there. Our
attempts to obtain information or specimens from the relevant museums remained,
unfortunately, without significant success. Below are given the names of the more
important collections of fossil Apidae or museums where such material is to be
found.

ARMBRUSTER COLLECTION. In the sediments of a small volcanic lake at Randeck,
Wiurttemberg, honey bees have been found by numerous collectors. Possibly the
oldest known specimens are those in the Oscar Fraas Collection, Stuttgart. In
1926, however, William Scheuthle of Goeppingen began his search for fossil honey
bees, and just over a year and a half later was assisted by Professor L. Armbruster.
When the former died in November 1928, substantial discoveries had already been
made and the collection eventually passed entirely into the possession of Professor

AO SINE, INTIS) 159

Armbruster. Other collections believed to contain Randecker Maar honey bees
are those of Eduard Scheer of Goeppingen, Bernard Hauff of Holzmaden, Karl
Schempp of Brucken, and Pfarrer Hermann of Holzmaden.

BAUCKHORN COLLECTION. See Siegburg an der Lahn (p. 160).

BursEY COLLECTION. The collection of amber fossils formerly belonging to Mr
Maurice Bursey of Surbiton, Surrey, now in the Muséum National d’Histoire
Naturelle, Paris, contains the important specimen FE. (Electrapis) apoides Manning

(p: 227).

COCKERELL COLLECTION. The late Professor T. D. A. Cockerell of the University
of Colorado collected and studied fossil bees. His main collections were made from
the Miocene lake deposits of Florissant, Colorado, particularly during the expedition
of 1906-7, when he was accompanied by his wife, Dr W. M. Wheeler and Mr S. A.
Rohwer. His collections of fossil bees appear to have been few, and are now in
the University of Colorado Museum, Boulder, Colorado, and the American Museum
of Natural History. Although more of his other fossil insects are in the British
Museum (Natural History), the only bee from his collection to be found there is
Anthophora melford: Cockerell.

DanziGc, Poland: Westpreussisches Provinzial-Museum. Many Baltic amber
specimens were to be found here, but the whereabouts of the collection is not known
at the present time.

KARLSRUHE i1.B., Germany: Badische Landessammlung fir Naturkunde,
Erbprinzenstrasse 13. Contains some of the Oeningen and Radoboj material
described by Oswald Heer. The building itself is in ruins, but the cellars are still
packed with material and it is possible, therefore, that some further types, at present
untraceable, will be found.

KLEBS COLLECTION. From time to time Baltic amber specimens were formerly
sold to private collectors and to museums, and it appears that this dispersal was
made by the firm of Messrs Stantien & Becker, who traded the amber and who later
were taken over by the Prussian State Amber Works. Messrs Stantien & Becker,
in their day, placed the amber fossils in charge of Richard Klebs (1850-1911) who
was, in later years, state geologist to the Prussian Geological Survey. A set of 346
specimens, including the holotype of EF. (Roussyana) proava Menge (p. 236), was
acquired by the British Museum (Natural History) from Messrs Stantien & Becker
in 1892, but since the specimens bear labels marked ‘R. Klebs, Museum Stantien &
Becker’, this collection has usually gone under the name of ‘Klebs Collection’.
(See also p. 238.)

KOENIGSBERG, East Prussia, U.S.S.R.: Geologisch-Palaontologisches Institut
und Museum der Universitat. The most important types and described specimens
of Baltic amber bees were included in this collection. They are now housed in the
Museum of the Humboldt University, Berlin.

Matnz (Rhein), Germany : Naturhistorisches Museum der Stadt Mainz. Contains
a large collection of insect remains from the Hydrobia-limestone of the district around
Mainz.

160 MONOGRAPH ON

Lonpon, British Museum (Natural History). See under Klebs Collection,
Swinhoe, R. C. J., Luke Thomas Collection and Krantz, F. This collection includes
some 52 specimens, including the types of Heyden (1862), and others from the
Baltic amber, Rott am Siebengebirge, Florissant, Béttingen, East African copal,
Burmese copal and South Australia.

MARSEILLE, France: Musée. According to Armbruster (1938) and Roussy (per-
sonal communication), the specimen of ‘Apis’ aquitaniensis de Rilly (No. 5979)
(p. 250) is in this collection, and Meunier (1915) states that a specimen of Antho-
phorites mellona Heer (p. 196) is also in it. No confirmation, however, can be ob-
tained after repeated enquiries. The Marseille Museum is rich in fossil insects
from the Tertiary of southern France.

Roussy CoLLecTION. This is the private collection of Monsieur Louis Roussy,
Aigle, Switzerland, and contains mainly Baltic amber specimens, including Electrapis
(Roussyana) palmmickenensts (Roussy), p. 233.

SCHEELE COLLECTION. This important collection of several thousand specimens
of Baltic amber has been purchased by the Geologisches Staatsinstitut, Hamburg.
Several important specimens of Electrapis are included.

SIEGBURG AN DER LAHN, Germany: Stadtisches Heimatmuseum. This museum
contains the valuable collection of fossil insects made by Hugo Bauckhorn. It is
particularly rich in specimens from Rott, previously studied by Meunier and
Seaez.

STATZ COLLECTION. Fossil insects from Rott am Siebengebirge have been
collected for over a century and are to be found in many museums and institutions.
The collection made by the late Georg Statz of Cologne has become famous for the
number and quality of its specimens. Moreover, Statz described and figured large
numbers of his specimens. His collection includes several dozen fossil bees some of
which have been well described and play an important role in the reconstruction of
the phylogeny of the group. It is much to be regretted that this collection has been
removed to Algiers, where it is in the hands of the collector’s daughter.

STUTTGART, Germany : Wirttembergische Naturaliensammlung. This collection
includes the material from Boettingen Swabian Alb, a locality situated not far from
Randeck. One of us, while studying the fauna of this area, discovered a fossil
swarm of honey bees, and casts were made of some of the bodies comprising it.
Unfortunately it is believed that the original specimens were destroyed during the
war, but the casts are still preserved in the British Museum (Natural History).

SWINHOE COLLECTION. Specimens of insect inclusions in the dark and pale
varieties of Burmese amber from the Hukong Valley were collected by R. C. J.
Swinhoe and presented to Professor T. D. A. Cockerell. Some were later presented
by Cockerell to the British Museum (Natural History). Only the pale Burmese
amber is known to contain bees.

Timon-Davip CoLLeEcTIon, Marseille. This collection consists of fossil insects
from Camoins (Bassin de Marseille) and other French localities, and includes the
fossil bee Halictus ruissatelensis Timon-David (p. 164).

FOSSIL BEES 161

LuKE THoMAS COLLECTION. A number of specimens of fossil bees are from East
African copal. These, and other insects, have been presented to the British Museum
(Natural History) at various times by individual donors. The Luke Thomas
Collection contains twelve pieces of copal including fossil bees belonging to the
Meliponini. They were presented to the British Museum (Natural History) in
1945 by Col. H. Burrows. All are from the east coast of the mainland of Africa,
facing Zanzibar.

WasHINGTON, D.C., U.S.A.: United States National Museum; Smithsonian

Institution. Contains part of the collection, mainly from Florissant, described by
Professor T. D. A. Cockerell.

WICKHAM COLLECTION. The late Professor H. F. Wickham collected material
from the Miocene shales of Florissant, Colorado, particularly at Wilson Ranch.
This collection, which contains Andrena percontusa Cockerell (p. 167), is in Yale
University.

ZURICH, Switzerland: Geologisches Institut und Museum der Eidgendéssischen

Technischen Hochschule. This collection includes much of the material, described
by Oswald Heer, from Oeningen.

Pie > YSteMATIC PALAEONTOLOGY: SYNONYMIC LIST OF FOSSIL BEES
(EXCLUDING APIDAE)

The following list is arranged according to Michener (1944, 1965). His phylo-
genetic groups and main diagnostic characters for families, relevant subfamilies and
tribes have been quoted, even though it is sometimes doubtful whether they are
applicable to fossil forms. Indeed, only completely inapplicable characters have
been omitted.

In compiling the list a considerable number of specimens have been examined.
In the case of unexamined material, no diagnoses of genera or species or systematic
descriptions are given. These must await a study of the fossil material in American
collections. Among such material, too, those specimens whose affinities have been
stated in a definite way by earlier authors, but concerning which one might hold
other views, have sometimes been accorded the benefit of the doubt, and noted
under the generic name assigned by the most trustworthy author. The sections
headed ‘incertae sedis’ therefore do not include specimens which are believed to be of
such a nature that more information about them may substantiate their present
suggested status. It is reserved for those which are too poorly preserved to be
classifiable, and for vague records which cannot be substantiated by specimens.

The publication containing the first valid name in accordance with the Inter-
national Rules of Zoological Nomenclature has been marked with an asterisk (*).

Superfamily APOIDEA Ashmead 1899
Family COLLETIDAE Bingham 1897

Diacnosis. Labrum broader than long; subantennal areas absent or at least
reduced to small triangular spaces ; lower sides of clypeus not bent parallel to long

162 MONOGRAPH ON

axis of body; subantennal sutures directed toward inner margins of antennal
sockets ; facial foveae often present ; galeae short post-palpally and usually short
pre-palpally, glossa short and often truncate or bifid, rarely round; labial palpi
short, segments similar but sometimes the first somewhat elongated and broadened ;
pre-episternal sutures usually complete ; metanotum usually horizontal; middle
coxae, as seen from the outside, much shorter than the distance from their summits
to posterior wing bases ; pygidial plate present or absent.

Many of the genera are also distinguished from practically all other bees in
having the posterior part of vein 2nd m-cu (2nd recurrent) of each forewing arcuate
towards the apex of the wing ; the basal vein is not strongly curved.

REMARKS. No fossil Colletidae have so far been found,? though Cyrtapis, a
genus of Halictidae, is according to Cockerell (1908a) in some respects reminiscent
of the Colletidae.

Family HALICTIDAE Ashmead 1899

DiaGnosis. Labrum broader than long, but in some females its apex is produced
to form a broad, acutely pointed process ; subantennal areas absent, though there
is usually a single subantennal suture directed towards inner margins of antennal
sockets ; lower sides of clypeus not bent back parallel to long axis of body ; facial
foveae ordinarily absent ; galeae elongated pre-palpally, not post-palpally, the pre-
palpal section tapering evenly to an acute base ; labial palpi with segments similar
to each other, and usually cylindrical, rarely are any of the first three elongate and
flattened. Glossa acute, usually short, without flabellum ; pre-episternal sutures
usually complete ; metanotum horizontal ; middle coxae with exposed parts much
shorter than distance from summits to posterior wing bases ; basitibial plate usually
present in females, except in parasitic genera ; pygidial plate present in females.

Basal vein usually strongly curved ; jugal lobe of posterior wings much more than
half as long as anal lobe.

The most distinctive features are the long pre-palpal parts of the galeae, the
usually strong arcuation or angulation of the 1st abscissa of basal vein (vein M), and
the normally complete pre-episternal suture.

DISTRIBUTION. Oligocene to Recent.

REMARKS. Apart from the genus Cyrtapis, which has colletid affinities, only true
Halictinae are known as fossils. By Oligocene times they appear to have been very
modern in wing venation and probably in their general appearance also.

Subfamily HALICTINAE Ashmead 1899

DIAGNOSIS. Supraclypeal area convex and protuberant if seen in profile ; clypeus
itself longer than labrum and not protuberant much beyond level of supraclypeal
area ; labrum in females commonly provided with a broad, pointed apical process ;

2 But see Appendix, p. 255.

FOSSIL BEES 163

antennal sockets not, or only a little, below middle of face ; pre-episternal suture
conspicuous and complete ; scopa, when present, usually very extensive and found
on posterior trochanters, femora, tibiae and basitarsi; prepygidial fimbria usually
divided by a longitudinal line without hairs or with only appressed hairs.

First abscissa of basal vein (vein M) strongly arcuate and first submarginal cell
ordinarily longer than third when, as is usually the case, three are present.

Sometimes the glossa is elongated and variously modified ; the labial palpi are
not elongated.

Remarks. This large subfamily is only very moderately represented by fossils.

Genus CYRTAPIS Cockerell 1908

*1908a Cyrtapis Cockerell : 339.
1909b Cyrtapis Cockerell ; Cockerell : 80.

GENOTYPE. Cyrtapis anomalus Cockerell by monotypy.

REMARKS. According to Cockerell (1908a), ‘the general appearance and the large
stigma, marginal cell, and bent basal nervure are all as in the Halictines. The form
of the second submarginal cell, with the first r.n. joining it before the middle, as well
as the second r.n. with the double curve, suggest affinity with the Colletines’.
Cyriapis may well have retained primitive elements, but because its evolution is
towards the Halictines it is better to look upon it, for the time being at least, as a
representative of the latter. Special mention of the type was made by Cockerell

(1909b).

Cyrtapis anomalus Cockerell 1908

*1908a Cyrtapis anomalus Cockerell : 339.
1909b C. anomalus Cockerell ; Cockerell : 80.
1931 Cyrtapis anomalus Cockerell; Salt: 145.

Hoiotype. Cockerell Collection, whereabouts unknown.
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Mouth parts not known. In the circumstances, and because vein
2nd m-cu (2nd recurrent) has a strong double bend, the species is considered to be a
little anomalous.

Genus HALICTUS Latreille 1804

*1804 Halictus Latreille : 182.
1935b Halictus Latreille ; Richards: 170.
1938 Prohalictus Armbruster : 48, fig. 74 (type species: P. schemppi Armb.).
GENOTYPE. Apis quadricincta Fabricius, by subsequent designation of Richards

(1935b : 170).

164 MONOGRAPH ON

Halictus ruissatelensis Timon-David 1944

1943 Sphecodes Timon-David : 256.
*1944 Halictus ruissatelensis Timon-David : 42.

Ho.totyrPe. Timon-David Collection, Marseille.
DISTRIBUTION. Oligocene: Bassin de Marseille.

REMARKS. Originally included by Timon-David (1943) in Sphecodes, but later
(1944) placed, more correctly, in Halictus. We have not examined the specimen.

Halictus florissantellus Cockerell 1906

*1906 Halictus florissantellus Cockerell : 43.
1907 Halictus florisantellus Cockerell ; Handlirsch : 889.

HoLotypPe. Mus. Comp. Zool. Harvard, No. 2010 (= Scudder Coll. No. 921).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. The name ‘flortsantellus’ (Handlirsch 1907) is obviously a printer’s
error. Cockerell (1906) writes that only part of the wing venation is preserved, but
‘generic reference seems safe’. Approximately 6-5 mm long.

Halictus miocenicus Cockerell 1909

*t909a Halictus miocenicus Cockerell : 160.
1931 Halictus miocenicus Cockerell ; Salt: 146.

HoLotypPe. Cockerell Collection, whereabouts unknown.
DISTRIBUTION. Oligocene: Florissant.
REMARKS. According to Cockerell, the generic position is probably correct.

Halictus scudderiellus Cockerell 1906

*t906 Halictus scudderiellus Cockerell : 43.
1907. Halictus Scudderiellus Cockerell ; Handlirsch : 889.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2011 (= Scudder Coll. No. 1966).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Appears to be correctly placed. This bee is smaller than H. floris-
santellus, being only 4-5 mm long.

Halictus schemppi (Armbruster 1938)
*1938 Pyrohalictus schemppi Armbruster : 48, fig. 74.

HoLotypPe. Schempp Collection, Stuttgart-Weil.
PARATYPES. Three specimens, Armbruster Collection.
DISTRIBUTION. Miocene: Randeck.

FOSSIL BEES 165

ReMARKS. Armbruster refers to one of his specimens as a true Halictus, stating
(1938 : 48) that ‘the nervulus is positioned quite normally’. There is no reason for
believing the others are not conspecific.

Family ANDRENIDAE Latreille 1802

Diacnosis. Labrum broader than long, subantennal areas defined by two sub-
antennal sutures beneath each antennal socket ; anterior lateral margins of clypeus
usually concave ; lower sides of clypeus bent backwards more or less parallel to
long axis of body (except in many species of Andrena) but clypeus not strongly
protuberant ; facial foveae usually present ; galeae short pre-palpally and usually
so post-palpally ; labial palpi short, the segments similar or first segment elongated
and flattened, very rarely first two segments elongated ; glossa short or long, acute,
without a flabellum ; pre-episternal sutures absent or present below scrobal sutures ;
metanotum horizontal ; middle coxae short as seen externally, far shorter than
distance from summits to posterior wing bases. Basitibial and pygidial plates
present in females.

Jugal lobe of posterior wing at least nearly three-quarters as long as anal lobe.

The most characteristic feature of this family is the presence of defined sub-
antennal areas. Unfortunately, in some species they le below the bases of the
antennae, covered with hairs, and even when uncovered only weak sutures may be
revealed.

DISTRIBUTION. Eocene to Recent.

REMARKS. Only the subfamilies Andreninae and Panurginae are represented
among the fossils. Andvena was well characterized as a genus by the Eocene.

Subfamily ANDRENINAE Latreille 1802

Diacnosis. Facial foveae large and usually covered with minute hairs giving a
golden or reddish brown sheen in females, but in males foveae absent or virtually
undefined and not clothed with minute hairs ; first flagellar segment rarely a little
more than half as long as scape, usually much shorter; maxillary palpi six-
segmented ; segments of labial palpi cylindrical, similar to one another; glossa
usually short, acute ; pre-episternal sutures absent below scrobal sutures ; scopa
present from coxae to basitarsi of posterior legs.

Pterostigma moderate to large in size; marginal cell longer than distance from
apex to wing tip, and with the apex itself pointed on the costal margin of the wing
or very narrowly rounded.

REMARKS. Most fossil members of this subfamily belong to the important genus
Andrena, and two have been placed in the extinct genera Lithandrena and Pelandrena.

Genus ANDRENA Fabricius 1775

*1775 Andrvena Fabricius : 376.
1873 Buiareolina Dours : 288 (type species : B. neglecta Dours).
1912 Andvena Fabricius; Viereck : 613.

13

166 MONOGRAPH ON

GENOTYPE. Apis helvola Linnaeus 1758, by subsequent designation of Viereck
Gor2ichm3).
DISTRIBUTION. Eocene to Recent.

Andrena wrisleyi Salt 1931
*1931 Andvena wrisleyi Salt : 141, fig. I.
HoLotypPe. Formerly in the Geol. Museum, Univ. Koenigsberg, but now in the
Humboldt Museum, Berlin.
DISTRIBUTION. Eocene: Baltic amber.
REMARKS. Well figured and described.

Andrena clavula Cockerell 1906

*1906 Andrena (?) clavula Cockerell : 45.
1907. Andrena (?) clavula Cockerell ; Handlirsch : 889.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2014 (= Scudder Coll. No. 6963).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Considered by Cockerell (1906) to be congeneric with A. sepulta.
He remarked, however, that ‘the shape of the abdomen is like that of a 9 Ceratina, or
possibly certain Halictines, but the venation does not agree with these’.

Andrena grandipes Cockerell 1911

*to11a Andyrena grandipes Cockerell : 73.
1931 Andrena grandipes Cockerell; Salt: 145.
1938 Andrena grandipes Cockerell; Rodeck : 295.

HototyPe. Univ. Colorado Mus., No. 8599 (with counterpart).
DISTRIBUTION. Oligocene: Florissant.

ReMARKS. Cockerell (Ig11a) states ‘agrees closely with modern Andrvena in
most respects, but is peculiar for the very broad hind tibia and basitarsus, the small
eyes and the long slender marginal cell’.

Andrena hypolitha Cockerell 1908

*1908b Andrena hypolitha Cockerell : 328.
1931 Andrena hypolitha Cockerell; Salt: 145.
1938 Andrena hypolitha Cockerell; Rodeck : 295.

HoLotypPe. Univ. Colorado Mus., No. 18736.
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Differs from A. sepulta and A. clavula in various details of venation
etc. It is not clear, however, how the abdomen differs from A. clavula.

FOSSIL BEES 167

Andrena lagopus Latreille 1809

*1809q <Andrena lagopus Latreille : 151.
1873 Biareolina neglecta Dours : 288.
1913a Biareolina neglecta Dours ; Cockerell : 346.
1930 Andrena lagopus Latreille ; Schmiedeknecht : 942.

HoLotype. Recent, whereabouts unknown.
FOSSIL SPECIMEN. Possibly in the United States National Museum.
DISTRIBUTION. Oligocene: probably Florissant. Also Recent.

REMARKS. Schmiedeknecht, in identifying his specimen with a Recent species,
mentions but one character, namely that the side of the stigma on the marginal cell
‘is conspicuously angulate or subdentate about the middle’.

Biareolina Dours is a synonym of Andrena Fabricius, but since it possesses two
submarginal cells, it is sometimes accorded separate generic status, a practice not
followed by Michener (1944). Moreover, the reduction of the number of sub-
marginal cells from three to two occurs frequently within the same species, for
instance in Andvena, Halictus, Nomada and others.

Andrena percontusa Cockerell 1914

*1914 Andyrena percontusa Cockerell : 640.
1931 Andrena percontusa Cockerell; Salt: 145.

Horotyre. Wickham Collection, ? Peabody Museum, Yale.
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Venation like A. sepultaand A. clavula, butasmallspecies. Abdomen
differs in shape from A. clavula. According to Cockerell (1914), the species can be
distinguished from A. grandifes and A. hypolitha by the venation.

Andrena sepulta Cockerell 1906

*1906 Andrena sepulta Cockerell : 44.
1907. Andvrena sepulta Cockerell ; Handlirsch : 889.

HoLotyPeE. Mus. Comp. Zool. Harvard, No. 2013 (= Scudder Coll. No. 14288).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Cockerell (1906) states ‘the venation is not exactly like that of any
modern species with which I have compared it but the differences are unimportant’.

Andrena primaeva Cockerell 1909
(Bl<ay figy2)

1865 Drei Osmien; Heer : 386 (partim).

1876 Osmiae; Heer: 43 (partim).

1g909c Three species of Osmia ; Cockerell : 316 (partim).
*1go09c Andrena (?) primaeva Cockerell : 316.

1g09c Osmia primaeva (Heer MS) ; Cockerell : 316.

1931 Andrena (?) primaeva Cockerell ; Salt: 145.

168 MONOGRAPH ON

HoLotyPeE. FEidgendssische Technische Hochschule, Ztirich, No. SN.202, with
counterpart.

DISTRIBUTION. Miocene: Oeningen.

DESCRIPTION AND REMARKS. Although Cockerell’s references to Osmiae are, on
the whole, difficult to interpret, there is no doubt whatever that the specimen he
described as Osmia primaeva was the holotype. The insect is reasonably well
preserved but the head is lacking. The chitin is dark brown in colour which may
have been the colour of the original insect. The abdomen is plump and subglobose,
but, since it is a millimetre broader than the thorax, the junction of the two gives the
impression of a distinct constriction.

The thorax has a large convex mesoscutum and apparently strong tegulae; the
other features are indistinct.

The hind legs are fairly well preserved, the tibia and basitarsus being approxi-
mately equal, their combined length being 3-5 mm ; the tibia is broad, apparently
twice the width of the basitarsus, and both are copiously hairy; the femur is as
broad as the tibia. The tibia and basitarsus of the middle leg are not preserved in a
natural position, and we are inclined to consider Cockerell’s basitarsus as the tibia
and his tibia as the femur. Both femur and tibia are of the same width, but the
tibia is slightly longer.

The wings are not preserved in their entirety owing to a crack in the rock. For-
tunately, what is preserved is very distinct and includes the apical region of the fore
and hind wings.

Forewing: long, well-developed pterostigma ; marginal cell narrowing apically,
and arising broadly from the pterostigma; three submarginal cells, the first not
completely preserved, the second rhomboid, receiving the first recurrent vein slightly
distad of the middle and having its proximal cross vein slightly hooked at the base,
the third submarginal cell at least twice as large as the second, only half as broad on
the marginal cell as its own base, and receiving the second recurrent vein in its last
third.

Hindwing: radial sector and media continued to wing margin and cross-vein r-m
only very slightly oblique.

Cockerell’s (1909) statement that ‘all things considered the reference to Andrena
seems reasonably assured’ appears a fair conclusion. The venation of the hind wing
and the rather plump body precludes the specimen from being referred to Halictus.

MEASUREMENTS. Length of thorax and abdomen (head lacking) 7mm; length
of thorax 2:25 mm, width 3 mm; length of abdomen 4:75 mm, width 4 mm.

Andrena sp. indet. (Brischke)

1886 Andrena Brischke : 278.

1891 Andvena Brischke ; Scudder : 686 (5915).
1907 Andvena Brischke ; Handlirsch : 890.
1931 Andvena sp. Brischke ; Salt: 144.

DISTRIBUTION. Eocene: Baltic amber.

FPOSSIE BEES 169

REMARKS. Among Menge’s and Helm’s collections, investigated by Brischke,
were apparently two specimens generically determined as Andrena, but not described.

Andrena sp. indet. (Motschulsky)

1856 <Andvena Motschulsky : 28.
1931 Andyrena sp. Motschulsky ; Salt: 144.

DISTRIBUTION. Eocene: Baltic amber.
REMARKS. Motschulsky merely records the occurrence of the genus.

Genus LITHANDRENA Cockerell 1906

*t906 © Lithandrena Cockerell : 44.
1909b Lithandrena Cockerell ; Cockerell : 81.

GENOTYPE. Lithandrena saxorum Cockerell 1906 by monotypy.

REMARKS. According to Cockerell, ‘a genus of Andrenidae, allied to Andrena.
It differs from Andvena and Nomza in the second r.n. which is strongly bent in its
upper part, straight but oblique below ; from Andrena alone it differs in the propor-
tions of the submarginal cells ; and from Nomza also by the tip of the marginal cell,
which is pointed, and a little away from the costae. The general appearance is that
of an Andrena, but it cannot be referred to this or any other genus...’. Special
mention of the type was made by Cockerell (1g09b).

Lithandrena saxorum Cockerell 1906

*1906 Lithandrena saxorum Cockerell : 44.
1907 Lithandrena saxorum Cockerell ; Handlirsch : 889.
1909b Lithandrena antiquorum Cockerell ; Cockerell : 80.

HoLotyrPe. Mus. Comp. Zool. Harvard, No. 2012 (= Scudder Coll. No. 82109).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Length 86mm. Venation of forewing fully described (Cockerell
1906). Head as wide as thorax, black ; flagellum stout ; abdomen light in colour
with a broad entire dark band on each segment, legs hairy.

Genus PELANDRENA Cockerell 1909

*1g09a Pelandrena Cockerell : 159.
1909b Pelandvena Cockerell ; Cockerell : 81.

GENOTYPE. Pelandrena reducta Cockerell 1909a by monotypy.

REMARKS. Allied to Andvena but with two submarginal cells on the anterior wing.
Special mention of the type was made by Cockerell (1909b). The reason for the
erection of this genus is not clear. It may be based only on the two submarginal
cells, in which case its generic status is weak.

170 MONOGRAPH ON

Pelandrena reducta Cockerell 1909

*tgo9ga Pelandrena veducta Cockerell : 160.
1909b Pelandvrena veducta Cockerell ; Cockerell: 81.
1931 Pelandrena reducta Cockerell; Salt: 146.
1938 Pelandvrena veducta Cockerell ; Rodeck : 298.

HoLotyPe. Univ. Colorado Mus., No. 18741.
DISTRIBUTION. Oligocene: Florissant.

Subfamily PANURGINAE Leach 1815

Diacnosis. Facial foveae, when present, much smaller than in Andreninae, not
clothed with fine hairs, and frequently distinct in males as well as in females ; labial
palpi frequently with first segment much elongated ; maxillary palpi usually six-
segmented (reduced in some species of Perdita). Scopa primarily confined to pos-
terior tibiae and basitarsi.

Pterostigma usually rather large; marginal cell, which is normally not longer
than distance from its own apex to apex of wing, mostly obliquely truncate apically
but sometimes merely bent away from wing margin.

REMARKS. Only one fossil genus belongs to this subfamily.

Genus LIBELLULAPIS Cockerell 1906

*1906 =6Libellulapis Cockerell : 42.
1909b Libellulapis Cockerell ; Cockerell : 80.

GENOTYPE. Libellulapis antiquorum Cockerell 1906 by monotypy.

REMARKS. Cockerell (1913a), when discussing Libellulapis wilmattae, stated, ‘A
reconsideration of the genus leads me to place it in the Panurgidae where, by reason
of certain features of the venation, it appears to stand rather near to Panurgus’.
Eyes curiously prominent. Special mention of the type was made by Cockerell

(1909b).

Libellulapis antiquorum Cockerell 1906

*1906 = Libellulapis antiquorum Cockerell : 42.
1907 Libellulapis antiquorum Cockerell ; Handlirsch : 889.
1909a Libellulapis antiquorum Cockerell ; Cockerell : 80.
1938 Luibellulapis antiquorum Cockerell; Rodeck : 297.

HoLotypPe. Mus. Comp. Zool. Harvard, No. 2009 (= Scudder Coll. No. go06r).
PARATYPE. Univ. Colorado Mus., No. 18740 (= Scudder Coll. No. 8560).
DISTRIBUTION. Oligocene: Florissant.

ReMARKS. Cockerell (1906) considered this species different from Halictoides
maurus (Cresson), Hesperapis rhodoceratus (Cockerell) and Parandrena andrenotdes
Cresson in points of venation, and believed that it did not agree with Diandrena or
Biareolina.

FOSSIL BEES 17/1

Libellulapis wilmattae Cockerell 1913

*to13a Libellulapis wilmattae Cockerell : 344, figs 2, 3.
1931 Libellulapis wilmattae Cockerell; Salt : 146.
1938 Libellulapis wilmatiae Cockerell ; Rodeck : 297.
HototypPe. U.S. Nat. Mus., No. 58688. Counterpart, Univ. Colorado Mus.,
No. 18742.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. Cockerell (1913a) states, ‘Broad head and prominent eyes as in some
Panurgi, pygidial plate as in Dasypoda’.

Family ANDRENIDAE incertae sedis
Andrenidae gen. et sp. indet. (Handlirsch)

1907. (? Andrenidae) ; Handlirsch : 890.
1931 Andrenidae ? sp. (Handlirsch) ; Salt: 144.

MATERIAL. Hofmuseum, Vienna.
DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Hind leg with pollen. Head formation suggestive of Halictus or
Andrena.

Andrenidae gen. et sp. indet. (Scudder)

1881 Andrenidae ; Scudder : 290.

1883 Andrenidae ; Scudder : 280.

1891 (Andrenidae) (several) ; Scudder : 682 (5872).

1907 (Andrenidae) (mehrere) Scudder ; Handlirsch : 890.

DISTRIBUTION. Oligocene: Florissant.
REMARKS. Specimens appear to be poorly preserved.

Family MELITTIDAE Dumeril 1821

Diacnosis. Labrum broader than long ; subantennal areas absent ; subantennal
sutures directed, except sometimes at their upper ends, towards inner margins of
antennal sockets ; lower sides of clypeus not, or only a little, bent backwards ;
clypeus not protuberant. Facial foveae absent; galeae short pre-palpally and
usually so post-palpally ; labial palpi with segments similar to one another, cylin-
drical; glossa short or long, acute; pre-episternal and scrobal sutures usually
absent, the former always absent below the latter ; metanotum horizontal or sub-
horizontal ; middle coxae short as seen externally, much shorter than distance
from their summits to posterior wing bases ; scopa confined to posterior tibiae and
basitarsi, although trochanters and femora are hairy ; basitibial plates and pygidial
plate present in female.

DISTRIBUTION. Eocene to Recent.

172 MONOGRAPH ON

REMARKS. The subfamilies Ctenoplectrinae and Melittinae only are represented
among the fossils.

Subfamily CTENOPLECTRINAE Cockerell 1920

DIAGNOSIS. Scrobal and upper parts of pre-episternal sutures are present,
though of doubtful value for fossil forms ; inner hind tibial spur much broadened
basally and throughout its length provided with a comb of very many long fine
teeth ; posterior basitarsi of female nearly as broad as tibiae; pygidial plate of
male absent, that of female narrow apically, greatly broadened basally.

Pterostigma hardly twice as long as broad ; marginal cell considerably longer than
distance from its apex to wing tip and bent gradually away from wing margin for
nearly half length of cell ; two submarginal cells ; jugal lobe of posterior wing about
half as long as anal lobe.

REMARKS. Among Recent forms this subfamily contains only the genus Cteno-
plectra found from Australia and Africa to China. This genus, however, is not yet
known in the fossil state. On the other hand, the subfamily is represented by two
fossil genera Ctenoplectrella and Glyptapis. Moreover, Cockerell (19g09d) maintained
that the genera included in this subfamily are ‘so remote from the modern members
of that group that they at least form a distinct subfamily, Glyptapinae’. ‘That
group’, however, refers to the Megachilidae to which Ctenoplectra, and hence
Ctenoplectrella etc., was considered to belong. Therefore until the affinities of the
Ctenoplectra group, both fossil and Recent, have been more clearly defined, it is not
considered advisable to create a new subfamily.

Genus CTENOPLECTRELLA Cockerell 1909

*1go9c Ctenoplectrella Cockerell : 314.
1909d Ctenoplectrella Cockerell ; Cockerell : 19.

GENOTYPE. Ctenoplectrella viridiceps Cockerell 1909 by monotypy.

REMARKS. This genus appears to be closely related to Glypiapis. Both genera
have Ctenoplectra as ‘their nearest relative in the modern fauna’. C. viridiceps was
first mentioned by Cockerell (1909c) with some characters, but its full description
followed in a later paper (909d). [See also Appendix, p. 255.]

Ctenoplectrella dentata Salt 1931
*1931 Ctenoplectrella dentata Salt : 139.

HoLotyre. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Salt considers it possible that this.specimen is the male of Cteno-
plectrella viridiceps Cockerell and gives a good description and figures.

HOSS TE BEES 173

Ctenoplectrella viridiceps Cockerell 1909

*1go9c Ctenoplectrella viridiceps Cockerell : 314.
1909d Ctenoplectrella viridiceps Cockerell ; Cockerell : 19, fig. 14.
1931 Ctenoplectrella viridiceps Cockerell ; Salt: 145.

HoLotype. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.
DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Female, see above. This species was a small, black, stout bee, like
Glyptapis, hardly 5mm long. “The eyes were bare and the sting visible’. Full
description given by Cockerell (1909d).

Genus GLYPTAPIS Cockerell 1909

*1909c Glyptapis Cockerell : 314.
1909d Glyptapis Cockerell ; Cockerell : 13.
GENOTYPE. Glyptapis mirabilis Cockerell, by original designation.

REMARKS. This genus, together with Ctenoplectrella, is referred by Cockerell ‘to
a group from which the Megachiloids sprang’. The genus was mentioned (1909c)
in connection with G. mirabilis, but it was not described until later (1909d).

Glyptapis fuscula Cockerell 1909

*1900d Glyptapis fuscula Cockerell : 17, fig. 12.
1931 Glyptapis fuscula Cockerell; Salt : 145.

HoLotyrPe. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Cockerell (1909d) mentions that the ventral side of the abdomen is
clearly seen to be furnished with strong bristles, the beginning of a ventral scopa.
As distinctive characters of the species he gives the very dark wings and the large
stigma.

Glyptapis mirabilis Cockerell 1909

*1909c Glyptapis mirabilis Cockerell : 314.
1909d Glyptapis mirabilis Cockerell ; Cockerell : 14, figs 9, Io.
1931 Glyptapis mivabilis Cockerell; Salt : 146.
HoLotyPe. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.
DISTRIBUTION. Eocene: Baltic amber.
REMARKS. The genotype, selected by Cockerell (1g909d), is actually the monotype,
since the genus and species were first mentioned (1g09c) together, with a few

174 MONOGRAPH ON

descriptive words. The specific characters are the dark brown stigma and veins,
the length of the marginal cell (over 1-2 mm), the entirely black body and the punctate
mesothorax.

Glyptapis neglecta Salt 1931
*1931 Glyptapis neglecta Salt : 136, fig. 1.
HoLotypPe. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.
DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Well described and figured by Salt. This is the largest known
Glyptapis from the Baltic amber.

Glyptapis reducta Cockerell 1909

*1900d Glyptapis reducta Cockerell : 18, fig. 13.
1931 Glyptapis veducta Cockerell ; Salt : 146.

HoLotypPe. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Wings hyaline. Cockerell (1909d) regards the ferruginous stigma
and veins and the length of the marginal cell (0-935 mm) as diagnostic.

Glyptapis reticulata Cockerell 1909

*19009d Glyptapis reticulata Cockerell : 16, fig. 11.
1931 Glyptapis reticulata Cockerell ; Salt: 146.
1969 Glyptapis reticula Cockerell ; Kellner-Pillault : 521.

HototyPe. Formerly in the Geological Museum, University of Koenigsberg, now
in the Humboldt Museum, Berlin.

DISTRIBUTION. Eocene: Baltic amber.

RemMARKS. Differs from G. mirabilis chiefly in the metallic crimson tints on the
abdomen and the reticulate mesothorax.

Subfamily MELITTINAE Dumeril 1821

Diacnosis. Inner hind tibial spur unmodified ; posterior basitarsi more slender
than tibiae (except in male Haplomelitia) and shorter than tibiae ; distitarsi unusually
large.

Pterostigma slender, more than twice as long as broad, its margins basad of vein
are very rarely converging basally ; marginal cell longer than distance from its own
apex to wing tip, the apex itself very narrowly rounded on wing margin and bent
away from the latter, if at all, only apically ; three submarginal cells, second much

FOSSIL BEES 175

the smallest, the first being the largest ; jugal lobe of posterior wing nearly half as
long as anal lobe (rarely three-quarters).

The species of this subfamily are of moderate size, and black. The abdomen is
provided with pale pubescent fasciae, giving the appearance of certain species of
Halictus and Andrena.

REMARKS. This subfamily is represented, among fossils, only by the genus
Melitta.

Genus MELITTA Kirby 1802

*1802 Melitta Kirby : 130-134.
1935b Melitta Kirby ; Richards: 172.

GENOTYPE. Melitta tricincta Kirby 1802, selected by Richards (1935 : 172).

Melitta willardi Cockerell 1909

*1909e Melitta willardi Cockerell : 393.
1931 Melitta willavdi Cockerell; Salt: 146.
1938 Melitta willardi Cockerell ; Rodeck : 297.

HoLotyPe. Univ. Colorado Mus., No. 18737.
DISTRIBUTION. Oligocene: Florissant.
REMARKS. Cockerell (1g0g9e) writes, “Venation normal for Melitta, except that

the upper segment of the basal nervure is shorter’; also ‘the reference of this insect
to Melitta seems safe’.

Family MELITTIDAE incertae sedis
Dasypoda (s.1.) sp. indet. (Menge)

1856 Dasypoda Menge : 26.

1891 Dasypoda (vic.) Menge ; Scudder : 696 (6005).
1907 Dasypoda (vic.) Menge ; Handlirsch : 890.
1931 Dasypoda vic. Menge; Salt: 144.

MATERIAL. Whereabouts unknown.
DISTRIBUTION. Eocene: Baltic amber.

REMARKS. An ‘Andrenida, Grabwespen’ was reported (Menge 1856) to resemble
Dasypoda but ‘was distinguished from it by the hind basitarsus having a blunt
lateral process in front of the joint with the second tarsal segment, and the hind leg
being longer than the abdomen’ (transl.). The mandibles had a broad tip with
three teeth and the lancet-shaped tongue had labial palps protruding. Two speci-
mens existed, one clearly visible, the other covered with ‘schimmel’.

Family MEGACHILIDAE Latreille 1802

Diacnosis. Labrum longer than broad, widened basally to form a long line of
articulation with clypeus ; subantennal areas absent ; subantennal sutures directed

176 MONOGRAPH ON

towards outer margins of antennal sockets ; lower sides of clypeus not bent back-
wards parallel to long axis of body ; facial foveae absent ; first flagellar segment
much shorter than scape ; galeae short pre-palpally, long post-palpally ; labial palpi
with first two segments elongated and flattened, sheath-like, first segment usually
shorter than second; glossa linear, possessing a flabellum, pre-episternal and
scrobal sutures considerably reduced or absent; metanotum usually vertical,
sometimes horizontal ; middle coxae elongate, at least half, and usually more than
half, as long as distance from summits to hindwing bases ; basitibial plates absent ;
scopa, when present, confined to abdominal sterna; pygidial plate absent, except
in Lithurge.

Pterostigma small; submarginal cells two, the second as long, or nearly as long,
as the first ; marginal cell pointed or narrowly rounded at apex, which is usually a
little bent away from costal margin of wing and nearer to wing tip than to base of
cell.

The most distinguishing characters are the position of the scopa, the form of the
labrum and the direction taken by the subantennal sutures.

DISTRIBUTION. Eocene to Recent.

REMARKS. The subfamilies Lithurginae and Megachilinae are known as fossils,
the latter being well represented in the Oligocene.

Subfamily LITHURGINAE Newman 1834

DiaGnosis. Face of female elevated below antennal sockets ; epistomal suture
in most part absent ; mandibles in both sexes short and robust, tridentate, the
middle tooth longer and more elevated than the others; metanotum about one-
eighth as long as scutellum ; coarse spiculae of tibiae arranged in two longitudinal
rows on fore and middle legs, scattered on posterior legs ; posterior basitarsi slender,
parallel-sided, about as long as posterior tibiae ; second abdominal tergum unusually
small, posterior margin broadly rounded ; terga two to seven of male exposed and
unmodified ; male with pygidial plate, female with slender remnant thereof in the
form of large spine, flattened dorsally.

Jugal lobe of posterior wing about three-quarters as long as anal lobe.

REMARKS. Only one genus, Lithurge, is known to which both Recent and fossil
forms belong.

Genus LITHURGE Latreille 1825

*1825 Lithurge Latreille : 463.
1827 Lithurgus Berthold : 467.

GENOTYPE. Centris cornuta Fabricius 1787 by monotypy.

Lithurge adamitica (Heer 1865)

*1865 Apis adamitica Heer : 386, fig. 287.
1867 Apis adamitica Heer; Heer: 4, pl. 3, fig. 11.

FOSSIL BEES 177

1872 Apis adamitica Heer; Heer: 473, fig. 287.
1876 Apis adamitica Heer; Heer: 43, fig. 287.
1879 Apis adamitica Heer; Heer: 412, fig. 320.
1891 Apis adamitica Heer ; Scudder : 688 (5936).
1907 Apis adamitica Heer ; Handlirsch : 892.
1g0g9c Lithurgus adamiticus (Heer) ; Cockerell : 313.
1915 Apis adamitica Heer; v. Buttel-Reepen : 9, fig. 1.
1928 Apis adamitica Heer; Wheeler: 97.

1928 Lithurgus Berthold ; Wheeler : 97.

1931 Apis adamitica Heer; Statz: 45.

1938 ‘Apis adamitica’ (Heer) ; Armbruster : 86.

HoLotyPe. Eidgenéssische Technische Hochschule, Ztirich, V.S.4.

OTHER SPECIMENS. (1) Eidgendssische Technische Hochschule, Ztirich, SN.200 ;
(2) Badische Landessammlung fir Naturkunde, Karlsruhe, Inv.1216 (Oe.95).

DISTRIBUTION. Miocene: Oeningen.

DESCRIPTION AND REMARKS. Cockerell (I909c) considered this species to be
somewhat more primitive than recent Lithurge, possibly a near subgenus. According
to him it is separated from Afzs by its possession of only two submarginal cells.
Unfortunately the holotype, which we have studied, is not well preserved and we
have not succeeded in recognizing the submarginal cells. Armbruster (1938) who
contributed some original remarks on it did not consider it to belong to A fis.

The holotype, indeed, is a somewhat elongated insect with a slender and pointed
abdomen. The outlines of the wings are preserved but the venation is mostly too
indistinct for satisfactory recognition. It is clear, however, that the species does
not belong to the genus Afvs, and also that the preservation of the specimen appears
to have deteriorated since it was examined by Cockerell.

The following points are presented as still being of interest. The head and thorax
have preserved no recognizable diagnostic characters ; the colour of the body has
darkened from Cockerell’s ‘warm, red-brown’ ; the abdomen, as Cockerell stated, is
‘truncate basally, pointed apically’ and, since basal intersegmental membranes are
evident, may be somewhat distended ; legs are lacking ; wings are small in compari-
son with size of insect, and wing venation shows only the following points. Strong
costal vein ; straight and short basal vein, the anterior section of which is longest ;
short, stubby, pointed marginal cell; poorly developed pterostigma. The apex of
the wing, marginal cells and recurrent nerves are no longer clearly evident ; the
cross vein cu-a, however, can be detected slightly distal of the juncture of basal
vein on cu. While these characters obviously do not give sufficient grounds for
placing the specimen generically, we must recognize that Cockerell saw it while in a
better state of preservation and accept his decision in this matter.

A second specimen (Ztirich, SN.200) has been studied and tentatively assigned to
this species. Its head, thorax and abdomen (except the tip) are quite well preserved,
but the wings are present only in traces and the legs are missing. A strong costal
vein is present, and a marginal cell similar to Lithurge.

A third specimen (Karlsruhe, Inv.1216 (Oe.g5)) may also belong here. Abdomen
and thorax alone are preserved. The former, however, is well segmented and
pointed, and is very reminiscent of the holotype, from which it differs only slightly

178 MONOGRAPH ON

in size, i.e. length of abdomen 10 mm, as compared with 8-5 mm for the holotype.
The width of the abdomen is 4:25 mm and the overall length, as preserved, i.e.
without head, is 13 mm.

Finally, it is interesting to note that Heer’s figure (1865 : fig. 287) differs in wing
venation from his pl. 3, fig. 11 (1867), although both are presumably meant to repre-
sent the same specimen.

Subfamily MEGACHILINAE Latreille 1802

DiaGnosis. Face of female very rarely elevated below antennal sockets ; epi-
stomal suture complete ; mandibles variously toothed, apical tooth longest, preapical
tooth not elevated ; tibiae not spiculate; posterior basitarsi much shorter than
tibiae ; second abdominal tergum large, with the posterior margin usually straight,
so that base of abdomen lacks the compressed aspect of Lithurge ; abdomen of male
with fewer unmodified sterna than in Lithurge ; pygidial area entirely absent.

Jugal lobe of posterior wings less than half as long as anal lobe.

REMARKS. Two tribes, Anthidiini and Megachilini, are distinguished within this
subfamily, both known as fossils.

Tribe ANTHIDIINI Michener 1944

DiaGnosis. Posterior margin of scutellum usually vertical or even overhanging,
at a distinct angle to the dorsal surface ; integumental maculation, at least in part,
pale ; claws of female cleft or at least with an inner tooth.

Inner margin of pterostigma from its base to the base of medial vein little, if at
all, longer than width of pterostigma.

REMARKS. Two Recent genera are represented by fossils, and there is one extinct
genus. [See also Appendix, p. 256.]

Genus ANTHIDIUM Fabricius 1805

*1805 Anthidium Fabricius : 364.
1810 Anthidium Fabricius ; Latreille : 439.

GENOTYPE. Afis manicata Linnaeus 1758, selected by Latreille (1810 : 439).

Anthidium mortuum (Meunier 1920)

*1920 Eucera mortua Meunier : 731, pl. I, fig. 5, text-fig. 4.
1931 Eucera mortua Meunier ; Salt: 145.
1936 Anthidium mortuum (Meunier) ; Statz: 260, 285, 293, 294, pl. 13, fig. 38.

HoLotypPe. Heimatmuseum Siegberg (Lahn), Bauckhorn Collection.
DISTRIBUTION. Oligocene: Rott am Siebengebirge.

REMARKS. This specimen was first described as Eucera by Meunier (1920).
Statz (1936), on re-examining the specimen, placed it in Antiidium. He insisted
that the venation of the fossil form was identical with that of Anthidiwm, except for

FOSSIL BEES 179

the ‘nervulus’ which is ‘interstitial’ in the fossil and ‘postfurcal’ in the Recent species.
Since Statz was the more reliable worker of the two, we are placing this form in
Anthidium. The possibility of a Eucera occurring in Rott should not, however, be
wholly disregarded, since Statz is anxious to reinforce his identification by the
ecological argument that Anthidiwm builds its nest in stalks and twigs, which agrees
better with the Rott biotope than the steppe habitat of Eucera. Statz’ (1936 : 260),
however, states that Recent species of Eucera have the second recurrent vein meeting
the second submarginal cell, whereas in Meunier’s Eucera mortua it ends distad to
the second submarginal cross vein as in the genus Anthidium.

It is also necessary to point out that Statz’ discussion of this form is misleading.
He describes it (1936 : 285) under the heading of ‘new discoveries’, whilst his figure
proves it to be Meunier’s holotype. Three further disconnected paragraphs appear
(1936 : 260, 293, 294) in which no mention is made of the fact that he had no new
specimen. The identity with Meunier’s holotype is proved not only by the general
shape of the fossil, but also by the frayed margin of the right forewing.

Anthidium exhumatum Cockerell 1906

*1906 =Anthidium exhumatum Cockerell : 38.
1907. Anthidium exhumatum Cockerell ; Handlirsch : 888.

Ho.totyre. Mus. Comp. Zool. Harvard, No. 2003 (= Scudder Coll. No. 13709,
with counterpart, No. 11388).

PARATYPE. Scudder Coll. No. 8444.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. According to Cockerell, ‘not so typical an Anthidium as A. scudderv’.
The paratype agrees with the holotype in the width of the marginal cell and the
shape of the second discoidal. It differs in the strongly banded abdomen. He
concludes, ‘It appears to be a female and I have little doubt that it represents that
sex In A. exhumatum’.

Anthidium scudderi Cockerell 1906

*1906 Anthidium scudderi Cockerell : 38.
1907 Anthidium scuddert Cockerell ; Handlirsch : 888.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2002 (= Scudder Coll. No. 11381).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Apparently a typical Anthidiwm which can be clearly separated from
Megachile on points of wing venation.

Genus DIANTHIDIUM Cockerell 1900

*1900 Dianthidium Cockerell : 412.

GENOTYPE. Anthidium curvatum Cockerell (nec Smith) — Dianthidium § sayt
Cockerell, by original designation (1900).

180 MONOGRAPH ON

Dianthidium tertiarium Cockerell 1906

*1906 Dianthidium tertiarium Cockerell : 39.
1907 Dianthidium tertiari'um Cockerell ; Handlirsch : 888.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2004 (= Scudder Coll. No. 806).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. This average-sized (8 mm) dark brown bee can be assumed to be
correctly placed by Cockerell, in view of his intimate knowledge of the genus.

Genus LITHANTHIDIUM Cockerell 1911

*ro11b Lithanthidium Cockerell : 225.

GENOTYPE. Lithanthidium pertriste Cockerell 1911 by monotypy.

Lithanthidium pertriste Cockerell 1911

*ro1ib Lithanthidium pertriste Cockerell : 225.

1931 Lithanthidium pertriste Cockerell; Salt : 146.
1938 Lithanthidium pertriste Cockerell ; Rodeck : 297.
HoLotypPe. Univ. Colorado Mus., No. 18738.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. Only a single specimen known.

Tribe MEGACHILINI Latreille 1802

Diacnosis. Claws of female simple ; no obtrusive pale integumental maculation ;
abdominal pattern often present, resulting from bands of pale hairs ; posterior part
of scutellum, except in some parasitic forms, not separated by a sharp angle from

dorsal surface.
Inner margin of pterostigma from its own base to base of radial vein much longer

than width of pterostigma.
REMARKS. The Recent genera Heriades, Megachile and Osmia are represented by
fossils.

Genus HERIADES Spinola 1808

*1808 Heriades Spinola: 7.
1810 Heviades Spinola; Latreille : 439.

GENOTYPE. Apis truncorum Linnaeus 1758, selected by Latreille (1810 : 439).

REMARKS. According to Spinola’s usage, the name Heriades is feminine. The
masculine ending, however, is used here in conformity with modern usage.

FOSSIL BEES 181

Heriades bowditchi Cockerell 1906

*1906 Heriades bowditchi Cockerell : 41.
1907 Heriades bowditchi Cockerell ; Handlirsch : 889.
1938 Heriades bowditchi Cockerell; Rodeck : 296.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2007 (= Scudder Coll. No. 13761).
PARATYPE. Univ. Colorado Mus., No. 18739 (= Scudder Coll. No. 13436).
DISTRIBUTION. Oligocene: Florissant.

ReMARKS. Differs from H. laminarum in the pronouncedly larger head and the
banded abdomen.

Heriades halictinus Cockerell 1906

*1906 Heriades halictinus Cockerell : 40.
1907 Heriades halictinus Cockerell ; Handlirsch : 889.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2006 (= Scudder Coll. No. 10564).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Resembles a small Halictus in superficial aspect, but its true affinity
is indicated by wing venation. Differs from H. laminarum in its darker abdomen
and smaller size.

Heriades laminarum Cockerell 1906

*1906 Heriades laminarum Cockerell : 40.
1907 Heriades laminarum Cockerell ; Handlirsch : 888.

HoLotyPe. Mus. Comp. Zool. Harvard, No. 2005 (= Scudder Coll. No. 3062).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Cockerell maintains ‘it appears to be allied to the ordinary species of
Heriades’ .

Heriades mersatus Cockerell 1923

*1923 Heriades mersatus Cockerell : 371.
1931 Heriades mersatus Cockerell ; Salt: 146.

HoLotyPE. Supposedly in Colorado Museum of Natural History, but not referred
to by Rodeck (1938).

DISTRIBUTION. Oligocene: Florissant.
REMARKS. Nearest to H. halictinus. Author ‘sure’ of genus.

Heriades mildredae Cockerell 1925

*1925a Heriades mildvedae Cockerell : 420.
1931 Heriades mildvedae Cockerell; Salt: 146.
1938 Heriades mildvedae Cockerell ; Rodeck : 296.

14

182 MONOGRAPH ON

HoLotyPe. Univ. of Colorado Mus., No. 15075, with counterpart.
DISTRIBUTION. Oligocene: Florissant.

REMARKS. Cockerell remarks that this species differs conspicuously from all
other Heriades.

Heriades priscus Cockerell 1917

*1917 Heriades priscus Cockerell : 383.
1931 Heviades priscus Cockerell; Salt : 146.

HoiotyPe. U.S. Nat. Mus., No. 62543.
DISTRIBUTION. Oligocene: Florissant.

Remarks. The largest Heriades from Florissant. A robust, hairy, black insect.
Cockerell did not appear to have much doubt about its generic position.

Heriades saxosus Cockerell 1913

*1913b Heriades saxosus Cockerell : 233.
1931 Heriades saxosus Cockerell ; Salt : 146.
1938 Heviades saxosus Cockerell ; Rodeck : 296.

HoLotyPe. Univ. of Colorado Mus., No. 18638.
DISTRIBUTION. Oligocene: Florissant.
REMARKS. Nearest to H. laminarum Cockerell.

Genus MEGACHILE Latreille 1802

*1802a Megachile Latreille : 434.
1810 Megachile Latreille ; Latreille : 439.
1828 Megachile Latreille ; Curtis: pl. 218.
1841 Chalicodoma Lepeletier : 309.
GENOTYPE. Apis centuncularis Linnaeus 1758, selected by Curtis (1828).

RemArKS. The International Commission on Zoological Nomenclature recom-
mended that the type-fixation (i.e. Afis muraria Retzius 1783) of Latreille be set
aside and that of Curtis (1828) upheld (ICZN Opinion 219, 1954).

Megachile amaguensis Cockerell 1925

*1925b Megachile amaguensis Cockerell: 7, pl. 1, fig. 1.
1931 Megachile amaguensis Cockerell ; Salt : 146.

HoiotypPe. U.S. Nat. Mus., No. 69601.
DISTRIBUTION. Tertiary: Kudia River, Siberia.

REMARKS. Cockerell states, ‘The distance of the first recurrent nervure from the
base of the second submarginal cell is an unusual feature, which I do not observe in
living species’.

FOSSIL BEES 183

Megachile praedicta Cockerell 1908

*1908c Megachile praedicta Cockerell : 31.
1931 Megachile praedicta Cockerell; Salt : 146.

HototyrPe. Cockerell Collection, whereabouts unknown.
DISTRIBUTION. Oligocene: Florissant.
REMARKS. Genus confirmed by leaves cut through as though by Megachile.

Megachile sp. indet. (Brischke)

1886 Chalicodoma Brischke : 278.

1891 Chalicodoma Brischke ; Scudder : 694 (5985).
1907 Chalicodoma Brischke ; Handlirsch : 888.
1931 Chalicodoma sp. Brischke ; Salt: 144.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Chalicodoma Lepeletier = Megachile Latreille. An unnamed Chali-
codoma was said by Brischke to occur in the Baltic amber, two specimens apparently
being known to him.

Genus OSMIA Panzer 1806

*1806 Osmia Panzer : 230.
1810 Osmia Panzer; Latreille : 439.

GENOTYPE. Apis bicornis Linnaeus 1758, = Apis rufa Linnaeus 1758, selected
by Latreille (1810 : 4309).

REMARKS. It has been found difficult to determine the identity of the species of
Osmia known to Heer. The following notes may prove useful in clarifying this point.

In 1865 Heer mentioned ‘Drei Osmien’, having previously (1849) described one
specimen, Osmia antiqua, in the Karlsruhe Collection. But quite a number of so-
called Osmia are now in the collection at Zurich. These possess Heer’s manuscript
names and were actually placed by him in four groups: (1) Osmia primaeva; (2)
Osmia sp.? dubia; (3) Osmia tmmortua and (4) Osmia kirbyana. Which of these
were the three referred to in 1865 is not known.

Cockerell (1909) studied the species O. primaeva (see p. 167) and referred it to
Andrena. He did the same for O. kivbyana (see p. 202) and considered it, probably
rightly, to bea wasp. No more ‘Osmia’ examined by Heer were studied by Cockerell,
except a now unidentifiable insect 9:5 mm in length.

Of the remaining two groups, those specimens labelled as O. tmmortua are un-
fortunately indeterminable, while of the two specimens labelled ‘dubia’, one is
superficially much like tmmortwa and being also indeterminable it is best placed
there (p. 187). The second specimen, however, is good and has been selected as the
type of a new species (p. 185). However, ‘dubia’, apart from being a manuscript
descriptive term and not a name, is preoccupied by Osmia dubia (Germar). We
have, therefore, named it O. migra. [See also Appendix, p. 256.]

184 MONOGRAPH ON

Osmia carbonum Heyden 1862

*1862 Osmia carbonum Heyden: 75, pl. Io, figs II, 12.
1891 Osmia carbonum Heyden ; Scudder: 721 (6210).
1907 ‘Osmia carbonum’ Heyden ; Cockerell : 228.
1907 Osmia carbonum Heyden ; Handlirsch : 889.
1936 Osmia carbonum Heyden ; Statz: 260.

Hototyre. British Museum (Natural History), 58779 (Heyden’s 1862 fig. rr).
PARATYPE. British Museum (Natural History), 58768 (Heyden’s 1862 fig. 12).
OTHER SPECIMENS. Mus. Comp. Zool. Harvard.

DISTRIBUTION. Oligocene: Rott.

REMARKS. Heyden’s originals were badly figured. His fig. rz is of an insect
without visible wings, and the original of fig. 12, of which the counterpart existed,
was a somewhat smaller insect which has at least the more proximal portions of the
wings still preserved. Both are in the British Museum (Natural History) Collection,
but were badly damaged by exposure to water during the war so that no information
additional to that given by Heyden can be derived from them.

On the other hand, the specimens studied by Cockerell (1907) at Harvard were ‘a
very good specimen with reverse’ and another apparently unnamed one which may
be the counterpart of the paratype. Cockerell says, ‘it is evident that they were not
part of Heyden’s material as they do not agree with his figures and descriptions’.
This is certainly true as regards his ‘good specimen with reverse’, but his statement
relating to wingless originals of O. carbonum is not quite correct as far as the paratype
is concerned. Moreover, after mentioning the ‘goodness’ of the specimen with
counterpart at Harvard, he later quite casually states it ‘is indeterminate’, a state-
ment difficult to reconcile with others previously made. It now remains to be seen
whether Cockerell’s second specimen is the counterpart of Heyden’s paratype.
Should this prove to be the case, no change of nomenclature will be necessary. On
the other hand, a significant difference will necessitate placing O. carbonum under
the gen. indet. group, and a new name will be required for Cockerell’s specimens.

Osmia antiqua Heer 1849

*1849 Osmia antiqua Heer: 95, pl. 7, figs 2a, b.
1856 Osmia antiqua Heer; Giebel : 182.
1865 Drei Osmien; Heer : 386 (partim).
1876 Osmiae; Heer: 43 (partim).
1891 Osmia antiqua Heer; Scudder: 721 (6218).
1907 Osmia antiqua Heer ; Handlirsch : 889.
1g09c ‘Osmia’ Cockerell : 316.
1938 Osmia Armbruster : 87.

3 We consider it quite possible that the specimens described by Cockerell (1907) are, in part, the same
as those described by Heyden (1862) from the collection of Dr Krantz. The difficulty of identification
almost certainly lies with Heyden’s poor descriptions and illustrations. That certain irreconcilable
anomalies, however, do exist is quite clear, e.g. the statement that no signs of visible wings are present
in a specimen of Heyden’s Apis dormitans whereas all Cockerell’s specimens of Apis (from Rott) possess
them. In the circumstances, we have decided either to be cautious where actual similarities suggest
themselves, as in Osmia carbonum, or to regard the two sets of material as separate and distinct, as for
Apis dormitans.

FOSSIL BEES 185

HoLotyPe. Badische Landessammlung fir Naturkunde, Karlsruhe (Oe.91),
iny:1272:

DISTRIBUTION. Miocene: Oeningen.

REMARKS. Cockerell (Ig09c) said this ‘cannot apparently be referred to Osmia
or any other genus with certainty’. He, however, never saw the type and is believed
to have based his statement on Heer’s (1849) entry of a specimen from the Karlsruhe
Collection, and his rather poor illustration of it. Whether Heer (1865) intended this
specimen to be included in his ‘Drei Osmien’ is not known, but it is assumed to be so.

We have now studied the type and, although not well preserved, the following
characters can be discerned. Broad head with a suggestion of large compound eyes ;
parts of one fore and two hind legs, the latter, on the right, possessing distinguishable
hairs ; thorax well chitinized, probably not so broad as head; abdomen also well
chitinized, some segmentation visible, quite broad ; some details of wing venation
can be discerned, but since hind- and forewing overlap, and parts are missing, it is
only possible to state that what can be seen is not incompatible with the wing
venation of Osmza, and certainly the character of the hind legs, segmentation of
abdomen, colour, general habitus etc. do not rule out Osmia either.

MEASUREMENTS. Overall length, 12mm; breadth of head, 4-5 mm (approx.) ;
breadth of abdomen 5 mm ; length of abdomen, 7mm; length of forewing, 9 mm.

Osmia nigra sp. nov.
(Pl. 1, figs 5, 6)
Hototype. Ejidgendssische Technische Hochschule, Ziirich, Pl. I, 719.
PARATYPE. Same collection, Pl. I, 685.
DISTRIBUTION. Miocene: Oeningen.

DESCRIPTION AND REMARKS. The holotype (labelled Osmia sp.? dubia) is a
comparatively small, black, insect, which is preserved in two parts, the abdomen
being 5 mm distant from the head and thorax, but evidently belonging to it. The
head is very imperfectly preserved but is as broad as the thorax. The thorax
shows a well-developed mesoscutum. Parts of the middle and the two hind legs
are preserved, and what is believed to be the hind tibia is triangular-shaped. The
hind femur, however, is as strong and almost as broad as the tibia. Only the left
forewing is preserved, but most of the venation can be discerned and it agrees well
with Osmia: short, tapering marginal cell, broader at the base where it joins the
pterostigma about the middle ; pterostigma well defined but not particularly large ;
radial vein very strong; two subequal submarginal cells; basal vein short and
relatively straight, of two almost equal segments, only slightly deflected at their
junction ; cross-vein cu-a coincident with that of basal vein at point of juncture with
Cu, but posteriorly it arches slightly proximad. Abdomen clearly reveals inter-
segmental membranes in the distal parts.

The paratype is an insect very reminiscent of the holotype and can be regarded
as conspecific. The thorax is well developed with a broad convex mesoscutum. <A

186 MONOGRAPH ON

middle and two hind legs are apparently preserved, with triangular tibiae. Some
aspects of the abdomen are clearly shown. The specimen, however, is not generally
well preserved but nevertheless corresponds with the holotype in all those characters
that can be discerned except for a slightly lighter coloration. Since darker patches
exist on various parts of the body, however, it is not known which was the more
natural coloration for this species. No wings are preserved. MHeer’s manuscript
name for this specimen was Apis adamitica, but it was clearly not the type of the
latter species.

MEASUREMENTS. Holotype: length of abdomen, 6:+5mm; width, 4.5mm;
forewing length 5-5 mm (approx.). Paratype: length of abdomen 6mm; length
of thorax and abdomen (together) 10 mm.

Osmia sp. indet. (Menge)

1856 Osmia Menge: 26.

1891 Osmia Menge; Scudder: 721 (6217).
1907. Osmia Menge; Handlirsch : 889.
1931 Osmia sp. Menge; Salt: 144.

DISTRIBUTION. Eocene: Baltic amber.
REMARKS. Specimen referred to but neither described nor species determined.

Family MEGACHILIDAE incertae sedis
‘Apiaria’ dubia Germar 1849

*1849 Apiaria dubia Germar : 66, pl. 2, fig. 8.
1856 Osmia dubia Germar; Giebel: 182.
1891 Apiaria dubia Germar; Scudder : 688 (5935).
1891 Osmia dubia Giebel ; Scudder : 721 (6220).
1907 (Osmia) dubia Germar ; Handlirsch : 889.

DISTRIBUTION. Oligocene: Orsberg bei Bonn.

REMARKS. Nothing appears to be known about this specimen. The correctness
of Giebel’s generic determination is very doubtful.

Megachilidae gen. et sp. indet.

MATERIAL. Eidgendssische Technische Hochschule, Ziirich, Nos. Pl. I, 687 (with
counterpart) ; Pl. 1, 688 (with counterpart Pl. I, 689) ; X; and Pl. I, 718.

DISTRIBUTION. Miocene: Oeningen.

DESCRIPTION AND REMARKS. Two of the specimens bear Heer’s MS label
‘“mmortua’. The specimens appear to be bees, whose known parts are not incom-
patible with Osmia but whose real systematic position cannot be satisfactorily
ascertained owing to lack of diagnostic characters: The state of preservation is as
follows.

FOSSIL BEES 187

(x). Pl. I, 687 (and counterpart). A fairly well-chitinized insect with head,
thorax, abdomen and portion of one antenna preserved, but with almost all traces
of wings effaced. Body length 10mm; length of abdomen 5:5 mm, width 3 mm ;
length of thorax 3:25 mm, width 3mm. Apart from segments of the abdomen few
morphological details can be discerned.

(2). Pl. I, 688 and counterpart Pl. I, 689. A poorly-defined specimen, no longer
with visible wing venation though the outline of the wings themselves can be detected.
Tegulae quite prominent. Body length 11mm approx.; length of abdomen
55mm, width 375mm; length of thorax not accurately determinable, width
3°25 mm.

(3). One specimen marked “X, Hymenoptera dub.’ Body chitinized, dark brown
in colour and comparatively well preserved, but legs and wings completely lacking.
Two well-developed compound eyes present ; thorax, though less well outlined, is
featureless ; abdominal segments, five in number, clearly in evidence. Body length
10:25 mm ; length of abdomen 5 mm, width 3:25 mm; length of thorax 3:25 mm,
width 3 mm.

(4). Pl. I, 718, bearing Heer’s MS label Osmia sp.? (dubia). Head, thorax and
abdomen are preserved, but legs are lacking and the wing venation is too indistinct
for satisfactory recognition. Both the head and thorax are well chitinized and dark
brown in colour, but the abdomen is paler. Body length (allowing for curved
position) 10mm approx.; length of abdomen 5 mm, width 3:-5mm; length of
thorax 3-I mm, width 3 mm (approx.).

Family ANTHOPHORIDAE Dahlbom 1835

Diacnosis. Labrum usually broader than long, attached by rather short margin
to clypeus ; subantennal areas absent ; subantennal sutures directed towards inner
margins of antennal sockets ; lower sides of clypeus frequently bent back parallel
to long axis of body; facial foveae almost always absent ; galeae elongate post-
palpally, short pre-palpally ; labial palpi with first two segments elongate and
flattened, first segment at least as long as second; glossa linear, usually with a
flabellum ; pre-episternal suture usually absent below scrobal suture, present above
it ; middle coxae elongate, except in a few parasitic forms ; basitibial plates usually
present ; scopa, when present, occurs on the posterior tibiae and basitarsi, and very
rarely found on abdominal sternites as well; pygidial area often present.

Usually three submarginal cells ; when two only, second often much shorter than
first ; rarely only one closed submarginal cell.

DISTRIBUTION. Eocene to Recent.

Subfamily XYLOCOPINAE Latreille 1802

DiaGnosis. Clypeus not protuberant but relatively flat, lateral portions seen
from beneath a little bent backward and more nearly transverse than longitudinal ;
pygidial plate absent, represented in some females by a flat-topped apical spine
which, unfortunately, is hidden in a dense pygidial fimbria and hence would be

188 MONOGRAPH ON

difficult to see in fossils; scopa of female not forming a corbicula; inner apical
margins of posterior tibiae bare or hairy, without a hind tibial rake.

Michener (1944) also mentions a further distinctive feature of the clypeus of
Xylocopinae, though of doubtful use for fossils. The portions of the epistomal
suture between the anterior tentorial pits and the dorsolateral angles of the clypeus
are subparallel to one another, while below the pits the clypeus broadens abruptly.

REMARKS. This subfamily contains two distinctive tribes, the Ceratinini and the
Xylocopini, both represented by fossil forms.

Tribe CERATININI Latreille 1802

D1AGNosIs. Small, rather slender bees with little pubescence ; mandibles very
broad basally but abruptly narrowing to the slender, subparallel-sided apical
portions ; metanotum horizontal ; apex of seventh tergum of the female produced
to a small, sharp, median point.

Wing surfaces hairy throughout, not papillate ; marginal cell broad, longer than
distance from its own apex to wing tip, and considerably bent from the wing margin
apically ; jugal lobe of each hindwing is one-third to half as long as the anal lobe.

REMARKS. Only one fossil specimen so far known.

Genus CERATINA Latreille 1802

1802a Clavicera Latreille : 432.
*1802b Cevatina Latreille : 380.

GENOTYPE. (Hylaeus albilabris Fabricius 1793) = Apis cucurbitina Rossi 1792,
by monotypy.

REMARKS. A recommendation is before the International Commission on Zoo-
logical Nomenclature that the name Clavicera be suppressed.

Ceratina disrupta Cockerell 1906

*1906 Ceratina disrvupta Cockerell : 37.
1907 Cevatina Cockerell; Handlirsch : 891.
1973 Cevratina disrupta Cockerell; Daly: 15, pl. 2, text-fig. 4b.

HoLotype. Mus. Comp. Zool. Harvard, No. 2001 (= Scudder Coll. No. 9355).
DISTRIBUTION. Oligocene: Florissant.

REMARKS. This black, average-sized bee (8 mm) has the general appearance of
a Ceratina. Description of forewing given by Cockerell.

Tribe XYLOCOPINI Latreille 1802

Diacnosis. Large, robust bees with coarse, plumose hair ; metanotum vertical ;
apex of seventh tergum of female produced to a flat-topped spine, largely hidden in a
dense mass of hair ; basitarsi fully as long as the corresponding tibiae.

FOSSIL BEES 189

Wings elongate, papillate distally ; marginal cell very slender, its apex bent away
from the wing margin and the cell itself much longer than the distance from its own
apex to the wing tip; jugal lobe of hindwing much shorter than the anal lobe.

REMARKS. All fossils have been assigned to the genus X ylocopa.

Genus X YLOCOPA Latreille 1802

1802a Xilocopa Latreille : 432.
*1802b Xylocopa Latreille : 379.
1810 Xvylocopa Latreille ; Latreille : 439.

GENOTYPE. Apis violacea Linnaeus 1758, selected by Latreille (1810 : 439).

REMARKS. The name Xylocopa has been placed on the list of nomina conservanda
(ICZN Opinion 743).

Xylocopa friesei Statz 1936
*1936 Xylocopa friesei Statz : 284, 293, pl. 8, fig. 37.
HoLotyPe. Statz Collection.

DISTRIBUTION. Oligocene: Rott.

REMARKS. Wing, thorax and hind basitarsus all strongly suggest a Xylocopa.
According to Statz, the specimen is similar in size to Xylocopa senilis Heer, but since
only the thorax, right forewing and four legs (two very fragmentary) are known,
his statement can have but limited value.

Xylocopa hydrobiae Zeuner 1938
*1938 Xylocopa hydrobiae Zeuner : 127, figs 20, 21.

HoLotyPe. Nat. Hist. Mus. Mainz, Hydrobienkalk Collection, No. 123.
PARATYPE. Same collection, No. 5.

DISTRIBUTION. Miocene: Biebrich nr Mainz.

REMARKS. Vein 2nd r-m together with distal segment of Rs is Z-shaped.

Xylocopa jurinei (Heer 1865)

*1865 Bombus jurinei Heer : 386, fig. 296.
1867 Bombus jurinei Heer; Heer: 4, pl. 3, fig. 8.
1872 Bombus jurineit Heer; Heer: 473, fig. 296.
1876 Bombus jurinei Heer; Heer: 43, fig. 296.
1879 Bombus jurinei Heer; Heer: 411, fig. 338.
1891 Bombus jurvinet Heer ; Scudder : 690 (5953).
1907. Bombus jurineit Heer; Handlirsch : 892.
1g09c Xylocopa jurinei (Heer) ; Cockerell : 315.
1928 Bombus jurinet Heer; Wheeler: 97.
1928 Xylocopa Wheeler : 97.

190 MONOGRAPH ON

1931 Bombus jurinei Heer; Cockerell : 301.
1931 Xylocopa Latreille ; Cockerell : 301.
1938 Bombus jurinet Heer; Armbruster : 87.

HoLotyPeE. Eidgendssische Technische Hochschule, Zurich, No. Pl.I, 682a
(Heer 1865 : fig. 296).

OTHER MATERIAL. Same collection, No. Pl. I, 682b (Heer 1867: pl. 3, fig. 8).

DISTRIBUTION. Miocene (Sarmatian) : Oeningen.

REMARKS. If the specimen No. Pl. I, 682b is not the counterpart of the holotype,
a point very difficult to decide, then Cockerell (1909c) did not see the type, since his
specimen was that figured by Heer (1867).

Of this specimen, Cockerell (1g09c) states, ‘Only part of the venation can be made
out, but all that can be seen agrees with Xylocopa, and not with Bombus’. Arm-
bruster (1938) also suggests the name X ylocopa for this specimen. Having examined
the two existing specimens, we endorse Cockerell’s conclusion as applicable to both.
Moreover, the tibia and basitarsus of the hind legs bear a well-developed scopa, which
is typical of Xylocopa. Furthermore, in specimen Pl. I, 682b, the basitarsus of the
hind leg is approximately one and a half times as long as the tibia. This greater
length of the basitarsus is a character of Xylocopa, whilst in Bombus the basitarsus
is usually shorter than the tibia. The suture between mesoscutum and mesoscutel-
lum is very distinct in the fossils and reminiscent of Xylocopa. A prominent feature
of specimen PI. I, 682b is the strong and copious hairs protruding from the apex of the
abdomen, which again is observed in some Xylocopa.

The species is distinguished from Xylocopa senilis Heer by its larger size.

MEASUREMENTS. Holotype: length of thorax plus abdomen, 21 mm ; length of
abdomen 14mm, width 11 mm; length of thorax 7mm, width 8mm; length of
forewing I5 mm.

Specimen Pl. I, 682b: length of thorax plus abdomen 25 mm ; length of abdomen
18 mm, width 11-5 mm; length of thorax 7mm, width 8mm; length of forewing
16-5mm. The legs here are well preserved, the length of the tibia being 3-75 mm
and the length of the basitarsus6 mm. _ It will also be noticed that the measurements
of the thorax and the abdomen do not quite tally with those given by Heer.

Both specimens have the head missing.

Xylocopa senilis Heer 1849
(Pi a, es’, |)

*1849 Xylocopa senilis Heer : 93, pl. 7, figs 1a, b, c.
1856 Xylocopa senilis Heer ; Giebel : 181.
1865 Xvylocopa senilis Heer; Heer : 386, fig. 295.
1872 Xvylocopa senilis Heer; Heer: 473, fig. 295.
1876 Xvylocopa senilis Heer; Heer: 43, fig. 295.
1879 Xylocopa senilis Heer; Heer: 411, fig. 337.
1885 Xvylocopa senilis Heer; Scudder: 819, fig. 1109.
1887 Xylocopa senilis Heer; Scudder: 821, fig. 1126.
1891 Xylocopa senilis Heer ; Scudder : 734 (6334).
1907 Xylocopa senilis Heer ; Handlirsch : 891.

FOSSIL BEES 191

tg09c Xylocopa senilis Heer ; Cockerell: 415.
1931 <Xylocopa senilis Heer; Pongracz: 117, pls (5) 7, (6) 8.
1938 Xvylocopa senilis Heer; Armbruster: 88.

HoLotyPe. Badische Landessammlung ftir Naturkunde, Karlsruhe, Inv.1267.
PARATYPE. Same collection.

OTHER MATERIAL. Same collection Inv.1270 (as photographed by Pongracz
1931 : pl. (5) 7); Inv.1175 (Oe.89, no. 145); Inv.1175 (Oe.89, no. 146) ; Inv.1228
(Oe.g3). Also Eidgendssische Technische Hochschule, Zurich, no. 21a with counter-
part no. 21b, and another specimen doubtfully referred here, no. 21c.

DISTRIBUTION. Miocene: Oeningen.

REMARKS. We have studied specimens Inv.1175 (no. 145), Inv.1175 (no. 146)
and Inv.1228 from Karlsruhe, and specimens 21a, b and c from Ziirich. It is quite
clear that we have not seen the holotype in spite of the fact that we made a thorough
search for it at Karlsruhe, but we believe a photograph of it was published by
Pongracz (1931c: pl. (6) 8). The strongest evidence for this is the clear reference
to the well-preserved antennae. The paratype, as defined by Heer, was ‘ein sehr
undeutliches Stiick aus derselben Sammlung’. That we have not seen this specimen
is again evident. Possibly it is the second one photographed by Pongracz (1931 :
pl. (5) 7) although the outline here is quite clear, and not indistinct as Heer suggested.
Both types may now be lost but this cannot by any means be asserted since much
material is still stored below the ruins of what was once the Badische Landessamm-
lung, Karlsruhe.

In the absence of types it is difficult to know whether this is a genuine Xylocopa
or not. At least, the specimens at Karlsruhe that have been examined are known
to lack decisive detail for generic determination. The following are the measure-
ments and characters that can be more reliably ascertained.

Inv.1175 (no. 145). Overall length 20 mm ; breadth of abdomen 8 mm ; breadth
of thorax 5mm); length of wing on left 18 mm; length of wing on right 14 mm.
The impression of this insect is distinct and possibly a single leg is displayed on the
left.

Inv.1175 (no. 146). Overall length 21:25mm; breadth of abdomen 8 mm ;
breadth of thorax 5 mm ; length of wing on left 15 mm; that onrightr7mm. This
insect resembles no. 145, though it is not the counterpart. It is less distinctly
preserved.

Inv.1228. Overall length 19mm; length of abdomen 12mm; breadth of ab-
domen 8mm; breadth of thorax 6mm. This is a large brownish-black insect in
which a few proximal veins of the wing can still be discerned, but not enough to
serve any useful purpose. Very few other features are preserved.

On the other hand, the Zurich Collection contains a better preserved specimen
(no. 21a, b) which shows enough of the venation of the forewing to identify it as a
Xylocopa, the basal vein being short and the ‘2nd discoidal’ cell (i.e. 2nd M) elongate.
The copious hairs on the basitarsus of what appears to be the middle leg and the
shortness of the remaining tarsal segments support this identification. Since the
size agrees very well with that of Xylocopa senilis Heer, the specimen is assigned to

192 MONOGRAPH ON

this species without reservation. It is the best preserved specimen available, and
has the following measurements: overall length 22 mm ; length of abdomen 12 mm,
width 8-9 mm ; length of thorax 7mm, width 9mm; length of forewing 8 mm.
The head is rather crushed and thus of little value for measurement since it is too
much spread out. The broad thorax indicates a distinct transverse suture of the
notum.

Specimen no. 2Ic is poorly preserved, without recognizable venation. The
abdomen appears to be more slender than normal for the species. Its specific and
even generic identity is open to grave doubt, though it was labelled ‘“Xylocopa senilis
x Hr. (?)’ in Heer’s time. It is a smaller insect, possibly little more than 10 mm
long, with a greater constriction between thorax and abdomen.

Subfamily ANTHOPHORINAE Dahlbom 1835

DiaGnosis. Clypeus almost always protuberant so that lateral portions, seen
from below, are bent posteriorly and lie parallel to the long axis of the body; the
margins of the clypeus generally converge upward to the dorsolateral angles ;
pygidial plate present in nearly all females and most males; scopa of female not
forming a corbicula ; anterior coxae usually little broader than long ; inner apical
margins of posterior tibiae bare or hairy, without a tibial rake.

REMARKS. This very large subfamily is believed by Michener (1944) to be a
monophyletic unit. The characters, however, of parasitic forms, which have evolved
along different lines, are sometimes difficult to reconcile with those of non-parasitic
forms, and are not always compatible with the diagnostic characters given above.
The tribes Eucerini, Anthophorini and Melectini are represented by fossil forms.
[See also Appendix, p. 256.]

Tribe EUCERINI Latreille 1802

DiaGnosis. Large, pollen-collecting bees; clypeus of the male usually partly
yellow or white ; antennae of males usually much elongated ; flagellar segment in
the female and in some males nearly as long as the scape; posterior legs of the
male only rarely enlarged or modified ; arolia present ; scopa present ; males often
possess distinct pygidial and basitibial plates.

Wing pubescent throughout and only feebly papillate apically ; marginal cell as
long as, or longer than, the distance from its own apex to the wing tip, its apex
narrowly rounded and bent gradually away from the wing margin ; most (including
all North American) genera with three submarginal cells, the first and third longer
than the second ; second abscissa of vein M-Cu of the posterior wings usually nearly
twice as long as the transverse or somewhat oblique cu-a (comparable with
Emphorini) and the jugal lobe about half as long as the anal lobe; pterostigma
reduced, tapering little, if at all, beyond the base of vein r.

REMARKS. The long paraglossae in the Eucerini are of doubtful diagnostic value

for fossil forms. This tribe is one of the largest of the Anthophorinae but is re-
presented with reasonable certainty by only one fossil specimen (Tetralonia berlandt).

FOSSIL BEES 193

Another species, Eucera mortua, described by Meunier (1920) has now been placed
in Anthidium, but it is clear that Meunier felt very sure of his genus. The specimen,
therefore, should be re-examined. That this was not done by us was due to our
being mislead by Statz’s erroneous statements (see Anthidium mortuum (Meunier),

176).

Genus TETRALONIA Spinola 1838

1810 Macroceva Latreille (nec Meigen) : 339.
*1838 Tetralonia Spinola: 538.

GENOTYPE. Eucera antennata Fabricius 1793 (= Apis malvae Rossi 1790) by
subsequent designation of Sandhouse (1943 : 603).

Tetralonia berlandi Théobald 1937
*1937 Tetralonia berlandi Théobald : 131, pl. 2, fig. 11; pl. Io, fig. 1.

HototyPe. Museum at Marseille, Célas Collection, no. C.31.
DISTRIBUTION. Oligocene, Sannoisien du Gard, France.

REMARKS. This broad, black bee most closely approaches to the Recent T.
duvaucelu Lepel., according to Théobald. It is 10mm long, with the forewing
measuring 6mm in length. The specimen is adequately described and figured, and
Théobald’s figure certainly suggests an anthophorid.

Tribe ANTHOPHORINI Dahlbom 1835

DiaGnosis. Moderate-sized to large, pollen-collecting Anthophorinae ; clypeus
of male usually pale ; posterior angles of mandibles are beneath posterior margins
of eyes ; first flagellar segment often elongated, sometimes as long as the scape ;
scopa not large, consisting of simple, or slightly plumose, hairs, or in some cases
with some plumose hairs intermixed ; claws of male cleft, of female with median
tooth on inner side ; arolia mostly present ; eighth tergum of male is usually bilobed,
without a pygidial plate.

Wings largely bare and strongly papillate apically ; pterostigma small, usually
one and a half times as long as broad and ending at the base of veinr ; marginal cell
slightly shorter than the distance from its own apex to the wing tip, rounded at its
apex ; first submarginal cell, usually slightly longer on posterior margin than are the
subequal second and third cells; jugal lobe of posterior wing usually a little over
one-third as long as anal lobe ; second abscissa of vein M-Cu of posterior wings about
as long as cu-a.

REMARKS. The genera Anthophora and Anthophorites are represented by fossil
forms. The latter genus was erected by Heer (1849) for all fossil bees with affinities
with Anthophora. The type, Anthophorites mellona Heer, therefore, is in need of
further examination to establish its true relationship.

194 MONOGRAPH ON

Genus ANTHOPHORA Latreille 1803

1802a Podalivius Latreille : 430.
*1803 Anthophora Latreille : 167.
1810 Anthophora Latreille ; Latreille : 439.

GENOTYPE. Megilla pilipes Fabricius 1775 (= Apis acervorum Linnaeus 1758),
selected by Latreille (1810).

REMARKS. The International Commission of Zoological Nomemclature has
recommended that the generic name Podalirius Latreille (1802) be rejected (ICZN
Opinion I51, 1944).

Anthophora melfordi Cockerell 1908
(PLE Tier 1)

**1908b Anthophora melfordi Cockerell : 323.

1908d Anthophora melfordi Cockerell ; Cockerell : 575, fig. 6.
1931 Anthophora melfordi Cockerell; Salt : 145.

1938 Anthophora melfordi Cockerell ; Rodeck : 295.

HototyPe. British Museum (Natural History), I.g259. Counterpart in Colorado
University Museum, No. 18922.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. According to Cockerell (1908b), ‘This is the first genuine fossil
Anthophora’. The parts preserved include the head, thorax, first abdominal
segment, parts of three legs, most of the forewing overlying remnants of the hind-
wing and the extended mouth parts.

The specimen has been re-examined and the following points noted. The fore-
wing is well preserved, except distally. In the hindwing, however, since it is overlain
by the forewing, the venation is less readily ascertainable. Some veins appear to
be very badly preserved or not at all. The legs, although undoubtedly hairy, are
possibly not so to the extent Cockerell suggests, nor are we able to verify his state-
ment of the ‘copious’ quantity of hair on the ‘tongue’.

As regards the venation of the forewing, there is no need for us to differ from
Cockerell’s decision to place the specimen in Anthophora. The juncture, however,
where the first recurrent vein reaches the second submarginal cell, may be ‘near the
beginning of its last third instead of at the middle’, but it must also be mentioned
that this juncture is far from being clearly visible. Vein cu-a of the hindwing is
very oblique, as Cockerell states.

Genus ANTHOPHORITES Heer 1849

*1849 Anthophorites Heer : 97.
1909c Anthophorites Heer ; Cockerell : 315.

GENOTYPE. Anthophorites mellona Heer (1849), selected by Cockerell (1909¢c).

Remarks. According to Cockerell, ‘Anthophorites cannot be precisely defined,
but includes various fossil bees supposed to be more or less similar to Anthophora’.

FOSSIL BEES 195

The genus is obviously a repository for all Anthophorid-like insects that cannot
otherwise be satisfactorily determined.

The holotype of the genotypic species is now lost, but it was illustrated by Heer
with some wing venation and body characters, insufficient for the establishment of a
generic diagnosis. We have, however, refrained from sinking this genus because no
specimens attributable to the type species have been examined by us.

Anthophorites gaudryi Oustalet 1870

*1870 Anthophorites gaudryi Oustalet : 104, pl. 2, figs 11, 13.
1891 Anthophorites gaudryi Oustalet ; Scudder : 687 (5922).
1907 Anthophorites gaudryi Oustalet ; Cockerell : 227.

1907 Anthophorites gaudryi Oustalet ; Handlirsch : 890.
1937 Anthophorites gaudryi Oustalet ; Théobald : 433.

HoLotTyPe. Museum at Corent, France.
DISTRIBUTION. Oligocene (Stampian): Corent, France.

REMARKS. The specimen is poorly preserved and lacks the head. Scudder
(1891) compared it with A. parietaria. Cockerell (1907) writes, ‘has a curiously
fly-like appearance’, and ‘its hairs are not plumose’. Oustalet’s figure (1870: pl. 2,
fig. 11) is most difficult to interpret.

It should be noted that the species is not listed by Théobald (1937 : 375) where
A. mellona is given as occurring at Corent instead.

Anthophorites thoracica Heer 1867

*1867 <Anthophorites thoracica Heer: 6, pl. 3, fig. 14.
1891 Anthophorites thoracicus Heer ; Scudder : 687 (5925).
1907 Anthophorites thoracicus Heer ; Handlirsch : 890.

Hototypre. Possibly Badische Landessammlungen fir Naturkunde, Karlsruhe.
DISTRIBUTION. Miocene: Radoboj.

REMARKS. Apparently based on one specimen from Radoboj. We particularly
searched for this specimen at Karlsruhe without avail. It may nevertheless still
reappear when the material now stored in the cellars there is unpacked and sorted.

Anthophorites longaeva Heer 1867

*1867 <Anthophorites longaeva Heer: 5, pl. 3, figs 12, 13.
1891 Anthophorites longaevus Heer ; Scudder : 687 (5923).
1907 Anthophorites longaevus Heer ; Handlirsch : 890.
1909c Anthophorites longaeva Heer; Cockerell : 315.

1938 Anthophorites longaeva Heer ; Armbruster : 87.

HoLotyPe. Heer’s (1867) fig. 13 from Radoboj. Whereabouts unknown.
PARATYPE. Heer’s (1867) fig. 12 from Oeningen. Whereabouts unknown.

OTHER SPECIMENS. Eidgendssische Technische Hochschule, Ztrich, Nos Pl. I,
be, Pl. 1-712 and Pl. I, 809.

196 MONOGRAPH ON

DISTRIBUTION. Lower and Upper Miocene: Radoboj and Oeningen.

REMARKS. Of the two specimens figured by Heer (1867), fig. 13, from Radoboj,
appears to be well preserved. Fig. 12, from Oeningen, however, is described as
‘sehr ahnlich A. mellona, aber betrachtlich grésser’. It is believed to have been
seen by Cockerell (1909c) who described it as a bee with large eyes, narrow face,
middle joints of a rather stout flagellum a little longer than broad, and an abdomen
broad at the base, with fine hair ; hind tibia 3-75 mm, hind basitarsus 2-5 mm and
broadened ; venation indistinguishable. The generic position of the specimen was
considered by Cockerell to be ‘wholly obscure’. The three Ziirich specimens can be
described as follows.

Pl. I, 683. A strongly chitinized insect having the head partly lacking ; thorax
well outlined, but without distinct diagnostic characters ; abdomen with segments
clearly indicated and covered with hairs ; hind legs poorly preserved, the basitarsus
and tibia apparently hairy. Body length 16mm; length of abdomen 10 mm,
width 5 mm; length of thorax 4mm, width 4 mm.

Pl. I, 712. Head, thorax and abdomen preserved, but not so highly chitinized as
Pl. I, 683; the strong costa of the wing can be discerned ; abdomen somewhat
distended ; a triangular-shaped chitinous area that is preserved may be a tibia.
Length of body 17mm; length of head 2:25 mm; length of thorax 4-5 mm, width
4:5 mm ; length of abdomen 10-25 mm, width 5-75 mm.

Pl. I, 809. A very weakly chitinized specimen on which few details can be dis-
cerned. Head, thorax and abdomen present, wings completely lacking and legs
practically absent. Length of body 17°-5mm; length of head 2mm; length of
thorax 5 mm, width 5 mm; length of abdomen 10-5 mm, width 6 mm.

These three specimens are not determinable. They are placed under this species
because the characteristics that can be discerned are compatible with the holotype.
They are labelled with the name A. longaeva in a nineteenth-century handwriting.

Anthophorites mellona Heer 1849

*1849 <Anthophorites mellona Heer : 97, pl. 7, figs 4a—c.
1856 <Anthophorites mellona Heer ; Giebel : 183.
1891 Anthophorites mellona Heer ; Scudder : 687 (5924).
1907 Anthophorites mellona Heer ; Handlirsch : 891.
1907. Anthophorites mellona Heer ; Cockerell : 229.
1909c Anthophorites mellona Heer; Cockerell : 315.
1914 <Anthophorites mellona Heer; Meunier : 193, pl. 6, figs 5a, b, text-fig. 5.
1937. Anthophorites melona (Heer) ; Théobald : 375.

HoLotyPe. Originally in Eidgenéssische Technische Hochschule, Ztrich, but
cannot now be traced.

OTHER SPECIMENS. One specimen in Mus. Comp. Zool. Harvard, and another,
described by Meunier, is believed to be in the Marseille Museum.

DISTRIBUTION. Miocene: Oeningen. Also reported from the Oligocene of Aix
and the Oligocene of Cereste, Corent, Célas, France.

FOSSIL BEES 197

REMARKS. Heer’s illustration is unsatisfactory. The thorax is very hairy, and
some wing venation can be made out. One specimen only was described by Heer,
from the Zurich collection, where however we have not succeeded in finding it.

Cockerell (1907), while investigating specimens in the Museum of Comparative
Zoology, Harvard, came across an example of Anthophorites mellona Heer; from
Oeningen, determined by Heer himself. Its venation could not be discerned. Since,
however, some wing venation must have existed in Heer’s original, this specimen
cannot be the holotype. Cockerell (1909c) fixed Anthophorites mellona as the
genotype.

Meunier’s (1914) specimen from Aix-en-Provence was thought by him to belong
to this species, though his description was meagre. He did, however, give photo-
graphs of two specimens. His pl. 6, fig. 5a is the specimen described ; it appears
to be well preserved. His pl. 6, fig. 5b is probably a photograph of Heer’s A.
mellona used for comparison, although no mention of this is given. Moreover, the
resemblance between the two specimens is probably merely superficial.

The reports from the Oligocene are even more doubtful. They appear in Théo-
bald’s (1937) table, for each of the three localities. There is only a single specimen,
however, and one Anthophorites is indeed known from the Oligocene of Corent,
France, viz. A. gaudryi (p. 195). It is listed by Théobald under that fauna (1937 :
433), but omitted from the list.

Anthophorites titania Heer 1849

*1849 <Anthophorites titania Heer : 99, pl. 7, figs 5a, b.
1856 <Anthophorites titania Heer; Giebel : 183.
1891 <Anthophorites titania Heer ; Scudder : 687 (5926).
1907 Anthophorites titania Heer ; Handlirsch : 891.
1909¢ Anthophorites titania (Heer) ; Cockerell : 315.

HoLotypPe. Badische Landessammlung ftir Naturkunde, Karlsruhe, Inv.1273
(Oe.g2).

PARATYPE. Eidgenéssische Technische Hochschule, Ziirich, No. Pl. I, 686 (not
figured by Heer).

DISTRIBUTION. Miocene: Oeningen.

DESCRIPTION. The holotype has been examined, and found to have the head
missing ; the thorax is much distorted, and somewhat hairy. Both hind legs are
in part preserved, one tibia and two basitarsi being very distinct. Both tibiae and
basitarsi are broad and very hairy. The tibia is less well preserved but appears to
be more than one and a half times as long as the basitarsus, without however being
much broader. The abdominal segments are clearly visible and covered with long
hairs. The basal vein is strongly curved at its base and soon straightens out.
Length of abdomen 9 mm, width 5:5 mm.

The paratype, stated by Heer (1849) to have once been in the Karlsruhe Collection,
is now at Ziirich and has been examined by us. It was also studied by Cockerell
(1909c) who considered its generic position to be obscure. This we hereby endorse.

aS

198 MONOGRAPH ON

The specimen is without head or legs, but the segmentation of the abdomen is clear.
A few details of the thorax can be discerned : the pronotum is narrow, well margined
but not quite reaching the tegulae ; the mesonotum is prominent and arched, but
rather small and well defined posteriorly by a transcutal suture ; the mesoscutellum
is clearly evident, the metathorax is delimited posteriorly by ‘a sharp edge separating
base from apical truncation’, and a definite constriction exists between it and the
abdomen. The wing venation has been almost effaced, but a distinct pterostigma
still remains and a strong radius is present. Length of body (without head)
13°25mm; length of thorax 4:25 mm, width 4-1 mm; length of abdomen 9 mm,
width 5:25 mm ; length of forewing 8 mm (as near as can be ascertained).

REMARKS. The shape of the basal vein of the holotype is a trait not uncommon
in Anthophora, as the examination of many Recent specimens has revealed. It
argues strongly against the alternative of assigning A. titania to Halictus. But the
characters preserved do not permit us to be certain that the species is an Anthophora
sensu stricto.

Anthophorites tonsa Heer 1849

*1849 <Anthophorites tonsa Heer: 99, pl. 7, figs 6a, b.
1856 Anthophorites tonsa Heer; Giebel : 183-184.
1891 Anthophorites tonsus Heer; Scudder : 687 (5927).
1907 Anthophorites tonsus Heer ; Handlirsch : 891.

HoLotyPe. Badische Landessammlung ftir Naturkunde, Karlsruhe.
DISTRIBUTION. Miocene: Oeningen.

REMARKS. One specimen only has been described, and it cannot at the moment
be found at Karlsruhe because of war damage. Heer’s (1849) illustration is un-
satisfactory.

Anthophorites veterana Heer 1849

*1849 Anthophorites vetevana Heer : 100, pl. 7, figs 7a, b.
1856 <Anthophorites veterana Heer; Giebel : 183-184.
1891 <Anthophorites veteranus Heer; Scudder : 687 (5928).
1907. Anthophorites veteranus Heer ; Handlirsch : 891.

HOLOTYPE. FEidgendssische Technische Hochschule (Lavater Collection), Ziirich,
No. 14.

OTHER SPECIMEN. Same collection, No. Pl. I, 684.
DISTRIBUTION. Miocene: Oeningen.

DESCRIPTION AND REMARKS. Heer’s figure of the holotype agrees well with the
holotype itself except that he omits the legs although he mentions them in the text.
It is not well preserved. The insect was robust, with a broad convex mesoscutum.
Tegulae prominent. The hind leg, particularly the basitarsus, is hairy and as broad
as the tibia. On account of a depression in the rock which crosses the tibia, the
relative lengths of the tibia and basitarsus are difficult to assess. Some venation

HOSS BEES 199

is present, but since the forewing overlies the hindwing, it is impossible to decipher
it satisfactorily. The strong radius, however, is distinct and also some proximal
portions of M+Cu and Cu. Length of body 14mm; length of thorax 4:25 mm,
width 5-5 mm ; length of abdomen 8-75 mm, width 6:5 mm ; forewing length rr mm.

A second specimen (PI. I, 684) labelled Anthophorites may tentatively be placed
here. It is 12-5 mm long, extremely poorly preserved and black in colour.

Tribe MELECTINI Westwood 1840

Diaenosis. Large, parasitic bees; clypeus of male black; posterior angle of
mandible lies beneath middle of lower end of eye; first flagellar segment not
unusually elongated ; scopa absent; basitibial plates absent ; claws with large
inner basal lobe or tooth ; middle tibial spurs unmodified ; eighth tergum of the
male usually feebly bilobate or bispinose and usually without a pygidial plate ;
seventh abdominal tergite of female with pygidial plate ; seventh sternite of female
somewhat produced, with the margins upturned to form a semitubular guide for
the sting.

Pterostigma several times as long as broad, narrow, and not tapering beyond the
base of vein r ; marginal cell shorter than the distance from its own apex to the wing
tip, broadly rounded apically, and usually not protruding beyond the last submarginal
cell; normally three submarginal cells, the first usually longer than either of the
others ; second abscissa of vein M-Cu of posterior wings shorter than vein cu-a,
sometimes virtually absent ; jugal lobe of hind wing very small.

Genus PROTOMELECTA Cockerell 1908

*t908a Protomelecta Cockerell : 341.
1909b Protomelecta Cockerell ; Cockerell: 81.
GENOTYPE. Pvotomelecta brevipennis Cockerell by monotypy.
REMARKS. According to Cockerell, ‘A very Melecta-like genus but with the
venation of the Eucerines and the large stigma of the Andrenids’. Special mention
of the type was made by Cockerell (1909b).

Protomelecta brevipennis Cockerell 1908

*1908a Protomelecta brevipennis Cockerell : 341.
1909b Protomelecta brevipennis ; Cockerell: 81.
1938 Protomelecta brevipennis Cockerell ; Rodeck : 298.

Ho.otyrPe. Univ. Colorado Mus., No. 18639.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. The affinities of this species are by no means clear. It would appear
to be an anthophorid, but its relationship with the melectines depends to some
considerable extent on theoretical considerations only. A reinvestigation of the
specimen is most desirable.

200 MONOGRAPH ON

Family ANTHOPHORIDAE incertae sedis
‘Anthophora’ effossa Heyden 1862

*1862 Anthophora effossa Heyden : 76, pl. Io, fig. Io.
1891 Anthophora effossa Heyden ; Scudder : 687 (5921).
1907. Anthophora effossa Heyden ; Handlirsch : 890.
1907. Anthophora effossa Heyden ; Cockerell : 228.

1936 Anthophora effossa Heyden ; Statz : 260.

HoLotyPe. British Museum (Natural History), 58781.
DISTRIBUTION. Oligocene: Rott.

REMARKS. It is necessary to distinguish between the ‘Anthophora effossa’ of
Cockerell (1907) which appears to be an ant (p. 202), and the Anthophora effossa
Heyden which Cockerell apparently did not see. He maintained that Heyden’s
figure and description did not agree with his specimen, though whether he was
sufficiently indulgent towards Heyden’s work in this respect is not clear. Heyden’s
figures, however, are entirely useless for diagnostic purposes, and the holotype was
damaged during the war, so that details are no longer recognizable. A further point
to be noted is the very arbitrary way in which Heyden referred his specimen to
Anthophora. For these reasons one is left with no alternative but to place it incertae
sedis.

Anthophoridae gen. et sp. indet. I

1856 Anthophora Latreille; Menge: 26.

1886 Anthophora ? Latreille; Brischke : 278.

1891 Anthophora ? Latreille ; Scudder : 686 (5919).
1891 Anthophora ? Latreille ; Scudder: 686 (5920).
1907. Anthophora ? Latreille ; Handlirsch : 890.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Since Brischke had access to, and reported on, Menge’s specimens,
it is quite possible that Brischke refers here to Menge’s Anthophora. Four specimens
appear to have existed, though Menge mentioned only one.

Anthophoridae gen. et sp. indet. 2

1831 Hymenopt. aus der Fam. der Bienen; MHiinefeld: 2000 (i.e. I100). (Specimens
determined by Burmeister.)

1891 (Apidae) Scudder : 682 (5873).

1907 (Apidae) Burmeister ; Handlirsch : 892.

1931 Apidae, Hiinefeld; Salt: 144.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. This entry has been wrongly attributed to Burmeister. It was
merely the latter’s list that was used by Hiinefeld.

FOSSIL BEES 201

Superfamily APOIDEA (except Apidae) incertae sedis*

Apoidea gen. et sp. indet. 1

1881a Apidae; Malfatti: 98.
1891 (Apidae) Malfatti; Scudder : 682 (5874).
1907 (Apidae) Malfatti; Handlirsch : 892.

DISTRIBUTION. Oligocene: Chiavone, Italy.
REMARKS. Poorly preserved.

Apoidea gen. et sp. indet. 2

1881 Apidae ; Scudder : 290.

1883 Apidae; Scudder: 280.

1891 (Apidae — several) Scudder : 682 (5875).

1907 (Apidae — mehrere) Scudder ; Handlirsch : 892.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. Scudder reported that several species occurred, ‘most of them badly
preserved ; the largest appears to be a Bombus’.

Apoidea gen. et sp. indet. 3
1907 (? Apidae) ; Handlirsch : 893.
MATERIAL. ? Hofmuseum, Vienna (Bosniaski Collection).
DISTRIBUTION. Miocene: Gabbro, Italy.

REMARKS. Poorly preserved bee with short angulated antennae, habitus related
to ‘Apis melisuga’ (see p. 248).

Apoidea gen. et sp. indet. 4

1895 (Apidae n.g. resembling Eucera) ; Scudder: 122.
1907 Apidae n.g. ? Scudder ; Handlirsch : 890.

DISTRIBUTION. Miocene: Oeningen.

REMARKS. Some confusion exists over this entry. There are two difficulties.
Firstly, according to Scudder (1895) and corroborated by Handlirsch (1907), this
specimen may first have been considered a representative of Bombus grandaevus,
but was later relegated to ‘a new genus of Apidae, in which the neuration more
closely resembles Eucerva’. Secondly Cockerell (1g08b) investigated a specimen,
from Scudder’s collection, which he referred to the genus Calypiapis. The doubt
that arises here is whether or not Cockerell was referring to the entry of Scudder in
his reference to a Eucera during his discussion of Calyptapis. See Bombus floris-
santensts (p. 212).

4 See also Appendix, p. 256.

202 MONOGRAPH ON

Apoidea gen. et sp. indet. 5
1928 Apidae Leach; Pongracz: 156.

DISTRIBUTION. Miocene (Sarmatian): Piski, Hungary.

MATERIAL. Magyar Allami Foldtani Intézet, Budapest.

REMARKS. According to Pongracz, the shape of the wings and the pointed
abdomen indicate Osmia; the large, contiguous eyes, however, point to Apis.
Since only the male of Afzs has contiguous eyes, whereas the pointed abdomen is
not characteristic of the male, the specimen cannot be referred to Apis. On account

of the general unreliability of the diagnoses of Pongracz it is impossible to say to
which group of insects the fossil belongs.

FOSSILS ERRONEOUSLY CLASSIFIED AS APOIDEA
(excluding Apidae)®

‘Formicinae’ (Meunier)

*1920 Andrena tertiavia Meunier : 730, pl. I, fig. 4.
1931 Andrena tertiavria Meunier ; Salt: 145.
1936 Andrena tertiavia Meunier ; Statz : 260.
1936 Formicinen; Statz: 260.

HoLotTyPe. Heimatmuseum, Siegburg an der Lahn.

DISTRIBUTION. Oligocene: Rott.

REMARKS. Venation largely effaced. Statz (1936) writes: ‘Nach der Bildung
des Kopfes und der Fihler, sowie der Thoraxsegmentierung und dem Bau der
Beine gehért das Insekt gemass der Ansicht von Bischoff, Berlin, zu den Formicinen.’
(‘According to Bischoff of Berlin, this insect belongs to the formicoids on the basis
of the structure of the head and antennae, as well as the segmentation of the thorax
and leg anatomy.’]

‘Formicinae’ (Cockerell) (nec Heyden)

1907 Anthophora effossa Cockerell : 228.
1907 “An Ant’; Cockerell: 228:

DISTRIBUTION. Oligocene: Rott.

REMARKS. An ant, to be found among a series of specimens from Rott, in the
Mus. Comp. Zool. Harvard.

Polistes kirbyanus Cockerell 1914

1909c ‘Three species of Osmia’ ; Cockerell : 316 (partim).
1g09c A Wasp; Cockerell : 316.

*1t914 Polistes kivrbyanus Cockerell : 638.
1914 Osmia kirbyana Heer MS; Cockerell : 638.

5 See also Appendix, p. 257.

HOSS LE SES 203

Horotype. Eidgendssische Technische Hochschule, Ztirich, No. Pl. I, 681, with
counterpart.

DISTRIBUTION. Miocene: Oeningen.

REMARKS. The view is taken that among the ‘three species of Osmia’ referred to
by Cockerell (1909c) as being at Ztirich, one, the wasp, was called erroneously ‘Osmia
kirbyana’. Not only was the generic identification wrong, but Heer appears to have
given this specimen merely a manuscript name.

LARVAL CHAMBERS
‘Genus’ CELLIFORMA Brown 1934

*1934 Cellifoyrma Brown : 539.
1935 Celliforma Brown : 527.

DiaGnosis. Fossil fillings of chambers purported to have been made by unknown
digging or nest-building Hymenoptera.

GENOTYPE. Celliforma spirifer Brown 1934 by monotypy.

DISTRIBUTION. Cretaceous to Recent.

REMARKS. The name can hardly be maintained under the International Rules of
Zoological Nomenclature, since it refers to objects made by animals, not to the
animals themselves. It is used by Brown merely for the sake of convenience, in
conformity with the practice adopted in palaeontology for fossil burrows, worm
casts and footprints.

Celliforma favosites Brown 1941

*tg4ta Celliforma favosites Brown: 54, pl. 1, figs I, 2.
1941 Celliforma favosites Brown ; Bequaert & Carpenter : 51.
1941b Celliforma favosites Brown ; Brown: 105.

HototypPe. U.S. Geological Survey Collection.

DISTRIBUTION. Upper Cretaceous, Colorado and Montana Beds: Gunlock,
south-west Utah, U.S.A.

REMARKS. Regarded by Brown as the comb of a wasp’s nest resembling that of
Polistes, but the possibility that it belongs to a social bee cannot be entirely dis-
missed. Bequaert & Carpenter (1941) questioned the correctness of Brown’s deter-
mination, to which Brown (1941b) appears to have satisfactorily replied.

Celliforma spirifer Brown 1934

1934 Upper Eocene fossil molds; Brown: 534, fig. 3.
*1934 Celliforma spirifer Brown : 539.
1935 Celliforma spirifer Brown ; Brown: 527.

HototyPe. U.S. Geological Survey Collection.
DISTRIBUTION. Eocene, Bridger Formation : Wyoming, U.S.A.

204 MONOGRAPH ON

REMARKS. This cell resembles that of an Anthophora, with a rounded base and a
spiral seal. Brown, however, emphasizes that many anthophorid bees construct
such cells, hence little systematic significance can be attached to these features,
except that they render probable the existence of Anthophoridae or similar bees.

Celliforma germanica Brown 1935

1907. Anthophora (Podalirius) ? sp. Schiitze : 25, pl. facing p. 56, figs 22, 23.
*1935 Celliforma germanica Brown: 527.

LECTOTYPE. Wiirttembergische Naturaliensammlung, Stuttgart (Schiitze’s fig.
22), here selected.

OTHER SPECIMENS. Same collection, Schtitze’s fig. 23, and a third specimen
referred to on p. 26.

DISTRIBUTION. Oligocene, Bunter Breccia: Wilheim, Ries, Wirttemberg.

REMARKS. Schiitze recognized these as burrows of Hymenoptera, and thought
they belonged to a species of Anthophora, but it is impossible to assign them to any
particular genus. It should be noted that they differ from C. spirifer and C. nuda
in the presence of a slight waist-like constriction. A spiral lid is not preserved.

Celliforma nuda Brown 1935

1915 Lithophaga nuda Dall: 129, pl. 24, fig. 4.
1934 Lower Miocene fossil molds ; Brown: 534, fig. 1.
*1935 Celliforma nuda Brown: 527.
HoLotypPe. U.S. Nat. Mus., No. 165187 or 165189 (Dall 1915: pl. 24, fig. 4).
DISTRIBUTION. Lower Miocene, Silex beds of Tampa Limestone: Florida, U.S.A.
REMARKS. Originally described (Dall 1915) as molluscan boreholes, but recog-
nized by Brown (1935) as hymenopterous. Dall had both molluscan shells of a
Lithophaga nuda which he had described previously, and also burrows which he
assigned to this species. Brown rightly identified as hymenopterous the burrow-
filling figured by Dall; hence it is the holotype of a species Celliforma nuda Brown
(nec Dall).
The cell resembles that of C. spivifer Brown in its structure, including the spiral lid.

Celliforma bedfordi sp. nov.
(PL 2, figs 7; 3}
Diacnosis. The innermost surface of the seal, that facing the main chamber, is
distinctly spiral in structure, of at least five whorls, and is markedly concave.
HoLotype. British Museum (Natural History), In.31434.
TYPE LOCALITY. Venus Bay, South Australia.

DISTRIBUTION. Pleistocene or Sub-Recent, calcareous travertine and consolidated
dune-rock, coasts of south and west Australia.

FOSSIL BEES 205

REMARKS. Specimens of these chambers were obtained, in 1933, by the British
Museum (Natural History) from Mr R. Bedford of Kyancutta Museum, South
Australia. They differ from those described by Lea (1925) in that they are tubular,
with septate cross walls, some 6—7 cm long by 2cm in diameter. According to the
correspondence of Mr Bedford, ‘the prevalent type is only the main chamber of a
structure of which the missing portion consists of a considerable number of small
chambers separated by concavo-convex septa’. The latter, however, are believed
to be the various layers of a single seal to the main chamber.

Celliforma septata sp. nov.
(Plt, ae.o: Pl 2) fig. 7)
DiacGnosis. The innermost surface of the seal, facing the main chamber, is a

plug which is not spiral. It appears to be just slightly convex. Seal composed of
four septa.

Hototypre. British Museum (Natural History), In.31435.
PARATYPE. Same collection, In.34133.
TYPE LOCALITY. Venus Bay, South Australia.

DISTRIBUTION. Pleistocene or Sub-Recent, calcareous travertine and consolidated
dune-rock, south and west Australia.

REMARKS. These chambers appear to correspond to C. bedfordi n. sp. in size and
shape, but differ in the form of the seal.

Celliforma sp. indet.
1932 ‘Tubular burrows in natural chalk’; Buxton: 3.

DESCRIBED SPECIMEN. British Museum (Natural History), Dept. of Entomology.
DISTRIBUTION. ? Bronze Age, not later than King Solomon, Megiddo, Palestine.

REMARKS. Perhaps made by Osmia or Anthophora, according to Buxton (1932).
No seal described. A second group of ‘mud nests’, submitted by Mr P. L. O. Guy,
Oriental Institute, Chicago University, was also mentioned by Buxton. This
appears to be a group of puparia reminiscent of the Australian forms described by
Lea (1925), but somewhat smaller and made of different material. The latter
were reasonably assumed to be the work of beetles still living in the vicinity. No
such conclusion can be hazarded of the organisms responsible for the ‘mud nests’
of Megiddo.

‘Genus URUGUAY Roselli 1939
Uruguay auroranormai Roselli 1939

*1939 Uruguay auroranormai Roselli : 73, figs 17-20, 22, 31(1—4).
1941 Uruguay auroranormai Roselli; Bequaert & Carpenter : 53.
1941b Uruguay auroranormai Roselli; Brown: 110.

206 MONOGRAPH ON

DISTRIBUTION. Cretaceous or later: Uruguay.

ReMARKS. Unfortunately the original work in which Roselli described the
hymenopterous cells has not been seen, but it is fairly obvious that Bequaert &
Carpenter’s comparison of Brown’s Cretaceous fossil, Celliforma favosites (p. 203),
with U. auroranormai is far-fetched. It would appear from the descriptions of
Bequaert & Carpenter (1941) and Brown (1941b) that the cells are not dissimilar to
those described in the present work from south and west Australia and may well be
the work of bees. Moreover, as Brown has pointed out, the fact that they occur in
Cretaceous strata does not imply a Cretaceous origin for them.

V. SYSTEMATIC DESCREPIZONS OF GEOSSIE APT DA

This section contains a detailed study of the fossil Apidae. The material is
arranged according to tribes, and within them stratigraphically.

Family APIDAE Latreille 1802

DIAGNOSIS. Social bees, except possibly the tribe Euglossini (which has not so
far been found in the fossil state) ; second flagellar segment much shorter than scape ;
basitibial plates absent ; scopa of female, when present, forming a corbicula on each
posterior tibia, except in queens of Meliponini; inner apical margins of posterior
tibiae of non-parasitic females, except in queens of Meliponini and workers of
Lestrimelitta, provided with a tibial pollen rake ; pygidial plate absent ; wings hairy
throughout, except possibly in a few fossil forms ; usually (and in all North American
forms) the distance between the costal ends of the two recurrent veins is nearly twice
as great as the length of the second recurrent vein, or even more, and is longer than
the first recurrent.

CHALCOBOMBUS Group

Many fossil insects carry characters of such a type that their exact systematic
position is difficult to define. Often new families, subfamilies and tribes are created
for them but not infrequently this procedure defeats its own end by establishing
units of separation where none exist. Chalcobombus embraces such a group and it
seems to us inadvisable to erect a new tribe for it. This genus could be regarded as
having affinities with the Bombini, or else it may have been the progenitor of
Sophrobombus which we consider to be an ancestral Meliponine, and thirdly, it is not
without affinity with the genus Electrapis which links it with the Apini. For these
reasons we have not assigned a tribal name to the genus.

Genus CHALCOBOMBUS Cockerell 1908

*1908b Chalcobombus Cockerell : 327.
1909d Chalcobombus Cockerell ; Cockerell : 12.

GENOTYPE. Chalcobombus humilis Cockerell 1908, by subsequent designation of
Cockerell (1909d : 11).

FOSSIL BEES 207

DESCRIPTION. Apidae of compact build, about 8mm long; metallic tints
sometimes in evidence; head broad, vertex usually with coarse, erect bristles,
sparingly plumose, facial quadrangle variable ; eyes not usually hairy, and ocelli
large, close together, in a slight curve ; scape comparatively long, flagellum variable
in size; mouth parts elongate, ‘tongue’ hairy but slender with an apical circular
flabellum ; mandible with at least two marginal notches ; abdomen subglobose ;
scutellum prominent and hairy; legs hairy, hind ones considerably plumose,
pulvillus large, claws with short inner tooth; antenna cleaner with exceptionally
long hairs, lining the notch or ‘auschnitt’, its clasp unusually large, fringed with
very fine hairs along its inner surface, lacking a lobe, variable in size; along the
distal margin of the hind tibia a crude rake of strong stiff setae, and flanking its
anterior margin an inner tibial spur with one edge of stiff hairs, closely set. This
spur is fairly closely pressed against the legs.

Forewing. Stigma distinct, slightly variable in size, triangular ; marginal cell
deepish, not particularly long ; three subequal submarginal cells, the first sometimes
the largest, but all three exhibiting variations in shape that are of specific value ;
second submarginal cell receiving first recurrent nervure beyond or near the middle ;
third submarginal cell receiving second recurrent nervure a short distance from its
end ; junction of cubito-anal cross vein and basal nerve coincident.

Hindwing. Cross vein cu-a somewhat oblique, or the upper half vertical ;
hamuli variable in number (8-11).

REMARKS. The genus was only mentioned in 1908 in a list of Chalcobombus
species and in a genealogical tree. It was fully described in 1909. The above
description is based entirely on Cockerell’s published papers.

Chalcobombus hirsutus Cockerell 1908

*1908b Chalcobombus hirsutus Cockerell : 326.
1909d Chalcobombus hirsutus Cockerell ; Cockerell: 12, text-fig. 8.

HOLoTyPE. Geological Museum, University of Koenigsberg.
DISTRIBUTION. Eocene: Baltic amber.
DiaGnosis. Head and thorax with long whitish hairs.

DESCRIPTION. A compact black bee, 8 mm long; head and thorax, particularly
the scutellum, with long, white or yellowish hairs apparently intermixed with brown
ones, and all sparingly plumose ; face, particularly below, broad ; scutellum con-
spicuously projecting ; wings reddish-fuliginous; distal corner of second sub-
marginal cell produced, third submarginal cell longer than second, narrowed more
than half above.

Basitarsus as in C. humilis; legs with much coarse fuscous hair; flagellum is
given as about 0-255 mm long and 0-170 mm thick, last joint 0-323 mm long.

REMARKS. This species was first listed with characters in 1908, but not fully
described until 1909.

208 MONOGRAPH ON

Chalcobombus humilis Cockerell 1908

*1908b Chalcobombus humilis Cockerell : 326.
1909d Chalcobombus humilis Cockerell ; Cockerell : 12, text-figs 6, 7.
1931 Chalcobombus humilis Cockerell; Salt: 145.

HOLOTYPE AND PARATYPE. Geological Museum, University of Koenigsberg.
DISTRIBUTION. Eocene: Baltic amber.
Dracnosis. Head and thorax above with short, black hair.

DESCRIPTION. Black, about 8-5 mm long; head and thorax above with short,
black hair ; vertex and front dull, with sparse, short, coarse, erect, black bristles,
sparingly plumose ; second ‘antennal joint’ barrel-shaped ; wings dark fuliginous ;
stigma triangular and distinct, marginal cell ending rather obtusely away from costa
and possessing a small appendiculation ; second submarginal cell greatly narrowed
above, narrower on marginal than is the third ; third submarginal cell narrowed at
least half above.

Basitarsus squarish ; flagellum a little variable in size and possibly in thickness —
Cockerell quotes lengths of 1-955 mm and 2-125 mm, and thickness 0-187 mm.

REMARKS. This species, selected by Cockerell (1g09d) as the genotype, was
first listed with characters in 1908 but not fully described until 1909.

Chalcobombus martialis Cockerell 1908

*1908b Chalcobombus martialis Cockerell : 326.
t909f Chalcobombus martialis Cockerell ; Cockerell : 22.
1931 Chalcobombus martialis Cockerell; Salt : 145.

HoLotyPe. Geological Museum, University of Koenigsberg.
DISTRIBUTION. Eocene: Baltic amber.

DiaGnosis. Head, mesothorax and scutellum dull black, prothorax and tubercles
coppery red.

DESCRIPTION. About 8 mm long; mesothorax, scutellum and vertex dull black,
but prothorax and tubercles coppery red ; face and abdomen black but legs metallic
copper-red or crimson, moreover, the hind margins of the segments very narrowly
but conspicuously silvery white, the dark part next to the margins stained with red,
especially on the first segment ; scape shortish ; hind basitarsus twice as long as
broad ; third submarginal cell is broad above ; first recurrent nervure joins second
submarginal cell near middle ; wings pale reddish brown ; stigma large but not of
great width ; second recurrent nervure with strong double curve.

Thick-set bee ; pubescence black or dark fuscous, not abundant ; the antennal
joints appear to be known with some precision, as follows : third longer than second,
fourth very short, about twice as broad as long, fifth conspicuously longer than
fourth.

REMARKS. This species was listed with characters in 1908, but not fully described
until 1909.

FOSSIL BEES 209

Tribe BOMBINI Latreille 1802

DiaGnosis. Moderate-sized to large hairy bees; clypeus slightly protuberant,
its lower lateral portions being slightly bent backward ; malar space long ; labrum
about twice as broad as long ; proboscis usually not reaching beyond middle coxae ;
scutellum rounded; posterior tibiae of male neither swollen nor fossate; two
posterior tibial spurs always present ; pterostigma often small, not tapering beyond
base of vein r; apex of marginal cell separated from wing tip by a distance equal
to at least half of cell ; jugal lobe of hindwing becoming reduced or absent.

REMARKS. No primitive forms are known of this tribe, the earliest appearing in
the Oligocene (Aquitanian), and these are poorly preserved. The first convincing
bumble bees are of Oligocene age, from Florissant. [For Probombus see Appendix,

p. 250.]

Genus BOMBUS Latreille 1802

*1802a Bombus Latreille : 437.
1906 Calyptapis Cockerell: 42 (Genotype Calyptapis florissantensis Cockerell 1906 by
monotypy).
1908b Calyptapis Cockerell ; Cockerell : 324.

GENOTYPE. Ais terrestvis Linnaeus 1758 [ICZN Opinion 220, 1954].

Diacnosis. Moderate-sized to large hairy bees; clypeus slightly protuberant ;
female with well-developed corbicula on each hind tibia, the latter also possessing
two strong spurs but no developed comb.

Basitarsus rectangular, strongly developed and inner surface covered with strong
coarse bristles ; depstum or auricle present. Posterior tibia of male flattened but
without corbicula.

Bombus crassipes Novak 1877

*1877 Bombus crassipes Novak: 92, pl. 3, fig. 4.
1891 Bombus crassipes Novak ; Scudder : 690 (5951).
1907 Bombus crassipes Novak ; Handlirsch : 891.
1928 Bombus crassipes Novak ; Wheeler : 97.
1931 Bombus crassipes Novak ; Cockerell : 309.
1936 Bombus crassipes Novak ; Statz : 260.

HoLotyPe. Whereabouts unknown.

DISTRIBUTION. Oligocene (Aquitanian) : Krottensee, Bohemia.

DESCRIPTION. Length of body 1omm, width 6mm; dark brown in colour,
densely hairy ; hind legs strongly developed, very hairy ; auricle possibly not so
well developed ; femurs of hind legs thick ; tibiae flat, widening posteriorly and
truncated ; basitarsus large, flat and truncated at both ends.

REMARKS. One specimen only, not well preserved, neither head nor wings being

discernible. Fortunately Novak’s description is helpful. It suggests that the speci-
men was a genuine Bombus. The figure, however, is very bad and possibly incorrect

210 MONOGRAPH ON

since three hind legs are displayed. What may be the right hind leg is in the correct
position, and except for the omission of the auricle and the tibial spurs is remarkably
reminiscent of Bombus. Of the left hind leg, the basitarsus only is satisfactorily
preserved. This, too, is Bombus-like. The third ‘leg’ is depicted in an anterior
position where in fact the wing might be expected ; perhaps only a mass of chitin is
intended, which has been given the semblance of a basitarsus by the artist. The
hind leg is very hairy and the corbicula is clearly indicated.

Bombus abavus Heer 1867
(PL 2ig.2)

*1867 Bombus abavus Heer: 5, pl. 3, figs 9, Io.
1891 Bombus abavus Heer ; Scudder : 690 (5948).
1907. Bombus abavus Heer ; Handlirsch : 892.
1909c Xylocopa abavus (Heer) ; Cockerell : 314.
1928 Bombus abavus Heer; Wheeler : 97.

1928 Xylocopa abavus (Heer) ; Wheeler : 97.
1931 Bombus abavus Heer; Cuckerell : 309.
1931 Xvylocopa Cockerell : 309.

1938 Bombus abavus Heer; Armbruster : 87.

HoLotypPe. Said to be in Eidgendssische Technische Hochschule, Ziirich, but no
trace can be found in that collection, and it must be presumed lost.

MATERIAL. Two specimens in the Eidgendssische Technische Hochschule,
Zurich (Nos. Pl. I, 808, examined by Cockerell, and SN.201) ; one specimen in the
Badische Landesammlung fir Naturkunde, Karlsruhe (No. Inv.1268) (Oe.go)).

DISTRIBUTION. Miocene, Oeningen.

DESCRIPTION AND REMARKS. Cockerell and Armbruster, who did not examine
the type, were inclined to place this species in Xylocopa, possibly because of Heer’s
remark that the basitarsus is hairy. The venation, however, which Cockerell
wrongly maintained could not be made out, is somewhat like that of Bombus in
specimen No. Pl. I, 808. Moreover, the ‘globular’ abdomen of specimen SN.201 is
not a feature of Xylocopa. In the circumstances it is advisable to retain this form
in the genus in which it was originally placed, but its systematic position remains
uncertain. The three existing specimens may be described as follows.

Specimen PI. I, 808. Only thorax and legs are preserved with parts of the
wings, especially the greater part of one forewing. The basal vein is fairly long, and
its posterior segment is much longer than its anterior segment ; cross vein cu-a is
slightly distad of junction of basal nerve with Cu ; second recurrent nerve is angular
shortly before it meets the base vein of the second submarginal cell. The last
recurrent nerve is incompletely preserved, and the junction les in the proximal
half of this vein.

The wing differs from that of Xylocopa in that the anterior segment of the basal
vein, although short, is not so short as is usual for this genus, secondly in the com-
paratively slightly greater length of the basal vein, and thirdly in the fact that the

HOSS ys Es 211

second recurrent nerve meets the base vein of the second submarginal cell well away
from its apex. In these characters it resembles Bombus.

Forewing length at least 6-5 mm. The legs,® particularly the hind ones, are stout
and hairy but not so much as to warrant the exclusion of the species from Bombus.
The broad thorax reveals substructures of the notum, which was not without hairs.
Length of thorax 4mm, width 5:5 mm.

Specimen SN.201. Head not preserved. The fossil consists of thorax and
abdomen with parts of legs, and the merest trace ofa wing. The thorax and abdomen
are hairy and there is a strong suggestion of hairs on the legs. Since the type
figured by Heer indicated an insect which had part of the head and several legs
present, it is obvious this is not Heer’s type. Length of thorax 4 mm, width 5 mm ;
length of abdomen 6 mm.

Specimen Inv.1268 (Oe.90). This specimen was identified on the label as
Xylocopa sentlis. Like the other known fossils referred to this species (p. 191) it is
poorly preserved, lacking head, part of thorax, and all legs except one hind femur
and tibia. The abdomen, however, is complete, and the hairs and segmentation
very distinct. Moreover, wings are present, that on the right showing some im-
portant features of the venation. Dimensions: length of abdomen 10:5 mm,
width 7 mm ; length of forewing 10mm; length of hind tibia 3-5 mm. It is thus
clear that the specimen is larger than the other known specimens of B. abavus, but
since bumble bees vary much in size, this is no reason for excluding it from the present
species.

Diagnostic characters are seen in the shape of the marginal cell, particularly in
the concavity of its base, which in Xylocopa is convex and very acute proximally,
in the position of the insertion of the first submarginal cross vein (abscissa Rs) on
the posterior margin of the marginal cell which is much more distal in Xylocopa, and
also in the curve, or rather indentation, of the cubitus just prior to its distal splitting
into Cu, and Cu,, a character much less prominent or lacking in Xylocopa. These
are not conclusive arguments for placing the specimen in the genus Bombus rather
than Xylocopa, but since in addition the long tibia has strong hairs arranged in a
manner suggesting a corbicula, the evidence becomes overwhelming. That femur
and tibia are preserved, rather than tibia and basitarsus, is proved by the angle
between them. The basitarsus, indeed, is present also but only very incompletely
preserved ; superficially it resembles the distitarsi. Its erroneous interpretation as
distitarsi no doubt contributed to the earlier identification of the specimen with
Xylocopa.

The colour of the chitin and hairs is dark brown. Hairs are very conspicuous on
the apex of the abdomen but become much sparser on the base of the abdomen and
the thorax. Markings on the thorax indicate the presence of morphologically
distinct regions, but these are too poorly preserved to be of use for generic identifi-
cation.

6 Cockerell (1931) maintained, ‘The legs show a scopa, and the hind tibia is very broad, with a gently
curved longitudinal ridge visible on both sides, and, therefore, certainly natural. This ridge is normal
for Xylocopa, to which Bombus abavus Heer must apparently be referred’. This ridge, however, could
quite well, in part, be the outline of the corbicula of a Bombus, and in any case would almost certainly
have been masked by the scopa in a fossil Xylocopa.

212 MONOGRAPH ON

Bombus florissantensis (Cockerell 1906)

1895 Apidae n.g.; Scudder: 122.

*1906 Calyptapis florissantensis Cockerell : 42.
1907 Calyptapis florissantensis Cockerell ; Handlirsch : 890.
1908b Calyptapis florissantensis Cockerell ; Cockerell : 324, 326.
1909b Calyptapis florissantensis Cockerell ; Cockerell : 80.
1931 Calyptapis Cockerell ; Cockerell : 311.

Hototyre. Mus. Comp. Zool. Harvard, No. 2008 (= Scudder Coll. No. 4933).

MATERIAL. Specimen studied by Cockerell (1908), from the Oligocene shales of
Florissant.

DISTRIBUTION. Oligocene: Florissant.

Diacnosis. Length of body 15 mm ; length of anterior wing 8-5 mm ; head and
thorax black but abdomen rather pale reddish, the junctions of the segments dis-
tinguished by moderately broad, light bands ; although legs are hairy, as in Bombus,
and the hind legs exhibit a corbicula, the abdomen is not noticeably hairy, and the
inner tooth of the bifid claws is much smaller and shorter.

Wings hyaline, with pale veins, cubito-anal cross vein (cu-a) of forewing vertical
except where it bends away immediately prior to the junction with basal nerve. It
differs from Recent Bombus ‘in the form of the third submarginal cell and in the
somewhat less specialised second submarginal’ (Cockerell 1908b).

Stout-bodied, like Recent Bombus ; hind basitarsus flattened and quadrate, with
hairy margins. It is stated to be ‘broadly emarginate apically’.

REMARKS. The above description is based on Cockerell’s of 1908. He placed the
specimen in the genus Calyptapis because of the peculiarities of the second and third
submarginal cells. Apart from the vagueness of his statement, however, it is
precisely at this part of the wing that variation might be expected in an Oligocene
species. The erection of a new genus indeed appears to confuse the issue, for it
seems to us that we are here dealing with a perfectly legitimate Bombus able to bear
the most detailed comparison with Recent forms.

Formerly Cockerell regarded it as near Melissodes (Anthophoridae), but since 1908
as one of the Bombini. He apparently based his first description on a poorly pre-
served specimen already referred to Eucera by Scudder (1895), who said it was ‘an
insect which may possibly be Bombus grandaevus Heer, but which belongs to a new
genus of Apidae, in which the neuration more closely resembles that of Eucera’.
Later, a better specimen was discovered which clearly indicated the relationship
with Bombus.

Bombus proavus Cockerell 1931

*1931 Bombus proavus Cockerell : 309, pl. 1, fig. 8.
1936 Bombus abavus Cockerell ; Statz : 295.

HoLotypPe. Mus. Comp. Zool. Harvard, No. 29402, b.
DISTRIBUTION. Miocene, Latah Formation ; near Spokane, Washington, U.S.A.

FOSSIL BEES 213

Diacnosis. A robust insect, with long black hair, at least on the face ; forewing
15 mm long ; basal vein slightly proximal of junction of vein cu-a ; first submarginal
cross vein (abscissa Rs) not angulate ; first marginal cell on first media cell slightly
longer than second submarginal on same cell; second submarginal cell on the
marginal cell very little longer than the third on same cell ; lower side of third sub-
marginal conspicuously bent at insertion of second recurrent ; wing hyaline, slightly
brownish, the upper part of the marginal cell moderately infuscated.

REMARKS. Insect known from Cockerell’s description and figure. Its body is
so crushed that most features are obliterated, but the forewing is in an excellent
state of preservation. As with B. florissantensis (Cockerell) the third submarginal
cell in this specimen differs from Recent Bombus.

According to Cockerell this was the first true fossil Bombus to be described, and
all other species belong to other genera or are of doubtful reference to Bombus. He
also maintains that it is ‘nearest to European B. lapidarius’. Statz (1936) when
referring to this specimen erroneously named it B. abavus.

Tribe MELIPONINI Handlirsch 1924

DEscrIPTION. Moderate-sized to minute bees, in perennial colonies; Recent
forms with eyes bare, sparsely hairy in some queens ; labrum much broader than
long ; epistomal suture arched upward, as in Afzs, close to the antennal sockets,
and the clypeus not protuberant.

Posterior tibial spurs reduced or absent ; tibial comb always present in Recent
forms (Lestvimelitta excepted) and distinct ; in fossil forms if a comb is not present
there is always an inner but modified tibial spur. Basitarsus very rarely as broad
as tibia ; auricle or true depstum apparently always absent, while the inner surface
of the basitarsus may not have the whole area completely covered with strong
bristles ; claws of Recent females simple.

Ninth sternum of male virtually absent, being represented, if at all, by a narrow,
longitudinal, weakly sclerotic band beneath the genitalia; eighth sternum rather
narrow, having the form characteristic of the ninth in most bees ; sting reduced.

Wings extending beyond the body ; marginal cell often open, and forewing with
veins always weak (or weakening) distally ; pterostigma of moderate to large size,
extending well beyond base of vein r; jugal lobe of hindwing present.

REMARKS. The genus Sophrobombus is represented by fossil forms, but not
Melipona. The tribe ranges from Eocene to Recent and is today widely distributed
in the tropics. Many are found in East African copal.

Genus SOPHROBOMBUS Cockerell 1908

*1908b Sophrobombus Cockerell : 327.
1909f Sophrobombus Cockerell ; Cockerell : 21.

GENOTYPE. Sophrobombus fatalis Cockerell 1908, by monotypy.
16

214 MONOGRAPH ON

DiaGnosis. Small bees resembling Chalcobombus ; mandibles with at least a
convex cutting edge ; ocelli large and arranged in a curve.

Basitarsus flattened, quadrangular, short, having on the inner surface a small,
bristleless area around the posterior apex ; a ridge of short hairs exists on the inner
surface of the hind tibia; inner hind tibial spur present, fringed with hairs along
one edge, and acting as a ‘stop’ to the pollen rake.

Forewing with second submarginal cross vein absent ; basal vein going basad
some distance from vein cu-a ; pterostigma well developed ; marginal cell possibly
slightly open at apex.

REMARKS. Cockerell considered this genus to be a near descendant of Chalco-
bombus. It differs from it, however, in a number of points, particularly in the reduc-
tion of the wing venation. In this respect it is much nearer Tvigona which it closely
resembles in habitus, as well as in the hairless patch on the inner surface of the
basitarsus and the ridge on the inner surface of the hind tibia. There seems little
doubt therefore that it could have been an ancestor of the Meliponini. The presence
of the inner tibial spur, rather than hindering this view, fosters it.

The genus was first mentioned by Cockerell (1908) in a list of Baltic amber species
and in a genealogical tree. It was fully described in the following year.

Sophrobombus fatalis Cockerell 1908

(Pl.-2; figs'3;-4)

*1908b Sophrobombus fatalis Cockerell : 326.
tgo09f Sophrobombus fatalis Cockerell ; Cockerell : 21.
1931 Sophrobombus fatalis Cockerell; Salt : 146.

HoLtotypPe. Geological Museum, University of Koenigsberg.
MATERIAL. Scheele Collection, No. 1225, in Geologisches Staatsinstitut, Hamburg.
DISTRIBUTION. Eocene: Baltic amber.

Diacnosis. Length 8-8-5 mm. Short, thick-set insect, metallic dark coppery-
red, almost without pubescence. Vertex with very short fuscous plumose hair.
Mandibles broadened apically, with a convex cutting edge ; clypeus weakly punctate ;
scape comparatively long and curved. Thorax above with erect, strongly plumose
fuscous hair, twice as long on scutellum as on mesothorax ; scutellum projecting.
Hind tibia long and large ; hind basitarsus quadrangular, less than twice as long as
broad, the outer angles acute. Claws with an inner tooth a little beyond the middle.

DESCRIPTION. Cockerell gave the following description of the specimen: ‘Man-
dibles (apparently without visible teeth) with long and short bristles on the lower
margin ; blades of maxillae long and slender ; antennae very much closer together
than either is to the eye, second joint barrel-shaped, but its basal third narrower
and cylindrical, third joint much longer than fourth but not quite so long as second,
fourth broader than long; legs moderately clothed with fuscous hair; pulvillus
large ; 2-4 tarsal joints, small, cordiform ; wings pale fuscous, veins and stigma
dark brown; hind wing with vein cu-a a little oblique, with the lower endmost
apicad.’ We have examined specimen no. 1225 and find that it agrees closely with

FOSSIL BEES 215

the above description, except that the second flagellar joint, i.e. third antennal
joint, is somewhat longer than the first. This does not, however, justify creating a
new species. The detailed description of this new specimen is as follows.

A metallic, coppery-red insect, not hairy. Overall length, from head to wing tip,
8-5 mm; the actual body, however, is shorter, with an overall length from head to
the tip of the abdomen of only6mm. The extent of wing overlap is thus consider-
able but was possibly somewhat less in the living creature since the whole abdomen
is compact, which suggests considerable contraction. Other approximate measure-
ments are: width of thorax 2-75 mm, length 2-5 mm; width of head 2 mm, length
1:25 mm ; length of tibia 2-5 mm.

Head. The rather large head is not preserved in its entirety since a hollow occurs
in the amber near the mouth parts, and some ‘schimmel’ also has developed here.
The upper part, however, is distinct. The compound eyes appear to be naked ;
the vertex is covered with fuscous hair, possibly plumose; ocelli are large and
prominent, set in a broad shallow triangle ; malar space quite broad but short, if
the limits of observations are the actual limits of this region; one antenna quite
distinct, inserted on the lower half of the face ; scape well formed, curved ; basal
segment of flagellum, i.e. the pedicel, barrel-shaped, second slightly larger than first,
but tapering basally, third the shortest, broader than long but not much shorter
than the fourth segment, the terminal segment the largest, at least twice the size of
the remaining segments which are equal in shape and size. Eleven flagellar segments
are present.

Thorax above indistinct owing to cracked amber and some ‘schimmel’. It was
apparently well arched and covered with hairs which extended somewhat down the
sides. The projected scutellum can be observed.

Abdomen. Ovoidin shape. Although the tip of the abdomen can be seen very
distinctly, there is no trace whatever of a sting.

Legs. Only the hind legs have been sufficiently preserved for study. They are
remarkable for the great length of the tibia with a shallow scoop-like extension over-
lapping the base of the basitarsus on the outside, and the comparatively small,
quadrangular basitarsus, not much longer than broad, with the posterior distal
angle strikingly acute. In detail, the hind leg has the following characters.

Tibia. Corbicula present with the usual scopa, some hairs of which may be
slightly plumose ; the corbicula itself relatively free from hairs; the inner surface
of the tibia exhibits, posteriorly, a longitudinally arranged plateau-like region of
short close hairs, occupying a third of its area and extending slightly to the femur,
and bounded (though seen only under certain lighting conditions) by a dark rim
which is quite broad towards the base of the femur but narrow on the tibia except
for a broader zone at the pollen rake. Two appendages are to be found on the
tibia, a strong pollen rake and an inner tibial spur, the latter having become modified
to carry a blade of fine hairs, as though developing into a comb. That it is not the
latter is clear from its position on the inner surface of the tibia where it forms a
‘stop’ to the pollen rake. It also differs from the typical spur by being more closely
pressed against the leg and more intimately associated with the pollen rake.

216 MONOGRAPH ON

Basitarsus. The inner surface exhibits a well-developed pollen brush, consisting
of seven rows of strong bristles, with possibly 18-20 bristles to the middle rows,
though the latter number is difficult to ascertain with complete accuracy owing to
pollen clogging the brush. Apical and basal rows have at least four bristles less in
consequence of the development of small, bristleless areas around the posterior
apices.

Pulvillus large ; claws bulbous at base and strong, with very small tooth beyond
half-way ; remaining tarsal joints cordiform.

Wings. Fuscous, hairy, and rather more elongate than is normal for the Baltic
amber bees. A study of the venation of the forewing reveals the absence of the
second cross vein of the submarginal cells ; the basal vein becoming basad a short
distance away from vein cu-a ;’ the latter vein a little oblique, the posterior end
most apicad ; stigma well developed ; marginal cell pointed ; second submarginal
cell shorter than slightly open marginal ; second recurrent vein weak and sloping
strongly basad. Since the forewings are folded along the back of the insect, the
hindwings are not visible.

REMARKS. The absence of reference by Cockerell to denticles on the mandible
cannot be taken too seriously since such denticles, unless very prominently displayed
on an open mandible, are not easy to see. Cockerell’s specimen was listed with
characters in 1908, but not fully described until 1909.

Genus TRIGONA Jurine 1807

*1807 Tyrigona Jurine : 246.
1896 Meliponorytes Tosi: 352 (type species Meliponorytes succini Tosi).

GENOTYPE. Apis amalthea Olivier 1789, by subsequent designation of Latreille

(1810 : 430).

DiaGnosis. Stigma well developed, rounded below; wings relatively long,
extending beyond apex of abdomen ; hamuli about five or six, rarely more than nine
or fewer than four; body length from 2 to 8 mm, rarely exceeding II mm.

REMARKS. The genus Meliponorytes was erected by Tosi (1896) to embrace forms
ancestral to Recent Meliponini. All the fossils so far discovered and which have
recognizable descriptions, however, are believed to be referable to Tvigona. But
the creation of the new genus Meliponorytes for what is apparently a genuine Tvigona
must have been a deliberate procedure, since Tosi makes mention of the likeness of
his specimens to Tvigona and also to the subgenus Tetrvagona. Meliponorytes was
thus undoubtedly erected on account of those slight uncertainties attending fossil
forms, although the action here seems scarcely justifiable. Tosi’s descriptions and
illustrations have been very carefully studied and there seems little doubt that his
M. succini (see p. 221) is definitely a Tvigona of the subgenus Tetvagona, whilst the
subgenus cannot be stated for his M. sicula (p. 224). Kerr & Maule (1964) thought

7 Not, as stated by Cockerell for his specimen, a considerable amount, though it is the most we have
noticed among Baltic amber bees.

FOSSIL BEES PIG)

that M. succint belonged to Meliponula, but in a different subgenus from M.
boccande.

It will be recalled that all Tvigona species lack hind tibial spurs, yet all, except
Lestvimelitta, possess the hind tibial comb. The variations of the basitibial surfaces
are of subgeneric value. [For subgenus Nogueivapis see Appendix, p. 257.] |

Subgenus TRIGONA Jurine 1807
*1807 Tvrigona Jurine: 245.
GENOTYPE. Apis amalthea Olivier 1789.

Di1aGnosis. (From Schwarz 1948.) Mandible toothed usually from end to end
along its apex, or with three teeth on at least the outer two-thirds of the apex and
an angulation at the inner extremity. Hind tibiae a little longer than the combined
length of their femora and trochanters. The fringe along the posterior lateral
contour of the hind tibiae with plumose hairs in addition to the usually sparser
simple hairs. The base of the inner face of the hind basitarsi with a differentiated,
more or less oval, area of appressed hairs that contrast with the brush-like hairs on
the apical half to two-thirds of this inner face.

REMARKS. At least one fossil form is so far known, from the Pleistocene.

Trigona (Trigona) erythra Schletterer 1891

*1891 Tvigona erythrva Schletterer : 2.
1895 Melipona togoensis Stadelmann : 620.
1948 Tvrigona erythra var. togoensis Stadelmann ; Schwarz: 51, 61, 68, 132.
1961 <Avxestotrigona erythra (Schletterer) ; Moure: 239.
1961 <Avxestotrigona togoensis (Stadelmann) ; Moure: 239.

FOSSIL SPECIMEN STUDIED. British Museum (Natural History), In.17664.
DISTRIBUTION. Pleistocene: East African copal ; and Recent.

DESCRIPTION. Body length 5-75 mm; general colour brownish-black to black.
Head broader than thorax, hairs upright and relatively strong on vertex, ocelli
prominent, in a gently curved line, hind ocelli more distant from each other than
from compound eyes ; from the front ocellus there runs a prominent frontal line ;
face, particularly the clypeus, appears to be covered with a close white pubescence ;
length of scape little more than half the length of the flagellum, and all flagellar
segments, except the last, not longer than broad, the last being longer than broad,
and pointed.

Thorax markedly triangular, black, its surface matt, probably due to a short
pubescence and to very fine rugosity ; longitudinal line through scutum (median
mesoscutal line) masked by a crack in the amber, but the polished and shining furrow
between scutum and scutellum can just be recognized ; scutellum clearly projecting,
with strong chitinized hairs on its ridge.

Abdomen dark in colour with only few traces of brown, anterior segments smooth
and shiny, hind segments covered with short pubescence ; legs on the whole black

218 MONOGRAPH ON

but not sufficiently well preserved at extremities to judge whether the latter are
brown (most have been cut off short at the edge of the copal) ; corbicula well formed,
wide, neatly hollowed, inner surface difficult to see ; one basitarsus detached, the
other wanting ; claws not notched.

Marginal cell of forewing a little open at apex; one rather weak submarginal
cross vein; first recurrent distinctly angular before junction with vein M; seven
hamuli.

REMARKS. No diagnosis is given because the taxonomic position of the Recent
species is still sub udice, and we are not sufficiently acquainted with previous investi-
gations concerning this group to be able to make a useful contribution.

The fossil agrees with Recent specimens of Tvigona togoensis Stadelmann in the
British Museum (Natural History) and our description of the fossil agrees in all
essential respects with the original description of T. togoensis by Stadelmann.

The two Recent forms T. togoensis and T. erythra differ in the colour of the
abdomen, which is rusty red in the latter. T. togoensis is also slightly smaller.
Schwarz, however, regards these differences at best as of varietal value, and we have
followed him in regarding them as races of a single species. In the colour of the
abdomen, the fossil from East Africa agrees with the West African form T. togoensts.

Subgenus Hypotrigona Cockerell 1934

*1934 Hypotrigona Cockerell : 54.
1961 Liotrigona Moure : 223 (type species Tvigona bottegoi Magretti).
1961 Hypotrigona Moure : 220.

GENOTYPE. Tvrigona gribodot Magretti 1884, by original designation.

DiaGnosis. Minute bees, with dull tegument but without true yellow marks.
Clypeus flat, a little wider than twice its length and one-third as long as clypeocular
distance. Lateral portion of epistomal suture straight. Mandible bidentate.
Scape short. Flagellar segments very short. Bifurcation between M and Cu after
cu-an. Submarginal angle about go°. Hind basitarsus about twice as long as
broad. [See also Appendix, p. 256.]

Trigona (Hypotrigona) gribodoi Magretti 1884
(Pl. 2, figs 5-7)

*1884 Tvrigona gribodoi Magretti : 630.
1894 Trigona braunsi Kohl: 280.
1895 Tvrigona bottegoi Magretti : 156.
1934 Tvigona (Hypotrigona) gribodoi Magretti; Cockerell : 54.
1934 Tvigona (Hypotrigona) bottegot Magretti; Cockerell: 55.
1961 Hypotrigona gribodoi (Magretti) ; Moure : 220.
1961 Luiotrigona bottegoi (Magretti) ; Moure: 223. ,
1964 Tvrigona (Hypotrigona) gribodoi Magretti; Kerr & Maule: 7.
1964 Trigona (Liotrigona) bottegoi (Magretti) ; Kerr & Maule: 7.

FOSSIL BEES 219

SPECIMENS STUDIED. British Museum (Natural History), In.38982—38993,
In.17686, In.18209, In.18214, In.18220, In.18226, 58511, 58622.

DISTRIBUTION. Recent (common in Africa); Late Pleistocene: East African
copal.

Diacnosis. Length of body not exceeding 4mm; head and thorax black,
abdomen and eyes dark brown. There is no apparent natural testaceous coloration
of the scape, mandibles or anterior legs. The body is only sparsely haired except,
chiefly, for the ridge of the scutellum and the basitarsi ; hind tibia triangular, neither
clavate nor club-shaped, the inner face showing an elevated plateau-like area with a
posterior rim.

DESCRIPTION. Face round, black ; antennae of only moderate length, flagellum
broad and shortish ; ocelli prominent. Although the whole face appears uniformly
black, the copal is believed to mask its short white pubescence, except in In.38993.
The same is true of the hairs on the black thorax, but the stronger, more heavily
chitinized hairs of the scutellum can be distinguished. Hind tibia not copiously
haired, and because the hairs are fuscous they are sometimes difficult to see. All
the tarsi appear to be more strongly haired than the tibiae. Wings hyaline ; hamuli
five or six, stigma well developed, marginal cell open at apex, no submarginal cross
veins in evidence, except sometimes a stump for the first.

Some specimens show a very strongly reduced venation particularly of the
posterior apical region of the forewing. The yellow rim to the compound eyes, so
often noticeable in copal specimens, is believed to be due to the medium of their
preservation.

REMARKS. The following are brief statements on the insect inclusions of each
numbered piece of copal preserved in the British Museum collection. The first
twelve pieces come from the Luke Thomas Collection.

In.38982. One typical T. (Hypotrigona) gribodot, and also a second inclusion of
an insect believed to be a termite but whose head and thorax have been removed
during the polishing of the copal.

In.38983. A well-preserved specimen of JT. (H.) gribodot, with quite a dark
abdomen.

In.38984. Eight insect inclusions, i.e. six T. (H.) gvibodot, one dipteron and one
termite. Some are most excellently preserved, and one of the bees shows an
advanced reduction of the veins of the forewing.

In.38985. One T. (H.) gribodoi and various other chitinous remains of insects,
including aculeates. The bee is poorly preserved, portions of it having been worn
away during the polishing of the copal, but sufficient remains for it to be identified.

In.38986. Two T. (H.) gribodot, both well preserved.

In.38987. Copal too cracked for accurate diagnosis, but contains one insect
which appears to be a T. (H.) gribodot.

In.38988. A poorly preserved specimen of T. (H.) gribodot. Copal around the
insect somewhat cracked. A second inclusion, again rather poorly preserved, but
markedly petiolate, is believed to be an ant of the genus Crematogaster.

In.38989. AT. (H.) gribodot, well preserved.

220 MONOGRAPH ON

In.38990. A well-preserved T. (H.) gvibodot, and a second unidentifiable inclusion.

In.38991. Two T. (H.) gribodot, one very well preserved, the other somewhat
less so, on account of its position on the edge of the piece of copal, which is slightly
cracked.

In.38992. One well-preserved T. (H.) gribodot.

In.38993. Two T. (H.) gribodoi, one better preserved than the other. Here the
pubescence, so characteristic of the species, is plainly visible. Both show a little
coating of hair.

58511. All six insect inclusions appear to be T. (H.) gribodov. In copal stated
to be from Stettin on the Baltic, purchased from J. C. Rees 1867. There is, however,
strong evidence from the so-called ‘Baltic copal’ from Stettin was imported from
East Africa.

In.17686. Two T. (H.) gribodot. Copal: East Africa.

The characters of the following specimens agree more with the original description
of T. bottegot Magretti than with T. (H.) gribodot Magretti. Unfortunately, our
study of Recent and fossil material leads us to believe that the characters selected
by Magretti are not constant enough for the separation of the two species, and we
have thus been compelled to sink the former species under the latter.®

In.18229. <A well-preserved T. (H.) gribodot with a detached head of the same
species. The colour of the insect has been somewhat intensified by the copal and
part of the ventral surface of the thorax, abdomen and legs appear whitish. This is
probably because of air bubbles under the chitin of the insect.

In.18214. At first glance the insect in this piece of copal appears different from
the other specimens mentioned above, owing particularly to the yellow-brown
abdomen, ruby eyes and the reddish-tinted cloud on the apical region of the fore-
wings. This intensified colour, however, appears to be some effect of the copal.

In.18220. Two T. (H.) gribodot, and additionally two well-preserved detached
legs almost certainly belonging to the same species. One of the two insects is well
preserved but the upper portion of the head and thorax has been pared away during
the polishing of the copal and it is surrounded by cracks. The second insect is also
well preserved but again has a distinct reddish tinge, particularly noticeable in the
eyes.

In.18226. A single 7. (H.) gribodot, fairly well preserved but with the head
partly pared away. Thorax and legs show many white ‘air-bubbles’. On the abdo-
men the pale bands indicate exposure of the thin intersegmental membrane as a
result of distension of the abdomen.

58622. A single specimen surrounded by fairly copious air-bubbles in the copal.
However, from the side, the specimen is relatively clear to view and is seen to belong
to T. (H.) gribodot. A white rim to the eye and the white sheen to the wings are

8 The characters chosen by Magretti (1895 : 157) to separate T. bottegoi from T. gribodoi were: “per la
superfice minutissimamenta punteggiata, quasi liscia e nettamente splendente del capo, del mesonoto e
dello scudetto; per presentare l’area dell’ocello mediano e la linea mediana frontale leggiermente distinte
ed infine per la colorazione giallo-testacea dello scapo antennale, delle mandibole, del labro et delle due
paia di zampe anteriori.’ [...in having the surface very finely punctured, being almost smooth and
distinctly shiny on head, mesonotum and scutellum; in showing the area of the median ocellus and the

median frontal line faintly distinct and finally by the yellow-testaceous colour of the antennal scape, of
the mandibles, labrum and the two anterior pairs of legs. ]

FOSSIL BEES 221

believed to be spurious effects. It is slightly larger than the others, but the size of
worker bees often varies considerably.

Subgenus TETRAGONA Lepeletier & Serville 1828
*1828 Tetrvagona Lepeletier & Serville : 710.

GENOTYPE. Tvrigona elongata Lepeletier & Serville 1828 (= Centris clavipes
Fabricius 1804).

Diacnosis. The inner face of the hind tibiae unevenly gabled but with a sharply
elevated, median, plateau-like area that extends from the base almost to the apex
and contrasts strongly with a wide, flat, posterior rim (particularly widened at the
apex) ; the contour of the hind tibiae more or less clavate, with the apex in all cases
much wider than the base and sometimes conspicuously expanded ; plumose hairs
in addition to simple hairs usually present along the posterior lateral contour of the
hind tibiae or present over their outer face near the base, or both.

REMARKS. The above is verbatim from Schwarz (1948).

Trigona (Tetragona) succini (Tosi 1896)

*1896 Meliponorytes succini Tosi : 352, pl. 6, figs 1-8.
1907 Meliponorytes succini Tosi; Handlirsch : 892.
1909d Meliponorytes succinit Tosi; Cockerell: 5.

1925 Meliponorytes succini Tosi; Schréder : 254.
1928 Meliponorytes succint Tosi; Wheeler : 97.
1948 Meliponorytes succini Tosi; Schwarz: 8.

1964 Meliponorytes succini Tosi; Kerr & Maule: 2.

HOLOTYPE AND PARATYPES. Museo di Mineralogia dell’Univ. di Bologna.
DISTRIBUTION. Miocene, Sicilian amber.

Diacnosis. Length 4mm; head subquadrangular, mandibles apparently with
only two denticles ; inner surface of hind tibia plateau-like, of sharp elevation, with
the posterior and apical rim fairly flat and wide; inner face of hind basitarsus
contains a smooth, bristleless area ; the hairs along the posterior edge of tibia appear
to be plumose.

DESCRIPTION. Dark, except abdomen which tends to be reddish. Compound
eyes large, ovaland naked. Ocelli situated in a slightly curved line, the lateral ones
prominent, the median one lying in a furrow. Face slightly convex, pubescent, and
the facial area itself ellipsoidal, slightly hairy and prominent. Clypeus almost
triangular and provided with hairs. Vertex double-humped, divided by a median
sulcus. Antenna twelve-segmented, the scape long, about one-third the length of
the antenna, creased at point of insertion. The structure of the antenna is given in
some detail by Tosi, but cannot be said to differ materially from the Recent Tetragona,
unless Tosi’s implication of the rather naked first to third flagellar segments is
regarded as significant. Labrum extended, ‘becoming round and hollowed like a
bowl’. Mandibles with two denticles. Tongue hairy with dilated apex. Labial

222 MONOGRAPH ON

palps apparently four-segmented with long first segment. Thorax essentially as in
Recent Tetvagona, with long hairs on ridge of scutellum. Tegulae well developed.

Wings hyaline but hairy, not extending much beyond the abdomen which appears
to be fully extended (maybe distended). The main features of the forewing are as
follows. Stigma very distinct, lanceolate in form ; first submarginal cross vein not
strong ; marginal cell open at its extremity.

An antennal cleaner is present, and its ‘auschnitt’ lined with hairs. Basitarsus
as long as the remaining tarsal joints taken together and covered with setae.
Posterior tibia well developed and covered apically with thick, long hairs, but basally
with fewer and more robust hairs. Last tarsal joint larger than the others (except
the basitarsus) and cylindrical. Pulvillus large.

Abdomen ovoid, segments all the same size with longer hairs along posterior rim.
First abdominal segment hollowed anteriorly. Ventral surface hulled and its first
segment furnished with rows of longish hairs. A median sulcus follows the shape of
the rows.

REMARKS. This species has been placed in the subgenus Tetvagona not only on
general morphology, but particularly on the structure of the hind leg. If Tosi’s
figures are carefully examined the inner surface of the hind tibia shows both a main
ridge and also a clear dark zone posterior to it. This provides reasonable evidence
that the inner surface is indeed plateau-like, of sharp elevation, with a flat posterior
rim.

Schwarz (1948) who reviewed Tosi’s species came to no definite conclusion with
regard to them. He did, however, raise some meticulous points, and rightly or
wrongly insisted upon very minute and difficult detail. He refers, for instance, to
the two denticles on the outer half of the apex of the mandible instead of the inner
half. It is true Tosi illustrated them in such a position and number, and this may
have been accurate for a Miocene genus, but it must in any case have been an
achievement for Tosi to recognize denticles at all. The distinction of inner and
outer halves may have been well-nigh impossible. Again the appearance of oblique-
ness of the lateral ocelli is obviously influenced by the diffraction of light through
the uneven amber surface and cannot be given the significance which Schwarz
would like to attribute to it. The latter’s reference to the length of the wing and
to the stigma is non-committal.

Trigona (Tetragona) iridipennis Smith 1854
(PL est8)9)
*1854 Trigona iridipennis Smith : 413.
1858 Trigona laeviceps Smith: 51.
1921 Meliponorytes (?) devictus Cockerell ; Cockerell : 544, text-fig. 4.
1922 Tvrigona laeviceps Smith; Cockerell : 714.
1922 Meliponorytes (?) devictus Cockerell ; Cockerell : 714.
1964 Tvrigona (Tetragonula) devictus Cockerell ; Kerr & Maule: 11.

SPECIMENS STUDIED. British Museum (Natural History), In.20702 (holotype of
Meliponorytes (2) devictus Cockerell) and In.43809 (identified by Cockerell as Trigona
laeviceps Smith).

IMOVS SIL, IBVE IES 228

FOSSIL DISTRIBUTION. Probably Pleistocene, Burmese amber: Hukong Valley.

Diacnosis. Length 4mm. Head and thorax black, face with short, silvery
pubescence rather variably distributed, labrum and mandibles ferruginous, antennae
testaceous, margins of thorax and scutellum fringed with fuscous pubescence and
the sides of the metathorax with a silvery pubescence, elsewhere smooth and shiny ;
wings subhyaline, iridescent ; abdomen ferruginous or the base testaceous, smooth
and shining, extreme apex pale testaceous, ventral surface with short, silvery
pubescence ; apical joints of tarsi testaceous.

DeEscriPTIon. Cockerell’s description of specimen In.20702 apparently reflects
the distortion which would be encountered in looking through the spherical bead
containing the specimen, but the bead has now been cut and the following details
may be regarded as reliable. Length 4mm. Head and thorax black, the abdomen
and mandibles ferruginous, tarsi testaceous. The naked eyes are red but this cannot
be considered the natural colour, the head is broad but not unnaturally so, and the
ocelli are prominent, in a curve on the vertex, and have also taken on the reddish
tinge. The antennae appear normal in position, size and shape for T. (T.) iridi-
pennis ; the scape certainly cannot be considered long and the curve is only slight ;
the flagellum is thick and short.

The mesothorax is raised and the scutellum projects, but neither of these features
is striking. Hairs are not much in evidence but a few can be perceived on the
thorax and on the ridge of the scutellum ; this apparent nakedness, however, is
what must be expected of white pubescence on fossilization in amber.

Femora not especially robust, hind tibiae broadened at apex, triangular rather
than clavate. Hind basitarsi are at most only half the length of the tibiae and
cannot be described as large ; on the other hand, the abdomen is not short for the
size of the creature. Forewings subhyaline with a well-developed pale stigma, but
folded over the body so that the venation cannot be determined, except that the
junction of the basal nerve and cross vein cu-a is coincident. Claws simple, pulvilli
large.

The specimen In.43809 was referred by Cockerell to the Recent Tvigona laeviceps
Smith, which is synonymous with T. (Tetragona) iridipennis Smith.

REMARKS. Although the specimens look different, when examined superficially,
the abdomen of ‘M. devictus’ being long and that of ‘T. laeviceps’ short, broad and
curved downwards, they are conspecific since the differences in the abdomen are due
to accidents of fossilization, the first being inflated by gases of putrefaction, the other
strongly constricted, as if dried.

‘T. laeviceps’ Smith had already been synonymized with T. iridipennis Smith by
Schwarz (1948). On the other hand, the comparison of M. devictus with Recent
species has led us to its identification with T. (T.) ividipennis. Cockerell, therefore,
was mistaken in assigning these two specimens to two different genera and species.

The reasons for Cockerell’s mistake are interesting. In 1921 when Cockerell
investigated the first piece of amber, he was under the impression that it was of
Miocene age and in consequence placed his specimen under a Miocene genus,
Meliponorytes. It was in the following year that he received a communication to

224 MONOGRAPH ON

the effect that this amber was, in fact, very much more recent. Consequently the
specimen described in that year was referred to the modern Trvigona laeviceps Smith
which, as we have seen above, is a synonym of T. (Tetrvagona) iridipennis Smith.
Unfortunately Cockerell never rechecked the specimen he designated as Meli-
ponorytes (2?) devictus and in consequence the latter name has persisted in the litera-
ture until now.

Subgenus indet.

Trigona sicula (Tosi 1896)

*1896 Meliponorytes sicula Tosi: 355, fig. 9.
1928 Meliponorytes sicula Tosi; Wheeler : 97.
1948 Meliponorytes sicula Tosi; Schwarz: 9.

HOLOTYPE AND PARATYPES. Museo di Mineralogia dell’ Univ. di Bologna.
DISTRIBUTION. Miocene: Sicilian amber.

Diacnosis. Three times larger than T. (Tetvagona) succint, i.e. about 12 mm long ;
mesonotum a little less arc-shaped, hind basitarsus narrower and longer. Mandibles
with two terminal notches ; pterostigma wanting.

DEscRIPTION. The following characters are common with those of T. (Tetragona)
succint : (1) large oval eyes, (2) situation of ocelli, (3) antennae, (4) labial palps,
(5) claws and pulvillus, (6) hairy tongue, (7) prominent scutellem with long hair.
The species appears to have been rather dark in coloration and the wing very similar
to that of T. (T.) succint except for the lack of stigma.

REMARKS. The subgenus unfortunately cannot be determined. The specimens
themselves are sufficiently like succint to be considered to belong to Tvigona, although
they lack the stigma, but their large size is a little perplexing. Again, the specimens
are not well preserved, and in consequence inadequately described. Tosi wondered
whether the species was not a Melipona, but in view of the present limited distribu-
tion of the latter in South America, and the prevalent ideas of the phylogeny of this
group, the latter suggestion is not a happy one. Only a re-examination of the type
may shed light on the problem. The diagnosis given above is of necessity compara-
tive, in view of Tosi’s manner of description. Yet it must be remembered that the
stigma illustrated by Tosi is much as in some Tetvagona. Again, the comb and
pollen rake of the hind tibia are notoriously difficult to detect in some fossil forms
and are poorly developed in some Recent members of Tetvagona. Failure to detect
them, therefore, may be taken as an indication of their smallness, not of their
absence.

The decision to include the present specimens in Tetragona was mainly the result
of a study of their leg contours, as already mentioned. The wing venation, however,
caused difficulty. According to Tosi vein Cu, is redundant and vein Cu, meets
vein A at the distal and lower apex of cell 2nd Cu. Nevertheless, a slight furrow,
which appears to exist in this vicinity in Recent forms, could, if it existed in fossil
forms, have given rise to the latter illusion. Again, since the junction of the veins

POSSEE BEES 225

Cu,, A and Cu is unknown in Apidae, it is fairly obvious that a mistake has been
made here. With regard to cross vein cu-a, not only does it lie along a rather
abnormal slope but its relation to the basal vein is peculiar. Possibly in his figure
Tosi has rather unwittingly exaggerated the junctions of the veins in this region in
an effort to clarify their course.

More than eight specimens of these fossil bees were found in one piece of amber.

Tribe APINI Latreille 1802

DESCRIPTION. This Old World tribe is unique among bees in the form of the male
genitalia which, unfortunately, are as a rule not accessible in fossil forms.

The Apini have the following characters in common with the Meliponini : scutellum
raised, malar space short, ninth sternum modified. Forewing with first recurrent
vein angulate or curved. Hindwing with jugal and anal lobes.

The following features are diagnostic for Apini. Eyes densely hairy in Recent
forms. Sting well developed. Either one outer tibial spur on hind leg or none ;
hind basitarsus usually broad as tibia and always rectangular, with a depstum, or
the latter broadened into an auricle ; basitarsal brush without the small bristleless
or differentiated area ; claws cleft.

Wings never extending much beyond the body. Distal cells of forewing always
complete, marginal cell parallel-sided for a large part of its length in Recent species,
but not parallel-sided and smaller and deeper in early fossil forms ; pterostigma
becoming reduced, not tapering beyond base of vein r.

REMARKS. Most of the known fossil Apidae belong to this tribe, the most
important and interesting of all as it embraces the honey bee Apis mellifera (p. 249).
Asa result, the evolution of this remarkable insect can be reconstructed in a detailed
manner.

Genus ELECTRAPIS Cockerell 1908

*1908b Electvapis Cockerell : 326.
1909d Electvapis Cockerell ; Cockerell : 7.

GENOTYPE. Apis meliponoides Buttel-Reepen 1906, by monotypy (Cockerell
1908).

Diacnosis. Body morphology not very bomboid (p. 157) but wing venation
retaining bomboid characters. Either a single outer hind tibialspur ornone. While
the body may have a variable quantity of hair, that on the eyes is scant or non-
existent.

DESCRIPTION. Variable in size and shape; from 4mm to 14mm in length.
Normal disposition of tip of abdomen slightly ventral and proximal.

Head. Moderately large and somewhat square ; tuft of hair on the vertex often
present ; compound eyes moderately convex and ocelli prominent. Male has eyes
far apart and the facial quadrangle is squarish.

226 MONOGRAPH ON

Mouth parts are well known only from Cockerell’s descriptions, which may not
apply to all species. They are as follows. ‘Mandibles truncate, much as in Apis,
but the cutting edge with two notches ; labrum semicircular, tongue elongate and
both it and the labial palpi essentially as in Afzs ; spoon on end of tongue small,
maxillary palpi very small, thick, apparently 2-jointed, malar space very short ;
anterior end of clypeus straight’. The little that can be discerned in EF. apoides
Manning, however, suggests a single notch in the cutting edge of the mandibles, and
an elongate ‘tongue’ with a rather large flabellum.

Antennae moderately long, sockets large; scape relatively long and curved,
usually broadened distally and often articulating with a ‘knob-shaped’ pedicel or
first joint of the flagellum. Male antennae longer than female.

Thorax variable; (1) long and broad in proportion to the size of the insect,
tapering sharply distally, very reminiscent of some Tvigona; (2) less broad and
more slender, proportionately more reminiscent of Apis; (3) more spherical in
shape, usually well covered with hair, somewhat reminiscent of Bombus. Scutellum
always distinct and often prominent.

Front leg. Antenna cleaner always present but structurally slightly different
from Recent Apis.

Middle leg. Outer surface of tibia well provided with stout hairs. A single
spur present at apex of tibia, stouter than that of hind leg. Basitarsus with long
strong hairs.

Hind leg. Tibia much longer than basitarsus. Posterior fringe of hairs on the
tibia well developed and believed to contain some plumose hairs among the simple
ones. Corbicula and pollen rake present. Sometimes a ridge covered with fine
hairs occurs on the inner surface of the tibia. Basitarsus rectangular, with a brush
and often a small differentiated area.

Claws bifid. Pulvilli fairly large.

Abdomen. Usually oval in shape and, at most, only moderately clothed with
hair. In E. (Roussyana) palmnickenensis the hairiness of the sternites resembles an
abdominal brush, usually as broad as the tibia or nearly so. The shape of the
abdomen as figured by Buttel-Reepen (1906) may not be typical of the whole genus ;
another known type has one edge convex, more like that of a recent Afzs worker.
The ‘auschnitt’ of the auricle between tibia and tarsus is deep and well formed, but
there is only an incipient development of the depstum. A single outer spur is
typically present.

Forewing. Pterostigma variable in size and shape. Marginal cell moderately
long and generally broader than that of Recent Apis, ending away from wing margin
in a stub. Vein second r-m always curved, mostly S-shaped, but much more up-
right than in Recent Afzs and meeting M very slightly distad of the junction of M
and second m-cu ; first r-m at right angles to M; base veins of cells first R, and
first Rs form a Y with first submarginal cross vein ; the junction of first m-cu with
M is angular as described, but not illustrated, by Cockerell (1909d)._ This junction
is beyond half the wing length and thus much more distal than in Recent Afzs.

FOSSIL BEES 2D

The basal vein meets at the base or just proximad of stigma and is gently S-shaped.
It runs into Cu slightly proximad of the junction of the latter with cross vein cu-a,
or the junctions here coincide. Lower section of basal vein (M) not less than twice
the length of the upper section of basal vein (Rs). Vein Cu, is fairly short.

Hindwing. Hamuli variable (6-16, cf. Chalcobombus with 8-11). Cross vein
cu-a oblique, the slope being posterior and basad, as in Recent Bombus and opposite
in direction to that of Recent Afzs; M runs almost to wing margin; cross vein
r-m well defined, but only at most, slightly oblique, while Rs, like M, almost reaches
wing margin. The general shape is more elongate than Recent Apis, and the wing
possesses distinct and well-characterized anal and jugal lobes.

Wings, in general, hyaline, fuliginous or fuscous and of moderate size, rarely
longer than the body length though from their point of insertion on the thorax
they often extend beyond the abdomen. Sometimes a fairly large, dark-coloured
cloud (about one-twelfth total area) is present in the forewing. Most wings are
hairy, but there is some doubt as to whether this applies to all species.

REMARKS. Most authors have held Electrapis to be directly ancestral to Recent
Apis. This view is likely to be correct regarding only Electvapis s. str., here treated
as a subgenus. The other species show clear signs of deviation. On the whole,
however, one cannot separate these groups as distinct genera, as too many characters
overlap. It is in fact impossible to give clear diagnoses of subgenera, and the three
subgenera here distinguished are to that extent arbitrary. They are:

(1) a group with Afzs-like characters : Electrapis Cockerell 19094, s. str. ;

(2) a group with Bombus-like body and A fvs-like venation ; Protobombus Cockerell

1g08b ;

(3) a Trigona-like group: Roussyana Manning 1960.

Cockerell (1909d) and Zeuner (1944, 1951) considered that the species of Electrapis
were social insects. As evidence Cockerell pointed out the occurrence of many
specimens crowded together in a small piece of amber (EF. meliponoides), whilst
Zeuner considered the structure of the collecting apparatus. On other grounds too,
this contention can be supported. [See also Appendix, p. 257.]

Subgenus ELECTRAPIS Cockerell 1908

*1908b Electvapis Cockerell : 326.
1909d Electvapis Cockerell ; Cockerell : 7.

GENOTYPE. Apis meliponoides Buttel-Reepen 1906, by monotypy (Cockerell
1go8b).

Electrapis (Electrapis) apoides Manning 1960

*1960 Electvapis (Electrapis) apoides Mahning : 306, pl. 5, fig. I.
1970a Electvapis apoides Manning ; Kelner-Pillault : 16, pl. 2, figs 1, 2, text-fig. 5A.

HoLotypPe. No. 94 in the Bursey Collection (see p. 159), now in the Muséum
National d’ Histoire Naturelle, Paris.

228 MONOGRAPH ON

DISTRIBUTION. Eocene, Baltic amber.

DiacGnosis. Body 6:5 mm long, hairy. Forewing 6mm. Hind tibia compara-
tively short and stout, with a single, much reduced, spur; corbicula exceptionally
well developed with strong curved hairs on the posterior edge; basitarsus square
and flat, but with the posterior edge markedly convex. Wings fuliginous with a
cloud apically ; pterostigma large; eight hamulli.

DESCRIPTION. Specimen well preserved, blackish. Total length 6:5 mm;
thorax 2°75 mm long and 2:25 mm wide ; abdomen 2-75 mm long and 2:5 mm wide ;
forewing 6 mm ; hind tibia 2-75 mm.

Proboscis extended for approximately the length of the head, provided with a
large flagellum. Antennae very much like those of E. (Roussyana) proava (p. 236),
with a similarly shaped, knob-like pedicel.

The forelegs differ from those of other species of Electrapis in the shape of the
antenna-cleaner. The zone of articulation between tibia and tarsus is greatly
reduced, and the latter limbs are rounded, and not square, at their point of insertion.
The clasp, too, is more highly developed, now having a fully shaped lobe and a
distal tip projecting over and against the distal edge of the notch. The middle legs
cannot be studied in detail, since they lie too much under the body of the insect.
They are, however, well provided with hairs and appear to possess a single strong
spur.

The hind legs are well proportioned, and not lanky as in E. (Roussyana) proava,
and the hairs of the posterior margin of the corbicula are possibly plumose. On the
apical margin of the tibia, posteriorly, a fine-haired rake is present, and the ‘auschnitt’
of the auricle is well developed. Moreover, the internal spur on the hind tibia is
present and visible on the left leg of the specimen as viewed from behind. It is
thicker and about one and a half times as long as the hairs of the anterior edge of
the leg, but it is evidently in a state of reduction, being so small as hardly to serve
any useful purpose. The basitarsus, although squarish, is elegantly shaped, with
graceful curved outline, and not coarsely square as illustrated by Cockerell for
Chalcobombus. Its inner surface is equipped with strong, almost bristle-like, hairs,
which are arranged in rows as can be seen for those situated at the distal extremity.

The wing venation does not differ materially from that of E. (R.) froava. Possibly
the basal nerve of the forewing is slightly less curved, and the radius stronger. The
hindwing is preserved in its entirety and exhibits eight hamuli, and well-developed
jugal and anal lobes. Only one anal vein is known to be present, but the two lobes
are separated by a deep cleft. A line that could be regarded as 4H is probably no
more than a fault in the amber. Veins Rs and M continue almost to the wing
margin, cross vein r-m slopes basad posteriorly, and cross vein cu-a exhibits a similar,
but more oblique, slope.

On the whole, the creature itself has the habitus of a small Apis. The head is
well proportioned and neatly tapering, less square than some other species of
Electrapis, and the thorax has neither excessive width nor length. The abdomen,
however, is still rather small, though the habit of holding the tip ventrally, together
with the presence of an air bubble on this particular specimen, rather prevents

FOSSIL BEES 229

accurate observation of this point. The wings still project some distance beyond
the apex of the abdomen.

REMARKS. The specimen was kindly lent by the late Mr Bursey of Surbiton,
Surrey, to whom we acknowledge our thanks.

Electrapis (Electrapis) meliponoides (Buttel-Reepen 1906)
(PL. 3, figs 5, 6)

*1906 Apis meliponoides Buttel-Reepen : 158.
1907 Apis meliponoides Buttel-Reepen ; Handlirsch : 892.

_ 1908b Electrvapis meliponoides (Buttel-Reepen) ; Cockerell : 326.
1909d Electrapis meliponoides (Buttel-Reepen) ; Cockerell : fig. 4.
1915 Apis meliponoides Buttel-Reepen ; Buttel-Reepen : 10, fig. 3.
1915 Electrapis meliponoides Butt.-Reep.; Buttel-Reepen : 12.
1928 Apis meliponoides Buttel-Reepen ; Wheeler : 098.

1928 Electvapis meliponoides (Buttel-Reepen) ; Wheeler : 98.
1931 Apis meliponoides Buttel-Reepen ; Statz : 4o.
1931 Electvapis meliponoides Butt.-Reep. ; Statz: 41, figs 2, 3.

HOLOTYPE AND PARATYPE. Two specimens in the Westpreussisches Provinzial-
Museum, Danzig.

MATERIAL. Cockerell’s specimens (1 male, 7 females), Koenigsberg Collection.
Also Nos. 238 and 461 in Scheele Collection, Geologisches Staatsinstitut, Hamburg.
No. 238 is also Typen-Katalog des Geol. Staatsinstitut, No. 149.

DISTRIBUTION. Eocene, Baltic amber.

Diacnosis. Body 14mm. Body hairy but scant hair on eyes; basitarsus
comparatively long, broad and flat, posterior edge not convex ; wings fuliginous.

DESCRIPTION. Cockerell’s description can be summarized as follows (the measure-
ments quoted in hundredths of a millimetre by Cockerell are to a degree of precision
as unnecessary as it is misleading). Black, much erect fuscous hair on head and
thorax. Tibia and tarsus with coarse fuliginous hair; wing venation remarkable
only for the small, piceous, stigma ; it is thought, too, that the upper section of the
basal vein is comparatively short. Male specimen with eyes far apart; facial
quadrangle square ; antennae long, over 5 mm with last joint obliquely truncate.

The following description is based on specimen No. 238 (Scheele Collection), which
one of us (F. J. M.) was able to study in Hamburg. At first sight the fossil appears
to be poorly preserved, in fact so much is covered with ‘schimmel’ that only a few
segments of the abdomen can be discerned. The left forewing and hindwing, how-
ever, are both completely intact and the right forewing also shows much of the
venation. Moreover, many parts of the legs are preserved, as well as the sting, and
the hairiness of the thorax and the colour of the abdomen are clearly perceptible.
Unfortunately, the head is completely missing.

The specimen has the following measurements: overall length, without head,
12mm ; length of forewing, 9:5 mm; hindwing 6mm; basitarsus 2-25 mm.

The copious hairs on the thorax are erect and fuscous. The abdomen is black,
but each segment appears to have a band of dark brown distally ; it is hairy but the

17

230 MONOGRAPH ON

hairs are smaller and darker than those on the thorax. They are, however, longer
at the apex of the abdomen. The sting protrudes considerably and is strong and
straight. In Recent Afis a straight sting is characteristic of the worker, that of the
queen being distinctly curved.

The many fragments of legs cannot be entirely pieced together. It is evident,
however, that the claws are bifid and the distitarsi cordiform. The tibia, which
can easily be recognized as such, has unfortunately had the outer surface pared away
during the polishing of the amber. The corbicula, therefore, is lost, but the strong
hairs of the anterior edge, which have not been destroyed, give complete proof of its
earlier presence, if such were needed. There is no evidence for a tibial spur. This
does not imply absence in view of the fragmentary character of the legs. The basi-
tarsus is large, broad and flat, three-quarters the length of the tibia, the posterior
edge quite straight and the posterior distal angle quite acute. Unfortunately the
details of the structure at the junction of the basitarsus with the tibia are mostly
masked by decayed pollen so that the depstum is not known. The strong non-
plumose hairs of the anterior apex of the tibia, however, are present exactly as in
Recent Afzs and there is reason to suspect that the pollen rake has strong teeth.
There is no evidence whatever for the presence of a pollen comb, but the pollen
brush of the basitarsus is present. The strong hairs are arranged in rows with
pollen still adhering to them. Approximately 17 hairs comprise the bottom
row.

The hairy forewing has a large cloud, and is slightly papillate apically. The
wing venation is that of Electvapis but a slight elongation of the wing itself has
brought about a few minor modifications. These are a narrowing of the apex of the
marginal cell, an elongation of the third submarginal cell and the second M. Proxim-
ally there is an increase in the length of the basal nerve with a consequent compen-
sating alteration on the shapes of the first submarginal cell and the first M. The
junction of basal nerve and cross vein cu-a is coincident on Cu, and vein cu-a slopes
slightly apicad. Pterostigma piceous, but not small, though rather weak.

The hindwing is particularly interesting because of the increase of the number of
hamuli to 16.

The other specimen from the Scheele Collection (No. 461) is poorly preserved,
and covered with much ‘schimmel’. Only part of a hindwing is well preserved. The
insect is reminiscent of specimen No. 238, agreeing with it in all available characters.
Length, with head, at least 13 mm ; hind basitarsus rectangular and long, almost as
broad as tibia, and with a straight posterior edge ; hindwing in venation and shape
agreeing with No. 238 except that the base is slightly shorter and the lobes deeper,
and cross vein r-m appears more oblique. These characters can be accounted for
by the oblique angle at which the wing must be viewed through the amber. At
least 16 hamuli are present.

The remaining features are taxonomically unimportant. They include the
presence of two well-developed antennae inserted on the lower half of the face ;
thorax possibly well supplied with hairs ; wings extending to tip of abdomen and
probably beyond; and the presence of legs whose broad outlines only can be
ascertained.

FOSSIL BEES 231

REMARKS. Cockerell defined the hind basitarsus of his specimens as “broad and
flat, as figured by Buttel-Reepen (Apis meliponoides)’. His specimens are thus
considered to be identifiable with the latter.

Buttel-Reepen was rather obsessed with the idea that E. meliponoides was inter-
mediate in character between Recent stingless bees and Recent honey bees. What
is probably true is that both groups evolved from some common ancestral form and
that the Eocene EF. meliponoides represents one member of this evolving series.
Moreover, Buttel-Reepen stressed his point exclusively with regard to the hind
basitarsus, which alone he described. From the descriptions here given, however,
it is clear that the basitarsus has little that can be described as typically meliponid.

Electrapis (Electrapis) tornquisti Cockerell 1908

*1908b Electrapis (?) tornquisti Cockerell : 326.
1g09f Electvapis (?) tornquist: Cockerell ; Cockerell : 23.
1928 Electvapis tornquisti Cockerell ; Wheeler : 98.
1931 Electvapis (?) tornquisti Cockerell ; Statz: 41.
1931 Electvapis tornquisti Cockerell ; Salt: 145.

HoLotyPe. Geological Museum, University of Koenigsberg.
DISTRIBUTION. Eocene, Baltic amber.

Diacnosis. Length of body 11-5-12mm; robust and more Bombus-like than
the previous species ; head and thorax abundantly hairy, but face with scant hair ;
fourth antennal joint conspicuously shorter than fifth ; wing hyaline, stigma almost
lacking ; hindwings with 15 hamuli; abdomen broad.

DESCRIPTION. A hairy bee, more so than EF. (E.) meliponoides. Coloration of
thorax anomalous. The broad abdomen, although apparently black, has the hind
margins of the segments rather broadly seamed with white. Pubescence of abdomen
short and pale. Tegulae large with very fine delicately plumose hairs.

Venation differing from E. (E.) meliponoides in the rather distal position of the
cross vein cu-a which lies slightly distad of the junction of the basal vein with Cu.
It further differs in the lower part of this vein being strongly bent basad. It agrees
with it, however, in that section Rs of the marginal cell is concave.

REMARKS. Although this specimen, on general morphology, is more bomboid
than other Electrapis s. str., it has been retained in this section for convenience.
The species was listed with some characters in 1908, but was not fully described
until 1909.

Subgenus PROTOBOMBUS Cockerell 1908

*1908b Protobombus Cockerell : 327.
t909d Pvotobombus Cockerell ; Cockerell : 10.

GENOTYPE. Pvotobombus indecisus Cockerell 1908, by monotypy.

D1acGnosis. Wings comparatively apoid, and more bomboid, than in E. (Elec-
tvapis).

232 MONOGRAPH ON

DESCRIPTION. Body globose, compact and Bombus-like. Compared with E.
(Electrapis) it appears to have a greater range of notches on the mandibles (1-3),
the elevation of the scutellum is accentuated and the inner tooth of the claws is
shallower. It agrees with E. (Electrapis) in size (7-10 mm), in its hairless eyes,
comparatively large wings, gently convex posterior margin of the basitarsus, tibial
spur of the middle leg, large ocelli and large pulvillus. The hind tibial spur may or
may not be present.

Forewing. Stigma almost wanting. Marginal cell longer than in E. (Electrapis)
but still ending obtusely at a distance from wing margin. All the submarginal cells
exhibit peculiarities, in particular the second which is, according to Cockerell, “very
broad, pentagonal, much longer than third on marginal, and receiving 1st recurrent
nervure (1st m-cu) at or a little beyond end of its first third’. The latter feature is
remarkable in that it shows a definite advance towards the Synapis wing. This
point, however, is further emphasized by the first and second submarginal cross veins
being ‘equally oblique in opposite directions’, and also by the fact that the third sub-
marginal cell is produced apically, by implication presumably more so than in the
E. (Electrapis) wing. In these points the wing is markedly apoid. It is not a
Synapis wing, however, since the base of the third submarginal cell receives the
second m-cu vein one-sixth from its end, and the first m-cu has a bend or angle in
its upper part. Such characteristics are typical of the genus Electrapis as are also
the remaining characters quoted by Cockerell.

Hindwing. Vein cu-a strongly oblique with sometimes a slight double curve.

REMARKS. Cockerell considered this genus to be ancestral to Bombus. This
point of view, however, is difficult to maintain since Protobombus has a more apoid
wing than even E. (Electrapis), though its body is somewhat reminiscent of Bombus.
The wide divergence from Bombus may be more easily appreciated if the evolutionary
sequence of the second submarginal cell, in relation to the position at which it receives
the first m-cu vein, is summarized :

Chalcobombus — beyond (distal of) the middle.

Electrapis (Electrapis) — just beyond or at the middle.

Electrapis (Protobombus) — much [Cockerell’s term] before the middle.

Synapis — about one-third before middle.

Apis — about one-quarter before middle.

Electrapis (Protobombus) indecisus (Cockerell 1908)

*1908b Protobombus indecisus Cockerell : 326.
1909d Protobombus indecisus Cockerell ; Cockerell : 10, fig. 5.
1931 Pyrotobombus indecisus Cockerell; Salt : 146.

HOoLotTyPe. Geological Museum, University of Koenigsberg.
DISTRIBUTION. Eocene, Baltic amber.

DiaGnosis. About 10 mm long ; mandibles ena truncate with three notches
on cutting edge ; basitarsus broad.

FOSSIL BEES 233

DescripTIon. A blackish bee, but base of abdomen brownish, with sparse pale
hair. Thorax and occiput densely covered with rather short, plumose, whitish
hair. Legs and tegulae on the whole blackish with pale hair. The distinctive
feature of the pale fuliginous wing is a dark cloud at end of marginal cell.

REMARKS. Listed by Cockerell with a few characters in 1908, but not fully
described until 1909.

Electrapis (Protobombus) tristellus (Cockerell 1909)

*tgo9f Protobombus tristellus Cockerell : 24.
1931 Pvotobombus tristellus Cockerell ; Salt : 146.

HoLotyre. Geological Museum, University of Koenigsberg.
DISTRIBUTION. Eocene, Baltic amber.

DiacGnosis. About 7 mm long ; much erect hair on vertex ; wings approximately
6-7mm long; abdominal segments black with hind margins ‘rather broadly pale
reddish’ ; anterior margin of basitarsus gently convex ; glossa, as preserved, short ;
mandibles long, falciform with a single deep notch ; last two joints of labial palpi
rather large ; eyes wide apart above.

DESCRIPTION. A blackish bee ; mesothorax and scutellum as well as vertex with
much erect black hair, abdomen as described above. The forewing has a large,
conspicuous fuliginous cloud. Venation as in E. (Protobombus) indecisus, except
for a presumed double curve of the long, oblique vein cu-a of hindwing. Mouth
parts appear to be well preserved and differ in several respects from those of E.
(Electrapis) meliponotdes, for instance in the short broad glossa and the large
flabellum. The length of the mouth parts, however, is difficult to ascertain in fossil
forms, since they may not be completely extruded. Little reliance, therefore, should
be placed on this character for diagnostic purposes.

REMARKS. This specimen was not listed by Cockerell in 1908.

Subgenus ROUSSYANA Manning 1960

GENOTYPE. Apis palmnickenensis Roussy 1927, by original designation.

DiacGnosis. This subgenus embraces the more Jvigona-like members of the genus
Electrapis. These bees exhibit, however, all the characters of this genus, in so far
as they are known. They are distinguished by the habitus of the body which is
reminiscent of the stingless bees.

Electrapis (Roussyana) palmnickenensis (Roussy 1937)
(Pl.3, figs 1,,2)

*1937 Apis palmnickenensis Roussy : 66, fig. 13.
1938 Apis palmnickenensis Roussy ; Armbruster : 89.
1960 Electrapis (Roussyana) palmnickenensis (Roussy) ; Manning : 306.

HOLOTYPE AND PARATYPES. Roussy Collection, Aigle, Switzerland.

234 MONOGRAPH ON

MATERIAL. In addition to the types, one specimen in Geologisches Staatsinstitut
Hamburg (Scheele Collection, No. 218; also listed in Typen-Katalog des Geol.
Staatsinstitut, No. 150).

DISTRIBUTION. Eocene, Baltic amber.

DiaGnosis. Small, black insects about 4mm long; scutellum very hairy ;
sternites with rows of hairs. Inner surface of tibia with a ridge of short hairs cover-
ing about a quarter of the posterior surface. Basitarsus less than half the length of,
and narrower than, the tibia, its posterior upper apex exhibiting an area clear of
bristles and an incipient depstum.

DESCRIPTION. We have inspected M. Roussy’s piece of amber containing eight
specimens, two of which are complete. The following is the condition of the
individual specimens, their numbers agreeing with the semicircular alignment from
the left upper corner of the piece of amber towards the right lower corner.

1. Badly preserved, except fore and hind legs. Wing venation masked by
‘schimmel’. Possibly not more than 3-5 mm long.

2 and 3. Very much alike and well preserved ; 3-4 mmlong. Details of almost
all appendages can be discerned including mandible and proboscis of No. 3.
Spur on hind legs distinct. Wings of No. 2 not clear.

4. The best preserved specimen, but parts of some legs cut off. A considerable
amount of wing venation can be discerned. 3:5 mm long.

5. Much covered by ‘schimmel’ but much of the wing venation recognizable.
Antennae prominent.

6. Badly preserved and consists only of remains of thorax and legs.

7. Somewhat covered with ‘schimmel’ and wings cut off short. Insect bent on
itself. One antenna present and well preserved.

8. Remains of a head with mandible.

The wings extend well beyond the apex of the abdomen ; the head is broad, but
less so than the thorax. The antennae are inserted on the lower part of the face
and consist of a comparatively long, slightly curved scape and 11 segments to the
flagellum ; the pedicel appears to be knob-shaped. The eyes are prominent on the
side of the head, and the ocelli distinct. The thorax is strong and carries a prominent
scutellum. The abdomen is short and oval. The forelegs exhibit a well-developed
antenna cleaner with broad flap and deeply rounded ‘auschnitt’, whilst the hind leg
is characterized by a distinct corbicula with scopa, strong pollen rake, and short
outer tibial spur but no pollen comb. The basitarsus is rectangular and equipped
with a hairy pollen brush on the inside.

In addition to the eight specimens of M. Roussy, another has been found. It is
preserved in the Scheele Collection, Hamburg. Through the kindness of Professor
E. Voigt and Dr Walter Hantzschel it was lent to us for detailed inspection. Since
the piece of amber contains one complete insect, perfectly preserved, and two frag-
ments of legs, one in excellent condition showing tibia and tarsi, it has been possible
to obtain a reasonably complete description of it, and the following is based on this
specimen.

FOSSIL BEES 235

Size. From head to tip of wing 4:25mm. Length of forewing 3:25 mm,
extending 1:25 mm beyond the tip of the abdomen. Length of hind tibia and tarsus
2mm. This specimen appears to be longer than the Roussy specimens, but this is
deceptive. Though the wings extend conspicuously beyond the abdomen, the length
of the body (which cannot be measured accurately) is possibly no more than 3 mm.

Colour. Very dark brown, or blackish. A creamy shading on the notum may
be due to the natural colour of its copious hairs, or else it is due to the presence of
some pollen.

Head moderately large, black, face sparsely hairy, with cream-coloured pubescence
on the vertex. There is a reddish patch around the right ocellus, but this is almost
certainly the result of post mortem alteration.

The compound eyes are hairless and prominent, and almost reach the bases of the
mandibles, so that the malar spaces are much reduced. The three ocelli are large,
set on the vertex and arranged in a shallow triangle. The single median frontal
suture runs from the middle ocellus downwards.

The clypeus is trapezoid, tapering dorsally because of the convergent slope of the
lateral epistomal sutures, and it is not protruding. The mandibles appear to have
at least a convex cutting edge. The anterior part of the face terminates in a narrow
labrum.

The antennae are not more than moderately long, inserted close together im-
mediately above the frontoclypeal suture and apparently yellow and black. The
scape has slightly less than half the length of the flagellum. The flagellum consists
of eleven segments mostly subequal in size, though the terminal segment is longer,
and those near the pedicel thinner and more tapering, and also darker in colour.

Thorax. Well developed; pronotum short; scutellum very prominent. The
whole thorax copiously covered with hairs which are longest on the scutellum, but
shorter on the pronotum. Tegulae large.

Abdomen. Small, oval. Tergites not noticeably hairy, though the sternites
have rows of well-defined hairs. No intersegmental membranes can be discerned
and this is taken as evidence of the rather contracted nature of the abdomen.

Wings. Hairy and fuscous; pattern of wing venation as in Electrapis, with the
following points of interest ; in the forewing the pterostigma is large and deep ;
the anterior vein bounding the marginal cell, i.e. radius, is not coincident with the
wing margin and its course continues beyond the apex of the marginal cell towards
the wing apex ; cross vein cu-a lies distad of junction of basal nerve with Cu; the
hindwing has only six hamuli and both jugal and anal lobes are well developed.

Three anal veins are present in the hindwing. They can be seen if the left hind-
wing is viewed head-on, i.e. as inspecting the face of the insect.

Legs. Antenna cleaner present on foreleg, ‘auschnitt’ large, but since the legs
are flexed on themselves other structures cannot be clearly identified ; middle leg
has tibia at least as long as femur, broadening towards apex, and covered with strong
hairs, particularly posteriorly. Pollen is entrapped in these hairs. A distinct tibial
spur can be recognized on this leg.

236 MONOGRAPH ON

The hind leg appears to have a comparatively large trochanter distinctly tapering
towards the base; tibia longer than femur, triangular in shape, and with a well-
developed scopa; hairs appear somewhat plumose; corbicula moderately well
developed, smooth, with few hairs within the corbicula itself ; inner surface of tibia
with the ridge of short hairs occupying such a position posteriorly that the rim which
separates it from the posterior edge of the corbicula is very narrow but the anterior
rim is wide. The tibial appendages consist of a small, outer tibial spur, straight,
and non-serrate, and a strong pollen rake projecting on a shelf from the inner apex
of the tibia ; no pollen comb is present, but at the posterior apex of the tibia an
angle suggests an incipient auricle. The tibia is not strongly curved, even pos-
teriorly, but the isolated hind leg, because of the curve of the amber, deceptively
gives this impression.

The basitarsus is narrower than the tibia and rectangular. The two posterior
apices are strongly acute ; the inner surface consists of a brush of strong hairs, whose
disposition in rows is rather masked by the comparatively great length of the bristles ;
possibly five rows is the maximum, though the bristles per row cannot be computed ;
the ‘clear’ area is present around the posterior upper apex. Distitarsi cordiform.

The claws appear entire at first sight but close inspection reveals a single tooth at
about half-way down. Pulvillus well developed.

REMARKS. In 1938 Armbruster criticized Roussy’s description of these insects,
which admittedly was in a popular style. He failed, however, to appreciate their
significance since he was unacquainted with the T7gona-like group of Electrapis.
The very detailed description here given will, it is hoped, fill an awkward gap and
establish this group as a natural division of the primitive honey bees. (See also

Pp. 254.)

Electrapis (Roussyana) proava (Menge 1856)

(Pl. 3, figs 3, 4)

*1856 Apis proava Menge: 26.

1891 Apis proava Menge; Scudder: 689 (5938).

1907 Apis proava Menge; Handlirsch : 892.

1909d Apis proava Menge ; Cockerell : 7.

1915 Apis proava Menge ; Buttel-Reepen : ro.

1928 Apis proava Menge; Wheeler : 98.

1931 Apis proava Menge; Statz: 4o.

1931 Apis proava Menge; Salt : 143.

LECTOTYPE. British Museum (Natural History), In.43592, herein selected.
PARALECTOTYPE. British Museum (Natural History), In.18757.
DISTRIBUTION. Eocene, Baltic amber.

DIAGNOSIS. Body about 6 mm long (after some allowance has been made for a
contracted and bent abdomen), length of forewing 6:-5mm. Scant hair on body
as well as on eyes; hind leg rather lanky, due to a proportionately long tibia,
basitarsus broad, flat and rectangular, posterior edge convex ; wings hyaline, almost
hairless.

FOSSIL BEES 237

DESCRIPTION. The lectotype, on which the following description is based, is well
preserved, blackish in colour, with the terminal portion of thorax becoming paler,
the whole body being very scantily supplied with hairs. The actual length of the
body is 6mm, length of abdomen approx. 2:25 mm, width of abdomen 2-25 mm ;
length of thorax approx. 2°75 mm, width of thorax 2:5 mm; length of hind tibia
3°25 mm.

Head squarish, poorly supplied with hair; ocelli prominent. Mouth parts not
known. Antennae as recent Afzs, but pedicel ‘ball-shaped’ and scape broadened
distally, meeting the pedicel steeply.

Thorax proportionately very broad and thick, tapering sharply distally, not hairy,
reminiscent of some species of Tvigona.

The foreleg has an antenna cleaner of a somewhat primitive type, and well
characterized by the clasp, which is long and thin and completely overlaps the hollow.
The zone of articulation of tibia and tarsus is greatly extended.

The middle legs have a pronounced spur on the apex of the tibia. This spur is
longer and sturdier than the corresponding spur of the hind leg. The outer edge of
the tibia is well provided with hairs.

Hind legs fuscous, a little disproportionate on account of the great length of the
tibia, the latter provided with long, somewhat plumose, hairs along the posterior
edge and a neat row of short hairs down the inner median line. A single spur is
present but its position is such that it is difficult to see on the specimen. Basitarsus
rectangular, with convex posterior edge, covered with hairs which are strong on the
inner side. An auricle is possibly not present but a concavity is present between
tibia and basitarsus. One gains the impression that this zone is slightly more
primitive than in Electrapis (R.) palmnickenensis. There is a tibial rake, which is
stronger and coarser than that of the latter species.

The venation of the forewing is characterized as follows. Stigma comparatively
large, vein cu-a at right angles on A and meeting Cu slightly distally of the junction
with the basal vein. Upper section of basal vein (Rs) apparently proportionately
longer than that of E. meliponoides. :

Hindwing incompletely preserved. Eight hamuli, compared with six in E£. (R.)
palmnickenensis. One might be inclined to regard this as a meliponid feature, since
E. (E£.) tornquisti has 15, E. (E.) meliponoides 16 (though E. (E.) apoides has only 8),
Apis (Synapis) henshawi has 17, Recent A. mellifera 20-23 and A. dorsata 23, but
the smaller A. florea only 12. Recent Bombus species have 20-23 and Euglossa 27.
There appears, however, to exist a relation between the number of hamuli, on the
one hand, and body size and power of flight, on the other.

The part of the hindwing which is preserved includes cu-a, r-m and sections of M
and Rs, together with cell Cu. It agrees with E. (E.) apoides.

The second specimen, In.18757, is covered with ‘schimmel’, and is assigned to
E. (R.) proava with some reservation. It is larger than the lectotype, being 8-1 mm
long. The forewing is 6-8 mm long and shows a dark apical cloud covering about
one-twelfth of its area. The venation, in so far as it can be ascertained, appears to
be identical with that of the lectotype. We hold that this is Menge’s other specimen,
as will be shown under ‘Remarks’.

238 MONOGRAPH ON

REMARKS. It appears that Cockerell (909d) thought little of Menge’s detailed
description, and that he never saw the specimens, presumably because he expected
them to be in the Koenigsberg Collection.

We have satisfied ourselves that the two specimens, In.43592 and In.18757 of the
British Museum (Natural History), are in fact Menge’s types, though no specific
identification was attached to them. Our reasons are as follows. Two specimens
were available to Menge, one almost entire and the other covered with ‘schimmel’,
i.e. with the products of the gases of putrefaction. The first lends itself to descrip-
tion, but the second is in a very poor condition. Menge’s description of the first
tallies in a remarkable way with the British Museum specimen In.43592, and in a
negative manner so does the second, In.18757. Furthermore, the history of the
specimens supports this view. They were purchased in 1892 by the Trustees of the
British Museum from Messrs Stantien & Becker, first placed in the Zoology Depart-
ment and later (about 1904) transferred to the Department of Geology (now
Palaeontology).

Fossil insects preserved in amber were sold by the above-mentioned firm over a
long period of time. Their stocks were eventually taken over by the Prussian State
and became the care of the State Amber Collection in Koenigsberg. Before this
happened, however, Dr Richard Klebs, a geologist in the Prussian State Geological
Survey, was in charge of the amber collection of Messrs Stantien & Becker. A set of
346 specimens were purchased by the British Museum in 1892, which included the
two bees in question. These specimens bear labels marked ‘R. Klebs, Museum
Stantien & Becker’. This suggests that they did not come directly from the amber
works of the firm, but were acquired through Dr Klebs. There is other evidence
that not only the Menge Collection, but also part of the Germar and Berendt Collec-
tions was dispersed with the aid of this firm, and it would appear that Dr Klebs
bought up such collections in the first instance and handed them over to the firm at
a later date. For this reason specimens of such provenance are usually assigned to
the ‘Klebs Collection’.

If the specimens In.43592 and In.18757 were obtained by the British Museum in
1892 from Menge’s collection via Dr Klebs and Messrs Stantien & Becker it is evident
that Cockerell could not have found them in the Koenigsberg Collection in 1908.

It now remains to be shown that Menge’s descriptions indeed tally with those
given here as the result of independent study. Of the well-preserved specimen,
Menge says that it is 7mm long, with the wings 10mm, breadth 3-4mm. The
abdomen is bent so that the sting lies under the apex of the tarsi of the first pair
of legs. The tarsus of the foreleg comprises an antenna cleaner, or rather a sinus
(‘ausschnitt’). A spur is present on the tibia of the hind leg where it projects from
its outer surface, and bristles are present on the digits of the hind feet. The bristles
are a little irregularly shaped. The body itself is scantily haired and the ocelli and
compound eyes are bare. The latter are oblong and narrow, and separated from
each other. Twelve antennal segments are present, and these, together with the
shape of the eyes, indicate for Menge a worker bee.

This description is at variance with our observations in two respects. First, the
number of segments of the antennae of the fossil cannot be ascertained. Those of

FOSSIL BEES 239

the good specimen (and the other is too obscured by ‘schimmel’ for any reliable
observation) has one flagellum cut off short at the surface of the amber, the other
twisted into a difficult position. Menge, however, referred to twelve visible segments
for the worker.? The explanation of the discrepancy between descriptions and fossil
is believed to be the treatment of the specimen whilst in the hands of Messrs Stantien
& Becker. The amber pieces were reduced in size to the minimum, preferably made
rectangular, and embedded in a resin in a glass cell on a microscope slide. The
polishing process resulted in the loss of the distal portion of the antenna, and parts
of the legs. Menge, therefore, may well have been able to study the entire antenna.

The second particular in which the description differs from the fossil is Menge’s
reference to the abdomen being so bent as to lie under the tarsi of the front pair of
legs. The abdomen of the fossil does appear to be bent under the body but certainly
not to such a forward position as Menge suggested. Also no sting can be seen.
The abdomen, however, does touch the tarsus of a leg which must be one of the middle
pair, for both fore and both hind legs can be made out separately. But the dis-
position of the abdomen is so obscured by the poor state of the amber in this region
that it is possible for Menge to have made the above statement rather from con-
jecture than from actual observation, especially as he would have expected the
abdomen to be longer than it actually is, believing the species to be closely related
to Apis mellifera.

Though Menge was right in saying that the general morphology of the specimens
resembles that of the honey bee in many respects, closer study has now revealed
some features reminiscent of Tvigona, such as the proportionately broad thorax, the
short and round abdomen, and the small number of hamuli.

Genus APIS Linneaus 1758

*1758 Apis Linnaeus : 343, 575.
1810 Apis Latreille : 439.
1938 Hauffapis Armbruster : 43.

GENOTYPE. Apis mellifera Linnaeus 1758 (= Apis mellifica Linnaeus 1766),
designated by Latreille (1810 : 439).

D1acnosis. Medium-sized bees, with hair. Eyes usually hairy, clypeus not
protuberant ; well-developed hind tibia, but spurs and comb lacking ; basitarsus
rectangular and covered with bristles arranged in rows; no reduction in wing
venation, marginal cell always complete.

REMARKS. This genus comprises the Oligocene, Miocene and Recent Apini. They
are all extremely closely related to each other, and the division into two subgenera
is based on very unsatisfactory characters. Similarly, the numerous species and
subspecies that have been named cannot all be maintained if the standard of Recent
taxonomy is applied.

9 This number is incorrect of course, for Recent Apis workers, if it refers to the flagellar segments,

since it is the drone that has twelve and the queen and worker (females) eleven segments, but it was
probably computed to include the scape.

240 MONOGRAPH ON

Subgenus SYNAPIS Cockerell 1907

*1907. Apis (Synapis) Cockerell : 229.
1931 Synapis Cockerell; Statz: 42.

GENOTYPE. Synapis henshawi Cockerell 1907, by monotypy.

Di1aGNosIs. General morphology very much like that of Recent Apis but differing
in the following points: labrum less tongue-shaped ; third segment of flagellum
short but broad ; basitarsus, though variable in shape, never identical with Recent
Apis ; junction of veins cu-a, Cu and the basal vein coincident, or vein cu-a lies in a
slightly more proximal position.

REMARKS. This group of honey bees has been extensively treated by Statz in
1931 and 1934, who regarded it as a genus. Cockerell, however, did not grant it
more than subgeneric status, regarding the venation characters as important.
Cockerell’s view has been adopted by the present writers.

Apis (Synapis) cuenoti Théobald 1937
*1937 Apis cuenott Théobald: 4o1 ; pl. 8, fig. 16, pl. 28, fig. 8.
HorotypPe. Ecole des Eaux et Foréts, Nancy, No. F.173 (Coll. Fliche).
DISTRIBUTION. Oligocene (Stampian), Cereste, France.

Diacnosis. Length of body 13:75mm; eyes pubescent ; hair on body fairly
plentiful ; second m-cu meets M near distal corner of cell second Rs; forewing
length 12°5 mm.

REMARKS. Not so well preserved as some other representatives of Synapis.
Specific rank has been retained for this specimen since it is most difficult to reconcile
the wealth of hair, the pubescence of the eye (compare A. (S.) henshawi kaschket,
p. 243, the extraordinary length of the wing and the distal juncture of second m-cu
with M, with any other Synapis. Théobald compared it with Apis oligocenica (recte
A. (S.) henshawi), and suggested some connection between the two. It is thus likely
that upon re-examination the specimen may turn out to resemble this species more
closely than is at present apparent.

Apis (Synapis) henshawi Cockerell 1907

*1907 Apis (Synapis) henshawi Cockerell : 229.
1907 Apis (Synapis) henshawi Cockerell ; Handlirsch : 1357.
1915 Apis oligocenica Meunier : 210, pl. 21, fig. 4.
1915 Apis henshawi Cockerell; Buttel-Reepen : 14.
1931 Apis (Synapis) henshawi Cockerell ; Statz: 42.
1934 Synapis henshawi Cockerell ; Statz: 3, 6, tf. 1, 4, 5, 6.
1934 Apis oligocaenica Meunier ; Statz: 1.
1938 Apis oligocaenica Meunier ; Armbruster : 88.

HoLotyrPe. Museum of Comparative Zoology, Harvard University.
DISTRIBUTION. Oligocene (Aquitanian), Rott, West Germany.

FOSSIL BEES 241

Diacnosis. Body 12—16mm, forewing 8-5-10mm. Variably hairy, eyes naked.
Second m-cu meets base of cell second Rs on the abscissa of vein M beyond half-
way.

DESCRIPTION. The three subspecies included in this species vary in size, the body
being 12-16 mm. A. (S.) cuenoti would be included in this on the grounds of size,
but its forewing is longer than in any known specimen of A. (S.) henshawt. The
hairiness varies, and the hairs are perhaps sometimes plumose on the vertex. These
bees closely resemble A. mellifera in general appearance.

REMARKS. The first description of this species is by Cockerell (1907). He
examined three specimens from Rott which the Museum of Comparative Zoology,
Harvard, had obtained from the firm of Dr Krantz, Bonn, under the name of Apis
dormitans Heyden 1862. One of these specimens was named Synapis henshawi by
Cockerell, but since the other two bore Apis dormitans labels, he assumed that they
belonged to Heyden’s species, which Cockerell regarded as different from his S.
henshawt.

Curiously enough, Heyden also had used three specimens of the Krantz Collection
for his description of Apis dormitans, so that it seemed possible that Heyden’s speci-
mens were those that were eventually sold to Harvard University. That this is
not so, however, is revealed by the fact that Cockerell was able to describe wings
of all three Harvard specimens, whilst Heyden stated that one of his specimens was
‘without visible signs of wings’. Moreover, the holotype of Apis dormitans Heyden
is in the collection of the British Museum (Natural History), and its re-examination
(p. 251) has confirmed the view held by other authors that Heyden’s species is not
an Apis.

Apis dormitans Cockerell (nec Heyden), based on two specimens in the Harvard
Collection, is a true Apis (Synapis), as was demonstrated by Statz (1931). Hence it
becomes necessary to rename this form, the species name being preoccupied by
Heyden’s form.

Three different Synapis have been described from Rott. They all come from the
same beds and the same biotope. In view of the variation known to exist in large
Apis populations, they are here reduced to subspecific status.

Apis (Synapis) henshawi dormiens subsp. nov.
(Pl. 4, figs 1-5)

*1907 Apis dormitans (nec Heyden) ; Cockerell : 228.
1908 Apis dormitans Cockerell ; Handlirsch : 1357.
1931 Apis oligocenica Meunier ; Statz: 46 (erroneously placed here).
1931 Synapis dormitans (Cockerell) ; Statz: 42, 46, figs 1a, 2b, 3c, 4a, 5-9.
1934 Synapis dormitans (Cockerell) ; Statz: 1, figs 2, 8, Io.
1944 Synapis dormitans (Cockerell) ; Statz : 63, figs I, 3, 6.

HoLotypPe. Museum of Comparative Zoology, Harvard University. (Cockerell’s
first specimen, 1907, described as Apis dormitans.)

PARATYPE. Second specimen ; Museum of Comparative Zoology, Harvard.

242 MONOGRAPH ON

MATERIAL STUDIED. In addition to the holotype and paratype, two specimens
in the British Museum (Natural History), Nos 59634 and In.36655; also all the
specimens in the collection of Dr Statz, Cologne.

DISTRIBUTION. Oligocene (Aquitanian): Rott.

Diacnosis. Body 13 mm, forewing 8-5 mm. Hair on vertex not plumose, eyes
naked. Basitarsus with g rows of bristles, middle row numbering 24. Vein cu-a
a little proximad of junction of basal vein with Cu.

DESCRIPTION. The medium-sized form of the species. In addition to the
characters stated in the diagnosis, the following are worth noting: the flagellum is
not narrowed at the base, the basitarsus is 1-6 mm long and 0-9 mm wide and the
depstum only moderately developed.

REMARKS. Numerous other specimens exist, mostly collected by Dr Statz, to
whom we owe the detailed description and excellent reproduction of specimens. In
1931 Statz thought that Meunier’s A fvs oligocenica was synonymous with his Synapis
dormitans, but after the study of Meunier’s type he correctly synonymized it with
S. henshawt (q.v.).

Apis (Synapis) henshawi henshawi Cockerell 1907

*1907 Apis (Synapis) henshawi Cockerell : 229.
1915 Apis oligocenica Meunier : 210.
1934 Synapis henshawi Cockerell; Statz: 5, figs I, 4, 5, 6, 7.
1944 Synapis henshawi Cockerell; Statz : 63, figs 2, 5.

HoLotypPe. Museum of Comparative Zoology, Harvard University.

OTHER MATERIAL. Holotype of Apzs oligocenica Meunier, Heimatmuseum,
Siegburg an der Lahn. Five specimens in the Statz Collection.

DISTRIBUTION. Oligocene (Aquitanian): Rott.

DiaGNosis. Body 15-16mm, forewing 10mm. Hair on vertex plumose or
simple, eyes naked. Basitarsus with 10 rows of bristles (second very short), middle
row numbering about 28. Vein cu-a coincident with junction of basal vein with Cu.

DESCRIPTION. This is the largest of the three subspecies. Making allowance for
the distended abdomen (measured length 16-5 mm), Cockerell inferred a length of
I5 or 16mm. When Statz encountered the holotype of Apis oligocenica Meunier
in the Siegberg Museum, he satisfied himself that it was conspecific with A. (S.)
henshawi, and he described it in detail under that name in 1934. There appears
indeed to be a close agreement between Cockerell’s and Meunier’s types, except that
Statz states positively that the hair on the vertex is not plumose, whilst in Cockerell’s
specimen ‘this hair appears to be plumose’. Evidently, Cockerell did not feel quite
sure, and since the hair is simple in all other Synapis, he may have misinterpreted
his specimen.

The wings are strikingly short for the size of the body. The antenna cleaner is
deeply incised, semicircular, and thus differs from the shallower type present in
A. (S.) dormiens. The basitarsus has ten rows of bristles, whilst in A. (S.) dormiens

FOSSIL BEES 243

only nine appear to be present, and the number of bristles comprising the middle
rows is slightly larger.

Apis (Synapis) henshawi kaschkei (Statz 1931)

*1931 Synapis kaschkei Statz: 50, figs Io, 11.
1934 Synapis kaschkei Statz; Statz : 3, figs 3, 9.

Hototype. In Statz Collection, Cologne.
DISTRIBUTION. Oligocene (Aquitanian) : Rott.

DiaGnosis. Body 11-12 mm, forewing 9:25 mm. Basitarsus broad, and shorter
than in any known Apis. Vein cu-a very proximad of junction of basal vein with Cu.

DESCRIPTION. This is the smallest bee from Rott. It differs from the other two
Synapis in the structure of the basitarsus which is very short and almost square
(I-2 mm long and 1-0 mm broad). Statz, in saying that this basitarsus almost agrees
with that of Apis iridiae Friese from Sumatra, is, however, mistaken, as this species
has a pronouncedly oblong basitarsus. The depstum closely resembles that of
Recent Apis mellifera Linn., but the wing venation places this form in the genus
Synapis. There appear to be nine rows of bristles on the basitarsus, including the
second row which is very short. It has a smaller body and a longer forewing than
the other two Synapis.

REMARKS. Only three specimens are known, two of which, including the holotype,
have been seen by one of us. Its status may have to be reconsidered when more
material becomes known. For the time being, Statz’s description cannot be im-
proved upon. He figures the holotype (1931: figs 10, 11) and a second specimen
in his own collection (1934 : figs 3, 9), the latter figure being the diagram of the fore-
wing venation derived from the two forewings of this specimen.

Subgenus APIS Linnaeus 1758

*1758 Apis Linnaeus: 343, 375.
1938 Hauffapis Armbruster : 43 (Genotype: Hauffapis scheuthlei Armbruster).
GENOTYPE. Apis mellifera Linnaeus 1758 (= Apis mellifica Linnaeus 1766),
designated by Latreille (1810 : 439).
DiaGnosis. Third segment of flagellum shorter, but not broader, than the remain-
ing segments ; the broadly rectangular, neatly curved basitarsus always possesses a
distinct auricle ; cross vein cu-a well proximad of the junction of basal vein and Cu.

REMARKS. This subgenus contains, apart from Apis mellifera, with its race
A.m. indica Fabr., the large Apis dorsata Fabr. of India and the small A. florea Fabr.
of India.

The fossil forms from the Pleistocene belong to the A. mellifera group, and those
from the Miocene also are rather closer to this Recent species than to either of the
other two.

244 MONOGRAPH ON

Apis (Apis) armbrusteri Zeuner 1931

1929 ‘Scheuthle Biene’ Armbruster : 4.

1931 ‘Scheuthle Biene’ Statz: 44, 46.
*1931 Apis armbrusteri Zeuner : 1292, fig. 21, pl. 8, fig. 1.

1938 Hauffapis scheuthlec Armbruster : 43, 92, figs 1-5, 7, 8, 14, 20, 23, 26, 29, 33, 35, 38-40,

51, 53, 55, 58, 63-65, 67, 70-72, 76, 77, 79-
Hototyre. Wiirttembergische Naturaliensammlung, Stuttgart, Boettingen Col-

lection No. 47, individual No. 3 (head, thorax and abdomen). Cast: British
Museum (Natural History) In.38859 (thorax) and In.38856 (abdomen).

PARATYPES. Wurttembergische Naturaliensammlung, Stuttgart, Boettingen
Collection No. 47, individuals Nos 1, 2, 4-17. Casts: British Museum (Natural
History) In.38857 (thorax, No. 1), In.38858 (thorax, No. 2), In.38860 (thorax,
No. 4).

OTHER MATERIAL. The specimens described by Armbruster (1938) from the
Upper Miocene of the Randecker Maar, all here attributed to Apis armbrustert as
a species. Subspecific distinctions have, however, been retained as will be seen
below.

DISTRIBUTION. Upper Miocene, Wirttemberg.

Diacnosis. A bee resembling Apis mellifera in size and appearance, with wax
mirrors on the abdomen, the posterior edges of the sternites being apparently
straight. Wings proportionately slightly broader than in the Recent species, position
of the junction of the second m-cu with M, and shape of the first m-cu variable.

REMARKS. In 1931, one of us (F. E. Z.) described a fossil swarm of honey bees
from the thermal limestone of Béttingen on the Swabian Alb, Wiirttemberg, under
the name Apis armbrusteri. It had perished in peculiar circumstances, having
fallen into a fissure, on the edge of the crater of an Upper Miocene volcano, from
which a spring was issuing and which released carbon dioxide into the atmosphere.
Whilst these specimens afforded many details of the body structure, the wing vena-
tion could not be studied.

After the termination of the eruptions, but still in the Upper Miocene, lakes had
formed in the rather larger crater of the Randecker Maar which is situated in the same
volcanic area. Shales deposited in these lakes produced an abundant insect fauna,
again probably because either the water or the atmosphere was temporarily poisoned.
In 1926 Wilhelm Scheuthle discovered bees here.

In 1928 he was joined by Professor Armbruster, and their search resulted in the
recovery of a very large number of specimens which, however, were not scientifically
described until 1938, under the names of Hauffapis scheuthler, H. scheert and H.
scharmanni Armbruster, with a number of subspecies. These bees provide valuable
information about venation, being compressed in much the same way as the bees
from Rott. The variation observed is great, but no clear line can be drawn justifying
the maintenance of species. It is altogether very unlikely that in an area of 50 miles
there should have existed simultaneously four hardly distinguishable species of
honey bees with four more subspecies. On geographical, stratigraphical and mor-
phological grounds, therefore, the Béttingen and Randeck bees should be regarded

FOSSIL BEES 245

as one variable species. The specific names given by Armbruster have been allotted
subspecific rank, as they may be useful in the discussion of the variation, whilst
Armbruster’s subspecies have been synonymized with his species. In essence, this
appears to represent not only ours, but Professor Armbruster’s views, who in 1938
stated that he considered the Randeck bees identical with Apis armbrusteri Zeuner.
The new names were perhaps, in part, prompted by his wish to honour some of his
collaborators and colleagues.

All specimens so far recovered at Randeck and Bottingen are workers, except for
a single fragment of a drone from Randeck. The following four subspecies may be
distinguished.

Apis (Apis) armbrusteri armbrusteri Zeuner 1931

*1931 Apis armbrusteri Zeuner : 292, fig. 21, pl. 8, fig. 1.
1934 Apis armbrusteri Zeuner ; Statz: 7.
1938 Apis armbrusteri Zeuner ; Armbruster : 16, 45.

HOLOTYPE AND PARATYPES. See under species, above.

DISTRIBUTION. Miocene, ‘Béttinger Marmor’, Béttingen, Swabian Alb, Wtirttem-
berg.

DiaGnosis. Head 2mm, thorax 5mm, abdomen 12mm (distended). Meso-
thorax strongly vaulted, middle segment only slightly sloping. First abdominal
segment very short, vertical and concave anteriorly. Posterior edges of abdominal
sternites not or very little angular at mid-line.

DESCRIPTION. On a piece of red tufa measuring not more than 45 x35 x10 mm,
17 individuals are preserved in a layer and ina closely packed condition. The insects
appear as undistorted cavities and have to be studied with the aid of casts made in
an elastic material. The cavities have been broken in various positions, the
abdomina (being the largest) being available most frequently. In addition, 5
thoraces and 3 heads can be studied.

Head. The eyes appear to have been oval, not kidney-shaped. They are
separated on the vertex, hence the specimens may be regarded as workers.

Thorax. Almost spherical, strongly vaulted above. Prothorax short, the
shoulder-buckles not reaching the tegulae. Shoulder-buckles prominent laterally
in an angular fashion. Mesopleurae curved, about twice as high as long. Middle
coxae long and fitted into the thorax, not pointing obliquely backwards, but vertical
to the longitudinal axis of the insect. Mesonotum very strongly vaulted, more so
than in A. mellifera, and exceeding the scutellum in height. Scutellum sharply set
off and forming a prominent crosswise flange. Middle segment sloping at a smaller
angle than in A. mellifera or even in A. florea.

Abdomen. Stalk short, first segment concave anteriorly, short, as seen from
above. The posterior margins of the middle sternites are not angular, but more or
less straight. After careful investigation, this could not be attributed to distension

18

246 MONOGRAPH ON

or distortion of the abdomen ; it thus appears to be a sound diagnostic character.
Wax mirrors present.

REMARKS. The Bottingen bees differ from the three Recent species of Apis in
the shape of the thorax, whose mesonotum is so strongly vaulted that it exceeds
the scutellum in height, a feature that may be interpreted as a sign of high speciali-
zation. Similarly, the shape of the first abdominal segment suggests specialization
beyond the level of Apis mellifera. In other respects, however, the Bottingen bees
are more primitive than A. mellifera, especially in the shape of the abdominal
sternites, and perhaps in the slight slope of the middle segment of the thorax.

It is much to be regretted that no evidence could be obtained about the collecting
apparatus or the hind legs. Attempts made to expose the legs by carefully grinding
away the matrix proved unsuccessful.

A. armbrusteri was a social bee which in all probability lived much like A. mellifera.
The wax mirrors suggest that combs were built of this substance as is characteristic
of the genus Afis. The fact that 17 individuals were contained in a closely packed
condition in the small piece of rock recovered suggests that a much larger number of
bees perished and were fossilized, a swarm which would have fallen into the death
trap of the Béttingen fissure and perished almost instantaneously in its poisoned
atmosphere.

Apis (Apis) armbrusteri scharmanni (Armbruster 1938)
*1938 Hauffapis scharmanni Armbruster : 44, 113, fig. 78.

HoLotypPe. Armbruster’s specimen and fig. no. 78, in the collection of Dr Hauff,
Holzmaden.

DISTRIBUTION. Upper Miocene, Randecker Maar, Wirttemberg.

DiaGnosis. Approx. 9:Imm long; form compact; head massive; abdomen
plump; corbicula (tibia) very broad but short, much reduced at base; apex of
‘pollen kneader’ (counterpart of auricle, on the tibia) strong and angular; wax
mirrors elongated and hammer-shaped ; vein first R, possibly not arched.

DESCRIPTION. This is a small bee though the respective states of preservation
have to be taken into consideration in comparing it with A. armbrustert armbrustert.
The edge of the tibia, from pollen kneader to the end of the limb, forms almost a
straight line, asin A. a. scheuthlet. Second Rs (third submarginal cell) not narrowed
unduly ; second m-cu joining base of third submarginal cell more towards the
middle ; cubital basis (according to Armbruster) as long as the corbicula, but re-
puted to be shorter than in A. a. scheeri ; first m-cu only a little deflected; Cu,
straight. The tergites of the abdomen are very narrow but very long.

REMARKS. This is in several respects a curious bee and of the several forms here
discussed the one perhaps most deserving of specific rank. Unfortunately the type
appears to be lost ; Armbruster had to use ‘photograms’ as a substitute for the
specimen. Other photographs mentioned by the author appear to be of the same
specimen, and no paratypes are known to exist.

FOSSIL BEES 247

Apis (Apis) armbrusteri scheeri (Armbruster 1938)

*1938 Hauffapis scheert Armbruster : 43, 92, figs 50, 52, 62, 64, 66, 69, 71, 73.
1938 Hauffapis scheert var. gallaunt Armbruster : 45, figs 31, 59.
1938 Hauffapis scheeri var. vahdei Armbruster : 45, fig. 50.

LECTOTYPE. Schempp Collection, Stuttgart-Weil (Armbruster 1938: fig. 71),
herein selected.

PARALECTOTYPES. Collections of Scheuthle and Schempp.

OTHER MATERIAL. Holotypes of Hauffapis scheert gallauni Armbruster, No. 59
and of Hauffapis scheert rahdet Armbruster, No. 50.

DISTRIBUTION. Upper Miocene, Randecker Maar, Wiirttemberg.

DiaGnosis. Smaller, plumper than A. a. scheuthlet, with narrower third sub-
marginal cell, and relatively short, broad and boldly formed hind tibia and basi-
tarsus. Legs very reminiscent of those of Recent A. mellifera.

DESCRIPTION. Short dense hair, partly arranged in lines, occurring on the abdo-
men. The wax mirrors are long and narrow. The first m-cu is distinctly angular
in the holotype. In most respects this form resembles A. mellifera greatly, and
especially in the shape and size of the wing, and in certain details of the venation.

The length of the body varies ; a giant form was given the name of H. s. gallaunt
and a dwarf form H. s. vahdet. There is also much variation in overall size and
proportions. Armbruster gave the average length of the cubital basis as 2-96 mm,
though he illustrated specimens in which this measurement is 3-05, 3:2 and 3:3 mm,
which thus appears to be of little taxonomic value.

REMARKS. Of this very variable subspecies, 24 specimens have been recorded.
Of these, three are only doubtfully placed here. The total includes the ‘giant’ and
‘dwarf’ forms to which Armbruster gave special names.

Apis (Apis) armbrusteri scheuthlei (Armbruster 1938)
1929 ‘Scheuthle Biene’; Armbruster: 4.
4931 Apis-sp.; Zeuner : 296.
*1938 Hauffapis scheuthlec Armbruster : 43, figs 63, 65, 67, 70, 72, 76, 77, 79.
1938 Hauffapis scheuthlei var. seemanni Armbruster : 45, figs 8, 11, 21, 26, 53.
1938 Hauffapis scheuthlet var. zeuneri Armbruster : 45, figs 36, 38.

LEecToTYPE. W. Scheuthle Collection (Armbruster 1938 : fig. 63), herein selected.
PARALECTOTYPES. Armbruster, Scheuthle, Schempp and Hauff Collections.

OTHER MATERIAL. Holotype of Hauffapis scheuthlet seemanni (Armbruster Coll.
no. 8), and paratypes (nos II, 21, 26, 53). Holotype of Hauffapis scheuthlei zeunert
(Armbruster Coll. no. 36), paratype (same collection, no. 28).

DISTRIBUTION. Upper Miocene, Randecker Maar, Wirttemberg.

Diacnosis. Larger than A. a. scharmanni and A. a. scheert, size slightly exceeding
A. mellifera except in dwarf specimens. Third submarginal cell broader than in
A. a. scheert. Legs slender, somewhat resembling those of A. dorsata.

248 MONOGRAPH ON

DESCRIPTION. This is the largest of the three forms of bees from Randecker
Maar here recognized. Though the majority of the specimens are of the size of
large Recent A. mellifera, giant and dwarf forms have been described. A. s. var.
seemanni is a giant form, whilst the smallest specimens were named A. s. var.
zeunert. The body is covered in hair which is somewhat coarse.

Compared with A. a. scheert, the legs of A. a. scheuthlei are longer and more slender.
The hind tibia (corbicula), considered by Armbruster to be slightly curved in outline,
is less narrow towards the base. The basitarsus, too, has a straighter outline.
Armbruster is inclined to regard the legs as reminiscent of A. dorsata.

The first submarginal cell of the forewing often appears arched to meet C; the
second and third submarginal cells are broader than in A. a. scheert ; pterostigma
much reduced ; the first recurrent vein is markedly angular, with a small stump of
a vein at the apex of the angle.

REMARKS. Armbruster distinguished two subspecies, A. s. seemanni and A. s.
zeunert. These are here synonymized with A. a. scheuthler as they are merely
extreme size variants which would be found in any large population. Regarding
one of the specimens assigned to A. s. seemanni (no. 35), Armbruster (1938 : 103-4)
assigned its counterpart to A. a. scheuthlei (no. 33) and withdrew the identification
of no. 35 (1938 : 114), without, however, altering the text on the earlier pages. That
this should have been possible illustrates how small the difference is between these
size classes.

Armbruster attached some importance to the length of the cubital basis, i.e. the
base vein of cells first R,, first Rs and second Rs. He did so presumably because
it is a measure that can be taken on a comparatively large number of wings, includ-
ing those the total length of which cannot be measured. In A. a. scheuthler this
cubital basis is reputed to equal the length of the corbicula. In fact it varies from
3:06 to 3-40 mm. In his A. s. seemannt, the value is 3°67 mm. Of A. s. zeunert no
measurements are given, except for no. 64 (Scheuthle Coll.) which is regarded as
possibly referable to this form. Here it is ‘apparently hardly longer than 3-0 mm’.
The unreliability of this measurement has been clearly demonstrated by Alber
(1949), who studied the asymmetry of the wings of bees.

Altogether 44 specimens have been assigned to this variable ‘species’, which
Armbruster regarded as ‘possibly, nay even probably’ overlapping with Afzs
armbrustert Zeuner. The wax mirrors are preserved on many of the Randecker
Maar specimens, but their shape is not discussed. From the figures it appears that
the margins of the sternites were more or less straight, and the wax mirrors rect-
angular, as in the Béttingen form.

Apis (Apis) melisuga (Handlirsch 1907)
*1907_ (Apidae) melisuga Handlirsch : 893.

HoLotypPe. Collection of Count Bosnianski, with counterpart ; whereabouts
unknown.

FOSSIL BEES 249

DISTRIBUTION. Miocene, Gabbro, Italy.

DiaGnosis. 18mm long. Other features believed to be much as in Apis
mellifera.

DESCRIPTION. Proboscis at least as long as head. Hind leg with distinct bristles
along the edges. Venation not identifiable.

REMARKS. The broadened hind tibia and tarsus strongly suggest a true Ais.
It is impossible to decide, however, how far it resembles the Recent A. dorsata, or
differs from A. a. scheuthler. It is, therefore, retained as a separate species, in the
hope that fresh material will become available in the future.

Apis (Apis) mellifera Linnaeus 1758
(Pl. 4, fig. 6)

*1758 Apis mellifera Linnaeus : 343, 575.
1890 Apis mellifica Linnaeus ; Foord : 94, fig. I.
1909c Apis mellifera Linnaeus ; Cockerell : 317.

SPECIMENS STUDIED. British Museum (Natural History), Palaeontology Dept.,
58516, 58576.
DISTRIBUTION. Pleistocene, East African copal.

DESCRIPTION. The specimens here listed from East African copal do not differ
from Recent A. mellifera in any essential point. Body length of specimen no. 58576
is 11mm; thorax 3:25 mm long and 4mm wide; abdomen 6 mm long and 4mm
wide. Length of wings 8-75 mm.

REMARKS. Several honey bees are known from the copal of East Africa. Foord
(1890) mentioned two specimens side by side. His poor figure suggests the shape
of the basitarsus being identical with A. mellifera.

Cockerell (1g09c) examined two specimens from the Cambridge Collection. They
were at first believed to be enclosed in Yarmouth amber and hence to be of Eocene
age, but he considered it much more likely that sailors had brought them from East
Africa for the Yarmouth trade in amber fossils.

Specimen no. 58576 has been studied in detail. Its wing venation is remarkable
in two respects ; the first submarginal cross vein (first abscissa of Rs) is S-shaped,
and vein first m-cu (first recurrent) appears to lack the usual angularity. The
peculiar shape of the submarginal cross vein is known also in Synapis, as well as in
A. florea and to a lesser extent in A. dorsata. One might, therefore, regard it as a
primitive character, but since it occurs also in a specimen of A. mellifera from
Entebbe, Uganda, it appears to have persisted in some East African bees. The
lack of angularity of first m-cu is believed to be a mere accident of preservation ;
angularity indeed probably does exist, though it may not be very marked, and it is
masked in this specimen by a slight fold in the wing.

The abdomen of the East African copal bee has yellow bands, like its modern
counterpart.

250 MONOGRAPH ON

Family APIDAE incertae sedis
‘Apis’ aquitaniensis de Rilly 1930

*1930 Apis aquitaniensis de Rilly (not seen).
1931 Abeille mellifére ; Alphandéry : 3, fig. on p. 3.
1938 Apis aquitaniensis de Rilly ; Armbruster: 88.
1950 Apis aquitaniensis de Rilly ; de Rilly: 45.

HoLotypPe. Muséum Marseille, No. 5979, as stated by both Armbruster (1938)
and Roussy (personal communication). Repeated enquiries, however, have not
succeeded in eliciting its present whereabouts.

DISTRIBUTION. Oligocene (Aquitanian) : Aix-en-Provence.

Diacnosis. Length of body 16 mm, length of thorax 5 mm, length of abdomen
g mm, length of forewing 10 mm, length of hind leg 7-5 mm.

DESCRIPTION. According to Alphandéry, two antenna are visible on the head,
and one eye is well preserved. The rest of the head is badly crushed but its general
shape is triangular, like heads of bees with mandibles in general. The thorax is
also crushed and has a black depression in the centre. On its edges there are two
orange-coloured ridges. The left wing is beautifully spread and shows all the
characteristics of a honey bee. The third pair of legs is well spread out, and the
tibia shows a broadening suggestive of a pollen basket. Five abdominal segments
can be distinguished.

This description is an almost literal translation from Alphandéry (1931).

REMARKS. This specimen was first mentioned in an article by F. de Rilly (1930)
entitled L’abeille fossile francaise, of which nothing further is known. It may be
identical with an article said to have been written by that author in 1924, which
we have not seen either. Alphandéry’s description is vague. His reference to the
left wing showing ‘all the characteristics of the honey bee’ is perplexing when applied
to an Oligocene species, and his figure is useless. Armbruster, aware of this, was
fortunate enough to receive a facsimile in coloured plastic from F. de Rilly. This,
however, was scarcely the likeness of an Apis.

Through the kindness of Monsieur Roussy we have been able to inspect a similar
facsimile and we agree with Armbruster’s conclusion. But the question arises as
to whether the fault lies with the facsimile.

A paper by de Rilly (1950) appears to answer this question. In it he asserts once
again his claim that the specimen under discussion is an Apis, but with ‘quelques
caractéres archaiques (cellule supplémentaire de l’aile, éperon tibial)’. According
to an enlargement of the wing (published in 1924) and reproduced in facsimile,
however, the supplementary cell is that formed by a cross vein to the marginal
cell, which is thus a cell completely unknown in Apidae. Moreover, the pterostigma
is exceptionally large and deep. The éperon-tibial, which undoubtedly belongs to
the hind tibia (since the tibial spur of the middle leg, being always present in Apis,
affords no diagnostic character) is worth something, in that it rules out the genus
Apis proper. The few positive indications which have been made available in de

FOSSIL BEES 251

Rilly’s work (1950) prove that this specimen is not an Apis, and probably not even
a member of the Apidae.

‘Apis’ dormitans Heyden 1862

*1862 Apis dormitans Heyden: 76, pl. to, fig. 8.
1891 Apis dormitans Heyden ; Scudder : 689 (5927).
1907 Apis dormitans Heyden ; Cockerell : 228.
1907 Apis dormitans Heyden ; Handlirsch : 892.
1908 Apis dovmitans Heyden; Handlirsch : 1357.
1915 Apis (?) dorvmitans Heyden ; Buttel-Reepen : 13.
1915 Osmia dormitans (Heyden) ; Buttel-Reepen : 14.
1915 Apis dormitans Heyden ; Meunier: 210.
1931 Osmia dormitans (Heyden) ; Statz: 42.
1931 Apis dormitans Heyden ; Statz: 42.

Hototype. British Museum (Natural History), 58778.
DISTRIBUTION. Oligocene (Aquitanian) : Rott am Siebengebirge.

REMARKS. According to Carl von Heyden three specimens existed but the first
was very poorly preserved and he was far from happy about placing it in the genus
Apis. The second and third specimens were too meagrely described for recognition.
Buttel-Reepen considered that the whole habitus, the rounding off of the abdomen
and the great breadth of the thorax compared with the breadth of the head,
indicated, not an Afzs, but an Osmia. Meunier considered it to be some kind of
worker bee, but referred to it as a very enigmatical form. All authors agree that it
has not been proved to be an Afis.

The re-examination of the holotype, which was damaged by water during the
war, confirms the views of previous authors. No character is preserved which would
justify placing it in A fis, and the rounded end of the abdomen precludes it from being
included in this genus. The venation, which Heyden also shows in his figure, is too
poor for generic identification.

Two further specimens of ‘A. dormitans’, from Rott, now in the Museum of
Comparative Zoology, Harvard University, were investigated by Cockerell (1907).
They are specifically different from the specimens referred to above but are legitimate
members of Afizs (Synapis). Hence they require a new name, and have been
described as Apis (Synapis) henshawi dormiens n. subsp. (p. 241).

‘Apis’ styriaca Pongracz 1931
*1931 Apis styriaca Pongracz : 105.
1938 Apis styriaca Pongracz; Armbruster : 88.
HoLotypPe. Whereabouts not known.
DISTRIBUTION. Miocene: Parschlug.

REMARKS. A wing fragment only has been referred to, but not described. Its
affinities are wholly obscure, and the name must be treated as a nomen nudum.

252 MONOGRAPH ON

‘Bombus’ carbonarius Menge 1856

1856 Bombus Latreille ; Menge: 26.
*1856 Bombus carbonarius Menge: 27.
1886 Bombus Latreille ; Brischke : 278.
1891 Bombus carbonarius Menge ; Scudder : 690 (5950).
1907. Bombus carbonarius Menge ; Handlirsch : 891.
1909d Bombus carbonarius Menge; Cockerell: 5.
1928 Bombus carbonarius Menge ; Wheeler : 97.
1931 Bombus carbonarius Menge; Salt : 143.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Apparently Menge had five specimens in two blocks of amber, but
only two species were present, which he called B. carbonarius and B. pusillus. No
diagnosis or description was ever published.

‘Bombus’ muscorum Roussy 1937

non 1935a Bombus muscorum (Linnaeus) ; Richards: 73.
*1937 Bombus muscorum Roussy : 58.

DISTRIBUTION. Eocene: Baltic amber.
HoLoTyPE. Whereabouts unknown.

DESCRIPTION. A solitary bee, red and black, and abundantly hairy. No further
details.

REMARKS. Cannot be assigned to any genus on the characters given.

‘Bombus’ pusillus Menge 1856

1856 Bombus Latreille ; Menge: 26.

*1856 Bombus pusillus Menge: 27.
1891 Bombus pusillus Menge ; Scudder : 691 (5954).
1907 Bombus pusillus Menge; Handlirsch : 891.
1909d Bombus pusillus Menge ; Cockerell: 5.
1928 Bombus pusillus Menge ; Wheeler: 96.

DISTRIBUTION. Eocene: Baltic amber.
REMARKS. See ‘Bombus’ carbonarius Menge above.

‘Bombus’ antiquus Heyden 1859

*1859 Bombus antiquus Heyden: 12, pl. 2, fig. 4.
1862 Bombus antiquus Heyden; Heyden: 75.
1891 Bombus antiquus Heyden; Scudder : 690, pl. 31, fig. 4 (5949).
1907 Bombus antiquus Heyden; Handlirsch : 891.
1931 Bombus antiquus Heyden; Cockerell : 309.
1936 Bombus antiquus Heyden ; Statz : 260.

HototyPe. Whereabouts now unknown. Originally in von Dechen’s collection.

FOSSIL BEES 253

DISTRIBUTION. Oligocene (Aquitanian) : Rott am Siebengebirge.

REMARKS. Specimen lacking several parts and even Heyden (1862) considered
it ‘perhaps not different from Osma carbonum’, thus making its generic status very
insecure. There is no useful purpose served in synonymizing it with the latter
species, since Heyden’s statement is far too casual, and possibly had little real
significance. No subsequent author has considered the specimen to be ‘good’.

‘Bombus’ grandaevus Heer 1849

*1849 Bombus grandaevus Heer : 96, pl. 7, figs 3a, b.
1856 Bombus grandaevus Heer ; Giebel : 182-183.
1867 Bombus grandaevus Heer; Heer: 5, pl. 3, figs 6, 7.
1891 Bombus grandaevus Heer ; Scudder : 690 (5952).
1895 Bombus grandaevus Heer ; Scudder : 122.
1907. Bombus grandaevus Heer; Handlirsch: 891.
1928 Bombus grandaevus Heer ; Wheeler : 97.
1931 Bombus grandaevus Heer ; Cockerell : 309.
1938 Bombus grandaevus Heer; Armbruster : 87.

Ho.otyPe. Badische Landessammlung ftir Naturkunde, Karlsruhe. Not traced.
(Heer 1849 : fig. 3.)

ANOTHER SPECIMEN. As figured by Heer (1867).
DISTRIBUTION. Miocene: Radoboj.

REMARKS. Cockerell regarded this form as generically unclassifiable. Scudder
thought it was not a Bombus, but possibly belonging to Eucera. Handlirsch (1907)
also believed that it belonged near Eucera but to a new genus. Heer’s illustrations
provide no help. They suggest, however, that two different specimens were in
existence, those figured in 1849 and 1867.

We have made enquiries at Karlsruhe and visited the Museum, but the specimens
cannot be found. It is possible they may come to light when the material packed
away in the cellars during the war is sorted out.

‘Bombusoides’ mengei Motschulsky 1856

*1856 Bombusoides Mengei Motschulsky : 28.
1891 Bombusoides mengei Motschulsky ; Scudder : 691 (5955).
1907 Bombusoides mengei Motschulsky ; Handlirsch : 892.
1909d Bombusoides mengei Motschulsky ; Cockerell: 5.
1928 Bombusoides mengei Motschulsky ; Wheeler : 97.

DISTRIBUTION. Eocene: Baltic amber.

DEscRIPTION. A bumble bee not more than 14 lines (3-175 mm) long. No further
description provided.

REMARKS. On formal grounds this is the genotype of Bombusoi1des Motschulsky.
The holotype is lost. The size given is much too small for any bumble bee.

oe)

254 MONOGRAPH ON

Apidae gen. et sp. indet. 1 (Brischke)

1886 Bombus Latreille ; Brischke : 278.
1891 Bombus Latreille ; Scudder : 690 (5946).
1907. Bombus Latreille ; Handlirsch : 891.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. A single “‘Bombus’ mentioned but not described. Probably either
B. carbonarius Menge or B. pusillus Menge, both nomina nuda (p. 252).

Apidae gen. et sp. indet. 2 (Brischke)

1886 Melipona Brischke : 278.

1891 Melipona (vic.) Brischke ; Scudder : 716 (6175).
1896 Melipona Tosi : 356.

1907. Melipona (vic.) Brischke ; Handlirsch : 892.

DISTRIBUTION. Eocene: Baltic amber.

DESCRIPTION. ‘Venation is reminiscent of the South American genus Melipona
but differs in the structure of the legs.’

REMARKS. Two specimens appear to have been known to Brischke, but the
described specimens are now lost. The comparison with Melipona means little in
view of the existence of the meliponoid Electrapis.

Apidae gen. et spec. indet. 3 (Burmeister)

1832 Tvigona Burmeister : 636.

1836 Tvigona Burmeister : 577.

1856 Tyvigona Burmeister; Giebel: 184.

1891 Tvrigona (?) Burmeister ; Scudder : 733 (6324).
1896 Tvrigona Tosi: 356.

1907 Tvrigona (?) Burmeister ; Handlirsch : 892.

DISTRIBUTION. Eocene: Baltic amber.

REMARKS. Perhaps identical with FElectrapis (Roussyana) palmnickenensis
(p. 233). Specimen lost.

Apidae gen. et sp. indet. 4 (Scudder)

1881 Bombus Scudder : 290.

1883 Bombus Scudder : 280.

1891 Bombus Scudder : 690 (6947).

1907. Bombus Scudder; Handlirsch : 891.

DISTRIBUTION. Oligocene: Florissant.

REMARKS. One of the specimens of bees which Scudder knew from Florissant
was assigned to Bombus without further description or comment.

FOSSIL BEES 255

NOT APIDAE
Lithoblatta lithophila (Germar 18309)

*1839 Musca lithophila Germar : 222.
1864 Blattidium beroldingianum Heer : 300, fig. 8.
1869 Musca lithophila Germar ; Weyenbergh : 256, pl. 34, fig. 2 (no. 6511).
1869 Bombus (?) conservatus Weyenbergh ; Weyenbergh : 259, pl. 34, fig. 7.
1869 Apiaria veterana Weyenbergh ; Weyenbergh : 260, pl. 34, fig. 8 (no. 6480 nec 6431).
1886 Mesoblattina lithophila (Germar) ; Scudder: 464.
1886 Mesoblattina lithophila (Germar) ; Deichmiiller : 6, pl. i, figs 1-6.
1888 Mesoblattina lithophila (Germar) ; Oppenheim : 221, pl. 30, figs 6, 8.
1895 Mesoblattina lithophila (Germar) ; Meunier : 223.
1898 Mesoblattina lithophila (Germar) ; Meunier: pl. 25, figs 76-78; pl. 26, figs 85-86;
pl. 27, figs 92-93.
1898 Naucoris carinata Meunier : pl. 27, fig. 88.
1907 Lithoblatta lithophila (Germar) ; Handlirsch : 530, pl. 46, fig. 7.
1931 Bombus conservatus (Weyenbergh) ; Cockerell : 309.

HototyPe. Whereabouts not known, probably lost.

SPECIMENS STUDIED. Teyler Museum, Haarlem, Netherlands, No. 15324 (Bombus
conservatus) and No. 15305 (Apiaria veterana).

DISTRIBUTION. Upper Jurassic: Solnhofen, Bavaria.

REMARKS. Meunier regarded both B. conservatus and A. veterana as cockroaches
(Mesoblattina). His view was shared by Handlirsch who identified them with
Germar’s Musca lithophila and erected the new genus Lithoblatta. Thus the sup-
posed bees of the Upper Jurassic have proved to be Orthoptera.

The specimens in Haarlem which we have studied support this view. No. 15324
has an orthopteroid habitus, and No. 15305 has long antennae and orthopteroid
tibia. Cockerell, who saw B. conservatus, also held that it is ‘wholly obscure and
there is no reason for regarding it as a bee’.

VI. APPENDIX
by S. F. Morris

Colletidae gen. et sp. indet.
1927 Colletidae; Blair: 139.
MATERIAL. Whereabouts unknown.

DISTRIBUTION. According to Blair the material is ‘perhaps Upper Pliocene,
perhaps Quaternary or probably Sub-recent’. Oil sands ; near Fyzabad, Trinidad.

Ctenoplectrella splendens Kelner-Pillault 1970
*1970a Ctenoplectrella spendens Kelner-Pillault : 13, pl. 1, figs 3, 4.

MATERIAL. Holotype: Geologisch-palaeontologischen Institut der Universitat,
Géttingen. Paratype : No. 49, Zoologisches Museum der Humboldt-Universitat,
Berlin.

DISTRIBUTION. Eocene, Baltic amber.

256 MONOGRAPH ON

Anthidiini gen. et sp. indet.
1972 Anthidiini; Sando: 424.
MATERIAL. Gerard Germs Collection.

DISTRIBUTION. Recent (collected from rocks of Precambrian and Lower
Palaeozoic age).

Osmia sp.
1972 Osmiasp.; Sando: 421.

MATERIAL. United States National Museum ; United States Geological Survey,
Glacier National Park Museum ; Harvard University, Paleobotany Collection.

DISTRIBUTION. Recent (in rocks of Precambrian to Upper Palaeozoic age),
Montana, Wyoming.

Apoidea (? Anthophorinae) sp.

1969 Apoidea (? Anthophorinae) sp. ; Werner: 12, pl. 19, fig. 3.

MATERIAL. United States National Museum 257, B.218.
DISTRIBUTION. Early Pleistocene, Rita Blanca lake deposits.

Probombus hirsutus Piton (ms)

1940 Pyvobombus hirsutus Piton (ms) : 218.
1969b Probombus hirsutus Piton; Kelner-Pillault : 526 (nomen nudum).

DISTRIBUTION. Eocene, Menat, Puy-de-Déme, France.

Trigona (Hypotrigona) eocenica Kelner-Pillault 1970
*1970b Trigona (Hypotrigona) eocenica Kelner-Pillault : 437, figs 1-3.

MATERIAL. Holotype: Geologisch-palaeontologischen Institut der Universitat,
Gottingen. Paratype formerly in the Geological Museum of the University,
Konigsburg ; now in the Humboldt Museum, Berlin.

DISTRIBUTION. Eocene, Baltic amber.

Trigona (Hypotrigona) dominicana Wille & Chandler 1964
*1964 Tvrigona (Liotrigona) dominicana Wille & Chandler : 188, figs 1-5.

MATERIAL. Holotype and five paratypes in the collection of the Illinois Natural
History Survey.

DISTRIBUTION. Oligocene, Dominican Republic.

FOSSIL BEES 257

Trigona (Nogueirapis) silacea Wille 1959

*1959 Trigona (Nogueirapis) silacea Wille : 849, pl. 119, text-figs 1-4.
1964 Tvrigona (Nogueirapis) silacea Wille ; Wille : 120, fig. 1.

MATERIAL. Holotype: 12601. Paratypes: 12602-11. In University of Cali-
fornia Museum.

DISTRIBUTION. Middle Miocene, Simjovel, Chiapas Province, Mexico.

Electrapis minuta Kelner-Pillault 1970
*1970a Electrapis minuta Kelner-Pillault : 16, pl. 2, figs 3, 5B.

MATERIAL. Syntypes: Geologisch-palaeontologischen Institut der Universitat,
Gottingen ; Block 50a, Zoologisches Museum der Humboldt-Universitat, Berlin.

DISTRIBUTION. Eocene, Baltic amber.

Apis catanensis Roussy 1960

*1960 Apis catanensis avolii Roussy : 8, fig. 2.
1969b Apis catanensis avolit Roussy ; Kelner-Pillault : 524.

MATERIAL. One specimen in the private collection of Anastasi and Guiseppe
Avolio, jewellers of Sicily.
DISTRIBUTION. Miocene, near Catania, Sicily.

REMARKS. Roussy used the trinomen A fis catanensis avoli, but since the species
was not split the third name is superfluous.

? Apoidea sp.
1821 Vast numbers of bees and nests; Henslow: 501.

MATERIAL. Whereabouts unknown.
DISTRIBUTION. Parish of Kirk Balaft, Isle of Man; ? Pleistocene.

Not Apoidea
1962 Native bee; Carroll: 264, fig. 1.

MATERIAL. Melbourne University, Geological Dept., M.U.G.D. 3518.

DISTRIBUTION. ? Lower Jurassic, Koonwarra, South Gippsland, Victoria,
Australia.

VII. ACKNOWLEDGEMENTS

The task of an amanuensis is a difficult one, but to undertake the publication of a
manuscript on a subject with which one is unfamiliar, and whose authors one did
not work with or even know, requires a particular courage and dedication. Thus,

258 MONOGRAPH ON

great credit is due to Mr S. F. Morris for his initiative in revivifying the manuscript
and in undertaking much of the editorial work towards its publication. Thanks
are also due to Mr D. L. F. Sealy who prepared the paper for the press, and to
Professor O. W. Richards for his advice.

H. W. Bai
Keeper of Palaeontology

VIII. REFERENCES

[Although every effort has been made to trace all the references cited in the text, in a few cases, indi-
cated below by *, this has unfortunately not been possible; though bibliographic references can be given
in most instances, they must be cited as ‘not seen’ (by the Editor). It is evident that some works which
were available to the authors are not now to be obtained in any British library. The Editor gratefully
acknowledges the assistance of the librarians of the British Museum (Natural History) in his efforts to
trace the missing works. Though not cited in the text, references marked { are known to have been
consulted by the authors during the preparation of this paper. They are included as a contribution to
the general bibliography of the subject. S. F. M.]

ALBER, M. 1949. Asymmetrie der Bienenfliigel. Arch. Bienenk., Berlin, 26: 14-16.
*ALPHANDERY, E. 1931. Tvaité complet d’apiculture. vi+572 pp. Paris.
ARMBRUSTER, L. 1929. Wilhelm Scheuthle +. Bienenpflege, Wiirttemberg, 51 (1) : 5-8.
1938. Versteinerte Honigbienen aus dem obermiocanen Randecker Maar. Arch.
Bienenk., Berlin, 19: 1-48, 97-133, figs I-79.
ASHMEAD, W.H. 1899. Classification of the Bees, or the Superfamily Apoidea. Tvans. Am.
ent. Soc., Philadelphia, 26 : 49-100.
BEQUAERT, J. C. & CARPENTER, F. M. i941. The antiquity of social insects. Psyche,
Camb. 48 : 50-55.
BERTHOLD, A. A. 1827. Latyreille’s naturliche Familien des Thierreichs. viiit+602 pp.
Weimar.
tBiscnorr, H. 1927. Biologie der Hymenopteren. 598 pp. Berlin.
Biair, K. G. 1927. Insect Remains from Oil Sand in Trinidad. Tvans. ent. Soc. London
752187 —TAT.
BRISCHKE, D. 1886. Die Hymenopteren des Bernsteins. Schr. naturf. Ges. Danzig (n.f.)
6 (3) : 278-279.
Brown, R. W. 1934. Celliforma spirifer; the fossil larval chambers of mining bees. J.
Wash. Acad. Sci., Menasha, 24 : 532-539, 5 figs.
1935. Further notes on fossil larval chambers of mining bees. J. Wash. Acad. Sci.,
Menasha, 25 : 526-528.
1941a. The comb of a wasp nest from the Upper Cretaceous of Utah. Am. J. Scz.,
New Haven, 239 : 54-56, 1 pl.
1941b. Concerning the antiquity of social insects. Psyche, Camb. 48 : 105-110.
BuRMEISTER, C. H.C. 1832. Handbuch der Entomologie 1. xvi+698 pp., 16 pls. Berlin.
1836. A manual of Entomology. xii+654 pp., 33 pls. London.
TBuTTEL-REEPEN, H. von 1903. Die Stammesgeschichtliche Entstehung des Buienenstaates
sowie Beitrdge zur Lebensweise der solitdven und sozialen Bienen. 138 pp., 20 figs. Leipzig.
1906. Apisticae. Beitrage zur Systematik, Biologie, sowie zur geschichtlichen und geo-
graphischen Verbreitung der Honigbiene (Apis mellifica L.), ihrer Varietaten und der
ubrigen Apis-Arten. Mutt. zool. Mus. Berl. 3: 117-201, 8 figs.
1915. Leben und Wesen dey Bienen. 300 pp. Brunswick.
Buxton, P. A. 1932. Ancient workings of insects, perhaps bees, from Megiddo, Palestine.
Proc. ent. Soc. Lond. 7 (1) : 2-4.
CARROLL, E. J. 1962. Mesozoic fossil insects from Koonwarra, South Gippsland, Victoria.
Aust. J. Sci., Sydney, 25 : 264-265, 2 figs.

*

FOSSIL BEES 259

CocKERELL, T. D. A. Ig00. Observations on bees collected at Las Vegas, New Mexico, and
in the adjacent mountains. Ann. Mag. nat. Hist., London, (7) 5: 401-416.

1906. Fossil Hymenoptera from Florissant, Colorado. Bull. Mus. comp. Zool. Harv.

50: 33-58.

1907. A Fossil Honey-Bee. Entomologist, London, 40: 227-229.

1908a. Descriptions and Records of Bees, 19. Ann. Mag. nat. Hist., London, (8) 1:

337-344:

1908b. Descriptions and Records of Bees, 20. Ann. Mag. nat. Hist., London, (8) 2:

323—-334-

1g08c. A fossil leaf-cutting bee. Can. Ent., Ottawa, 40: 31-32.

1908d. Some results of the Florissant expedition of 1908. Am. Nat., Salem, Mass.,

42 : 5690-581, 12 figs.

1g09a. Two fossil bees. Ent. News, Philadelphia, 20: 159-161.

1g09b. A Catalogue of the Generic Names based on American Insects and Arachnids

from the Tertiary Rocks with Indications of the Type Species. Bull. Am. Mus. nat. Hist.,

New York, 26: 77-86.

1g09c. Some European Fossil Bees. Entomologist, London, 42 : 313-317.

1g09d. Descriptions of Hymenoptera from Baltic Amber. Schr. phys.-ékon. Ges.

Kénigsb. 50: 1-20, 14 figs.

tg09e. New North American Bees. Can. Ent., Ottawa, 41 : 393-395.

1g09f. Some Additional Bees from Prussian Amber. Schr. phys.-dkon. Ges. Kénigsb.

50: 21-25.

19ita. Fossil insects from Florissant, Colorado. Bull. Am. Mus. nat. Hist., New York,
30 : 71-82.

1911b. Descriptions and Records of Bees, 34. Ann. Mag. nat. Hist., London, (8)
7 : 225-237.

1913a. Some fossil insects from Florissant, Colorado. Proc. U.S. natn. Mus., Washing-
ton, 44 : 341-346, pl. 56.

1913b. Some fossil insects from Florissant. Can. Ent., Ottawa, 45 : 229-233.

1914. Miocene fossil insects. Pvoc. Acad. nat. Sci. Philad. 66 : 634-648, 5 figs.

1917. New Tertiary insects. Proc. U.S. natn. Mus., Washington, 52: 373-384, pl. 31.
1921. Fossil Arthropods in the British Museum. Ann. Mag. nat. Hist., London, (9)
541-545, 5 figs.

1922. Fossils in Burmese Amber. Nature, Lond. 109: 713-714.
— 1923. Two Fossil Hymenoptera from Florissant (Vespidae, Megachilidae). Ent. News,
Philadelphia, 34 : 270-271.
1925a. Descriptions and Records of Bees, 106. Ann. Mag. nat. Hist., London, (9)
16 : 416-423.
1925b. Tertiary insects from Kudia River, Maritime Province, Siberia. Proc. U.S.
natn. Mus., Washington, 68 (5) : 1-16, pls 1-2.
1931. Insects from the Miocene (Latah) of Washington, II. Hymenoptera and Hemip-
tera. Ann. ent. Soc. Am., Washington, D.C., 24: 309-312, pl. 1, figs 2, 4, 8.
1934. Some African Meliponine Bees. Revue Zool. Bot. afr., Brussels, 26: 46-62,
9 figs.
CurtTis, J. 1828. British Entomology 5: pls 195-241 with explanations. London.
Dati, W. H. 1915. A monograph of the molluscan fauna of the Orthaulax pugnax zone of
the Oligocene of Tampa, Florida. Bull. U.S. natn. Mus., Washington, D.C., 90: 1-173,
pls 1-26.
{Davia Torre, K. W. von & FRIESE, H. 1899. Die hermaphroditen und gynandromorphen
Hymenopteren. Ber. naturw.-med. Ver. Innsbruck 24: 96 pp., I pl.
Daty, H.V. 1973. Bees of the Genus Ceretina in America North of Mexico (Hymenoptera :
Apoidea). Univ. Calif. Publs Ent., Berkeley, 74: 1-114, 2 pls.
DEICHMULLER, J. V. 1886. Die Insecten aus dem lithographischen Schiefer im Dresdener
Museum. Mitt. K. miner.-geol. prahist. Mus. Dresden 7: 1-88, pls 1-5.

ol it |

260 MONOGRAPH ON

Dours, R. Z. 1873. Hyménoptéres du Bassin Méditerranéen. Revue Mag. Zool., Paris,
(3) 1: 274-325.
Fasricius, J.C. 1775. Systema entomologiae. 30+832 pp. Flensburg and Leipzig.
1805. Systema Piezatorum. 439+30 pp. Brunswick.
Foorp, A.S. 1890. Notes on a collection of East Coast amber belonging to Mrs Burwood of
Yarmouth. Tvans. Norfolk Norwich Nat. Soc. 5: 92-95, I pl.
tFRIESE, H. 1923. Die europaischen Bienen (Apidae). 456 pp., 33 pls. Berlin and Leipzig.
GERMAR, E. F. 1839. Die versteinerten Insecten Solenhofens. Nova Acta physico-med.,
Breslau and Bonn, 19 (1) : 187-222, pls 21-23.
1849. Ueber einige Insekten aus Tertiarbildungen. Z. dt. geol. Ges., Berlin, 1: 52-66,
pli 2:
GIEBEL, C. G. 1856. Fauna der Vorwelt mit stdter Beriicksichtigung der lebenden Thiere.
511 pp. Leipzig.
HAaANDLIrScH, A. 1906-8. Die fossilen Insekten und die Phylogenie dev vezenten Formen.
1430 pp., 51 pls.
1921. In Schréder, C., Handbuch der Entomologie 3 (7) : 117-304, figs 52-327. Jena.
HEER, O. 1849. Die Insektenfauna der Tertiargebilde von Oeningen und von Radoboj.
Neue Denkschr. Allg. schweiz. Ges. ges. Naturw., Zurich, 11 : 1-264.
1850. Die Insektenfauna der Tertiargebilde von Oeningen und von Teningen und von
Radoboj in Croatien, II. Neue Denkschr. Allg. schweiz. Ges. ges. Naturw., Zurich, 11:
264 pp., 17 pls.
1864. Ueber die fossilen Kakerlaken. Vyjschr. naturf. Ges. Ziivich 9 : 273-302, 1 pl.
1865. Die Urwelt der Schweiz. 622 pp., 11 pls, 368 figs. Zurich.
1867. Fossile Hymenopteren aus Oeningen und Radoboj. Neue Denkschr. Alig.
schweiz. Ges. ges. Naturw., Zurich, 22 : 1-42, 3 pls.
1872. Le Monde primitif de la Suisse. 801 pp., 11 pls, 368 figs. Geneva and Basle.
1876. The Primaeval World of Switzerland, 2. 324 pp., 11 pls, 370 figs. London.
1879. Die Urwelt der Schweiz. 2nd (rewritten and enlarged) ed. xix+713 pp., 12 pls,
417 figs. Zurich.
HENSLow, J. S. 1821. Supplementary observations to Dr. Berger’s account of the Isle of
Man. Trans. geol. Soc. Lond. 5: 482-505.
HEYDEN, C. von 1859. Fossile Insekten aus der Rheinischen Braunkohle. Palaeonto-
graphica, Cassel, 8: 1-15, pls 1-2, figs 1-13.
1862. Gliedertiere aus der Braunkohle des Niederrhein’s, der Wetterau und der Rohn.
Palaeontographica, Cassel, 10 : 62-82, pl. Io.
HUNEFELD, L. 1831. Derselbe iiber Bernsteininsecten. Isis, Leipzig 15: col. 1100 (numb.
in error 2000). [Determinations by C. H. C. Burmeister.]
tJacx, R. W. 1917. Parthenogenesis amongst the Workers of the Cape Honey Bee: Mr
G. W. Onions’ Experiments. Tvans. ent. Soc. London 49 : 396-403, pls 105-106.
fJones, F.W. 1943. Habit and Heritage. too pp. London.
JuRINE, L. 1807. Nouvelle méthode de classer les Hymenoptéves et les Diptéves 1. 319 pp.,
14 pls. Geneva.
TKELNER-PILLAULT, S. 1968. Les Abeilles fossiles. 13th Int. Congr. Ent., Proc., Moscow,
1: 256.
1969a. Abeilles fossiles ancétres des Apides sociaux. Proc. VI Congr. IUSSI, Berne:
85-93.
1969b. Les abeilles fossiles. Memorie Soc. ent. ital., Genoa, 48 : 519-534, Io figs.
1970a. L’ambre balte et sa faune éntomologique avec déscription de deux Apoides
nouveaux. Awnnls Soc. ent. Fr., Paris (n.s.) 6: 3-24, 6 figs.
1970b. Une Mélipone (s.1.) de l Ambre balte (Hym. Apidae). Annis Soc. ent. Fr., Paris,
(n.s.) 6: 437-441, 3 figs.
Kerr, W. E. & MAuLeE, V. 1964. Geographical distribution of stingless bees and its impli-
cations (Hymenoptera: Apidae). Ji N.Y. ent. Soc., New York, 72: 2-8, 4 figs, 6 maps.
Kirsy, W. 1802. Monographia Apum Angliae 1. xxii+258 pp., 14 pls. Ipswich.

tT

if

FOSSIL BEES 261

Kou, F.F. 1894. Zur Hymenopterenfauna Afrikas. Annin naturh. Mus. Wien 9 : 279-350,
pls 13-17.
LATREILLE, P. A. 1802a. Hiustoive naturelle des Fourmis, et recueil de mémoires et d’ observations
sur les Abeilles, les Avaignées, les Fauchers, et autres Insectes. xvit+445 pp.,12pls. Paris.
1802b. Histoive naturelle, générale et particuliéve des Crustacés et des Insectes 3. 468 pp.
Paris.
1803. Nouveau Dictionnaire d’Histoive Naturelle 18. 595 pp., pls 23-33. Paris.
1804. Nouveau LDictionnaive d’Histoive Naturelle 24. Tableaux Méthodiques....
233 pp. Paris.
1809. Genera Crustaceorum et Insectorum 4. 399 pp. Paris.
1810. Considévations générales sur Vordve naturel des Animaux composant les classes des
Crustacés, des Avachnides, et des Insectes; avec un Tableau méthodique de leurs genres,
deposés en familles. 444 pp. Paris.
1825. Familles naturelles du Régne Animal, exposées succinctement et dans un ordre
analytique avec l’indication de leurs genres. 570 pp. Paris.
Lea, A.M. 1925. Notes on some calcareous insect puparia. Rec. S. Aust. Mus., Adelaide,
3: 35-36, 1 pl.
LEPELETIER DE SAINT-FaRGEAU, A. L. M. 1841. Histoive naturelle des Insectes. Hyménop-
téves, 2. 680 pp. Paris.
& AUDINET-SERVILLE, J. G. 1828. In Latreille, P. A., Encyclopédie méthodique.
Histoive naturelle 10. 833 pp. Paris.
TLEUVENBERGER, F. 1931. Eine fossile Biene. Schweiz. Bienenztg, Aarau, 3: 3 pp., 1 fig.
LINNAEUS, C. 1758. Systema naturae.... tioth edition, 1. 824 pp. Holmiae.
MaGretTl, P. 1884. Risultati di raccolte imenotterologische nell’Africa orientale. <Annali
Mus. civ. Stor. nat. Giacomo Doria, Genoa, 21 : 523-636, pl. I.
1895. Imenotteri. Annali Mus. civ. Stor. nat. Giacomo Doria, Genoa, 35: 149-173,
4 figs.
MatFaTtTl, G. 1881a. Bibliografia degli Insetti fossili Italiani finora conosciuti. Atti Soc.
ttal. Sci. nat., Milan, 24 : 89—I00.
T 1881b. Due piccoli imenotteri fossili dell’Ambra siciliana. Atti Accad. Lincei Transunti,
Rome, (3) 5: 80-83, 2 figs.
TMANNING, F. J. 1952. Recent and fossil honey bees: some aspects of their cytology,
phylogeny and evolution. Proc. Linn. Soc. Lond. 163: 3-8, 1 pl.
1960. A new fossil bee from Baltic amber. rth Int. Congr. Ent., Verh., Vienna, 1 : 306-
308, pl. 5, fig. 1.
MENGE, A. 1856. Lebenszeichen vorweltlicher, im Bernstein eingeschlossener Thiere.
Programm ... offentlichen Prifung der Schiiler dey Petrischule : 1-32. Danzig.
MEUNIER, F. 1895. Note sur des empreintes d’Insectes des schistes de Solenhofen. Bull.
Soc. ent. Fy., Paris, 64 : ccxxiii—ccxxiv.
1898. Les insectes des temps secondaires. Avchs Mus. Teyler, Haarlem, (2) 6: 89-149,
pls 1-30.
1914. Nouvelles recherches sur quelques Insectes du Sannoisien d’Aix-en-Provence.
Bull. Soc. géol. Fy., Paris, (4) 14: 187-198, pls 6-7, text-figs I-11.
1915. Uber einige fossile Insekten aus den Braunkohlenschichten (Aquitanien) von
Rott (Siebengebirge). Z. dt. geol. Ges., Berlin, 67 : 205-217, pls 21-25, 6 figs.
1920. Quelques insectes de |’Aquitanien de Rott, Sept-Monts (Prusse rhénane). Proc.
Sect. Sci. K. ned. Akad. Wet., Amsterdam, 22 : 727-737, 2 pls, 8 figs.
MICHENER, C. D. 1944. Comparative external morphology, phylogeny and a classification
of the bees (Hymenoptera). Bull. Am. Mus. nat. Hist., New York, 82 : 151-326, 246 figs.
1965. A classification of the bees of the Australian and South Pacific regions. Bull. Am.
Mus. nat. Hist., New York, 130 : 1-362, 789 figs, 15 pls.
MotscHutsky, V. 1856. Lettres...aM. Ménétriés, 3. Etud. ent., Helsinki, 5: 3-38, 1 pl.
Mouwre, J.S. 1961. A Preliminary Supra-specific Classification of the Old World Meliponine
Bees (Hymenoptera, Apoidea). Studia ent., Rio de Janeiro, 4: 181-242.

ai

262 MONOGRAPH ON

Novak, O. 1877. Fauna der Cyprisschiefer des Egerer Tertiarbeckens. Sber. Akad. Wiss.
Wien 76: 71-96, pls 1-3.

OPPENHEIM, P. 1888. Die Insectwelt des lithographischen Schiefers in Bayern. Palaeonto-
graphica, Cassel, 34: 215-247, pls 30-31.

OusTALET, M. E. 1870. Mecherches sur les insectes fossiles des terrains tertiaires de la
France. Annls Sci. géol. Paris 2 (3) : 1-178, pls 1-6.

PANZER, G. W. F. 1806. Kritische Revision der Insektenfauna Deutschlands 2. Entomo-
logischer Versuch die Jtivineschen Gattungen der Linnéschen Hymenoptern nach dem Fab-
viciusschen System zu priifen. 271 pp., 2 pls. Nuremburg.

Piton, L. 1940. Paléontologie du gisement éocéne de Menat. These, Clermont-Ferrand.

TPLatTH, O. E. 1934. Bumblebees and their Ways. 201 pp., 10 pls. New York.

PonGrAcz, A. 1928. Die fossilen Insekten von Ungarn. Aumnls hist.-nat. Mus. natn. hung.,
Budapest, 25 : 91-1094, 55 figs.

1931. Bemerkungen tiber die Insektenfauna von Oeningen nebst Revision der Heer’schen
Typen. Verh. naturh.-med. Ver. Heidelb. 17: 104-125, pls 3-11.

TRicHarps, O. W. 1935a. Bombus muscorum (Linnaeus) and B. smithianus White (Hym.).
Trans. Soc. Br. Ent., Southampton, 2: 73-85, pl. 5.

1935b. Notes on the nomenclature of the Aculeate Hymenoptera, with special reference
to British genera and species. Tvans. R. ent. Soc. Lond. 83 : 143-176.

*RILLY, F. DE 1930. L’abeille fossile francaise. Apiculteur. [Not seen by authors.]

> 1950. [title ?] Ruchers (10) : 45.

RopeEck, H. G. 1938. Type specimens of fossils in the University of Colorado Museum.
Univ. Colo. Stud. gen. Ser., Boulder, 25 : 281-304.

RoseE.u1, F. L. 1939. Apuntes de geologia y paleontologia Uruguayas y sobre insectos del
Cretaceo del Uruguay o descubrimentos de admirables instintos constructivos de esa época.
Boln Soc. Amig. Cienc. nat. Kraglievich-Fontana, Nueva Palmira, 1 (2) : 32-102, 32 figs.

Ross, H. H. 1936. The ancestry and wing venation of the Hymenoptera. Ann. ent. Am.,
Washington, D.C., 29: 99-111, 2 pls.

*Roussy, L. 1937. Flore et Faune de l’ambre fossile de la Mer Baltique. Gaz. apic., Mont-
favet, 388 : incl. p. 66 & fig. 13 (March 1937).
1960. Insectes et abeilles fossiles de l’ambre de Sicile. Gaz. apic., Montfavet, 635 : 5-8,

2 figs.
SALT, G. 1931. Three Bees from Baltic Amber. Bernstein-Forsch., Berlin and Leipzig,
2: 136-147.

SANDHOUSE, G. A. 1943. The type species of the genera and subgenera of bees. Proc. U.S.
natn. Mus., Washington, 92 : 519-668.

SanpDo, W. J. 1972. Bee-nest pseudofossils from Montana, Wyoming and South-West
Africa. J. Paleont., Tulsa, 46: 421-425, fig. I.

SCHLETTERER, A. 1891. Hymenoptera in expeditione sub auspicio regii imperii Belgici
perfecta in regione Africae ad Congo flumen inferius collecta, determinata sive descripta
ab. Annls Soc. ent. Belg., Brussels, 35 : 1-36, pls 1-2.

SCHMIEDEKNECHT, O. 1930. Die Hymenopteren Nord- und Mitteleuropas. x-+1062 pp.,
127 figs. Jena.

SCHRODER,C. 1913-25. Handbuch der Entomologie 3. viii+1201 pp., 1040 text-figs. Jena.
[See Handlirsch 1921.]

tScuuiz, W. A. 1907. Alte Hymenoptera. Berl. ent. Z. 51: 303-333.

ScHUTZE, E. 1907. In Branca, W. & Fraas, E., Die Lagerungsverhaltnisse Bunter Breccie
an der Bahnlinie Donauworth-Treuchtlingen und ihre Bedeutung fiir das Riesproblem.
Phys. Math. Abh. K. Akad. Wiss. Berlin 2: 25-26, 1 pl., figs 22-23.

SCHWARZ, H.F. 1948. Stingless bees (Meliponidae) of the Western Hemisphere. Bull. Am.
Mus. nat. Hist., New York, 90: 1-546, pls 1-8, 87 figs.

SCUDDER, S. H. 1881. The tertiary lake basin at Florissant, Colorado, between South and
Hayden Parks. Bull. U.S. geol. geogr. Surv. Territ., Washington, 6 : 279-300.

FOSSIL BEES 263

ScuDDER, S. H. 1883. The tertiary lake basin at Florissant, Colo., between South and
Hayden Parks. fep. U.S. geol. geogr. Surv. Territ., Washington, 12 : 271-293.

1885. Myriopoda, Arachnoidea und Insecta. Jn Zittel, K. A., Handbuch der Palaeonto-

logie : 723-893, figs 894-1109. Munich and Leipzig.

1886. A Review of Mesozoic Cockroaches. Mem. Boston Soc. nat. Hist. 3: 439-485,

pls 46-48.

1887. Partie 1, Mollusca et Arthropoda. Im Zittel, K. A., Tvaité de Paléontologie 2.

897 pp., 1126 figs. Paris, Munich and Leipzig.

T 1890. A classed and annotated bibliography of fossil insects. Bull. U.S. Geol. Suvv.,
Washington, 69: IoI pp.
1891. Index to the known fossil insects of the world including myriapods and arachnids.
Bull. U.S. Geol. Surv., Washington, 71 : 744 pp.
1895. The Miocene Insect-fauna of Oeningen, Baden. Geol. Mag., London, (4) 2: 116-
122, ‘pl.)6.
SMITH, F. 1854. Catalogue of Hymenopterous Insects in the... British Museum 2: 199-465.
London.
1858. Catalogue of the Hymenopterous Insects. J. Proc. Linn. Soc., London, 2: 42-130,
pls-a;.2.
SNODGRASS, R. E. 1935. Principles of Insect Morphology. ix+667 pp., 319 figs. London.
SPINOLA, M. 1808. Insectorum Liguriae species novae aut rvariores,...[&c.] 2. ii+262+

Vv pp., 5 pls. Genoa.

1838. Compte-rendu des Hyménopteres, recueillis par M. Fischer pendant son voyage
en Egypte. Amnls Soc. ent. Fr., Paris, 7: 437-546.

STADELMANN, H. 1895. Beitrage zur Kenntniss der Gattung Melipona sens. lat.
Sber. preuss. Akad. Wiss., Berlin, Jahrg. 1895 : 615-623, 3 figs.

StTaTz,G. 1931. Eine neue Bienenart aus Rott am Siebengebirge. Wuss. Mitt. Ver. Natur-u.
Heimatk. Kéln 1: 39-60.

1934. Neue Beobachtungen iiber fossile Bienen aus dem Tertiar von Rott am Siebenge-

birge. Ayvch. Bienenk., Berlin, 15 (1) : 1-10, 10 figs.

1936. Ueber alte und neue fossile Hymenopterenfunde aus den tertiaren Ablagerungen

von Rott am Siebengebirge. Decheniana, Bonn, 93 : 256-312, pls I-13.

= 1943. Bienen, Blatter und Bliiten dem rheinischen Braunkohlenwald. Rhein. Heimatphl.,
Diisseldorf, 13 (for 1941): too—112, 14 figs. Also (same title) Dtsch. Imkerfihrer 16
(for 1942): 6 pp., 22 figs. [These works, not seen by the authors, are desiderata in that they
may contain new taxonomic names. ]

. 1944. [title ?] Umschau, Frankfurt-a.-M., 48 (4) : 63, figs 2, 5.

THEOBALD, N. 1937. Les insectes fossiles des terrains oligocénes de France. 473 pp., 29 pls.
Nancy.

Timon-Davip, J. 1943. Premiére étude des Insectes fossiles oligocenes du basin de Mar-
seille. C.r. hebd. Séanc. Acad. Sct., Paris, 216: 455-457.

1944. Insectes fossiles de l’Oligocéne inférieur des Camoins (Bassin de Marseille) II.
Hymenopteres. Bull. Soc. ent. Fr., Paris, 49: 40-45, 5 figs.

Tosi, A. 1896. Di un nuovo genere di Apiaria fossile nell’ambra di Sicilia (Meliponorytes
succini—M. sicula). Rev. ital. Paleont., Bologna, 2 : 352-356, pl. 6, 9 figs.

VIERECK, H. L. 1912. Contributions to our Knowledge of Bees and Ichneumon-flies, in-
cluding twenty-one new genera and fifty-seven new species of Ichneumon-flies. Proc. U.S.
natn. Mus., Washington, 42 : 613-648, 2 figs.

WERNER, F. G. 1969. Jn Anderson, R. Y. & Kirkland, D. W., Terrestrial insects of the
Rita Blanca lake deposits. Mem. geol. Soc. Am., Washington, 113 : 123-130, pl. 19.

WEYENBERGH, H. 1869. Sur les insectes fossiles du calcaire lithographique de la Baviere,
que se trouvent au Musée Teyler. Avchs Mus. Teyler, Haarlem, 2: 247-294, pls 34-37.

WHEELER, W.M. 10926. Les sociétés d’insectes ; leur ovigine, leur évolution. 4068 pp., 61 figs.
Paris. [English translation 1928.]

264 MONOGRAPH ON FOSSIL BEES

tWuitTinc, P. W. 1935. Sex determination in bees and wasps. Jj. Heved., Washington,
26 : 220.
WILLE, A. 1959. A new fossil stingless bee (Meliponini) from the amber of Chiapas, Mexico.
J. Paleont., Tulsa, 33 : 849--852, pl. 110, figs 1-4.
1964. Notes on a primitive stingless bee, Trigona (Noguetrapis) mivandula. Revta Biol.
tvop., San José, 12: 117-151, 21 figs.

& CHANDLER, L.C. 1964. Anewstingless bee from the Tertiary amber of the Dominican
Republic (Hymenoptera ; Meliponini). evita Biol. trop., San José, 12: 187-195, 5 figs.
ZEUNER, F. E. 1931. Die Insektenfauna des Béttinger Marmors. Fortschr. Geol. Palaeont.,

Berlin, 9 : 247-406, pls I-19.
1938. Die Insektenfauna des Mainzer Hydrobien-kalks. Palaeont. Z., Berlin, 20: 104-
151, pls 13-17.
1944. Fossil Honey-Bees. Brit. Bee-Keep. Ass., London: to.
1951. A Discussion of Time-rates in Evolution. Proc. Linn. Soc. Lond. 162 : 124-130.

IX. INDEX

New taxonomic names and the page numbers of the principal references are printed in bold
type.

Aix-en-Provence 250 parietaria 195
amber, see Baltic, Burmese, Sicilian thovacica (thovacicus) 195
Andrena 165-6, 167-9, 171, 175 titania 197-8
clavula 166, 167 tonsa (tonsus) 198
gvandipes 166, 167 vetevana (vetevanus) 198-9
hypolitha 166, 167 ants 202, 219
lagopus 167 Apiaria dubia 186
percontusa 167 vetevana 255
primaeva 167-8, 183; pl. 1, fig. 2 Apidae 206, 207-54
sepulta 166, 167 gen. et spp. indet. 254
tertiavia 202 incertae sedis 250-4
wrisleyt 166 Apini 206, 225, 226-49
spp. indet. 168-9 Apis 177, 202, 213, 226-8, 230, 232, 239,
Andrenidae 165, 166-71 240-2, 243, 244-51; see Synapis
Andreninae 165, 166-70 acervorum 194
Anthidiini 178, 256 adamitica 176-7, 186
Anthidium 178, 179 amalthea 216-7
cuvvatum 179 aquitaniensis 250-1
exhumatum 179 arvmbrusterit 244-6, 247-8
mortuum 178-9, 193 schavmanni 246, 247
scudderi 179 scheert 240, 247
Anthophora 193, 194, 198, 204-5 scheuthlet 246, 247-8, 249
effossa 200, 202 seemanni 248
melfordi 194; pl. 1, fig. 1. zeunert 248
Anthophoridae 187, 188-99 bicornis 183
gen. et spp. indet. 200 catanensis 257
gen. et sp. indet. 256 avolit 257
Anthophorinae 192, 193-9, 256 centuncularis 182
Anthophorini 192, 193, 194-9 cucurbitina 188
Anthophorites 193, 194-5, 196-9 cuenoti 240
gaudryi 195, 197 dormitans 184, 241, 251
longaeva (longaevus) 195-6 dorsata 237, 243, 247, 249
mellona 193-5, 196-7 flovea 237, 245, 249

melona 196 helvola 166

INDEX 265

ividiae 243
malvae 193
manicata 178
meliponoides 225, 227, 229, 231
melisuga 201, 248-9
mellifera 225, 237, 241, 243-8, 249; pl. 4,
fig. 6
indica 243
mellifica 239, 243, 249
oligocenica (oligocaenica) 240-2
palmnickenensis 233
quadricincta 163
vufa 183
styriaca 251
tervestvis 209
tvuncorum 180
violacea 189
SP. 247
Apoidea 161 et seqq.
fossils erroneously classified as 202-3, 257
gen. et spp. incertae sedis 201-2, 256-7
Armbruster, L. 158
auricle 158
A xestotrigona erythra 217
togoensis 217

Baltic amber 159-60, 166, 168, 172-5, 200,
207-8, 214, 228-39, 252-7
Bauckhorn, H. 160
Bedford, R. 205
bees’ nests 203, 256
Biareolina 165, 167, 170
neglecta 165, 167
Blattidium beroldingianum 255
Bombini 157, 206, 209
Bombus 190, 201, 209, 210-3, 226-7, 232,
237, 254
abavus 210-1, 212-3; pl. 2, fig. 2
antiquus 252-3
carbonarius 252, 254
conservatus 255
crassipes 209-10
florissantensis 201, 212, 213
grandaevus 201, 212, 253
jurinet 189-90
lapidarius 213
muscorum 252
proavus 212-3
pusillus 252, 254
tervvestvis 209
Bombusoides mengei 253
Béttingen 160, 245
Burmese amber 160, 223

Burrow, Col. H. 161
Bursey, Mr 229

Calyptapis 201, 209, 212
florissantensis 209, 212
Camoins 160
Catania 257
Celliforma 203, 204-6
bedfordi 204-5; pl. 1, figs 7, 8
favosites 203, 206
germanica 204
nuda 204
septata 205; pl. 1, fig.9; pl. 2, fig. 1
spirifer 203-4
sp. indet. 205
Centris clavipes 221
cornuta 176
Ceratina 166, 188
cucurbitina 188
disrupta 188
Ceratinini 188
Cereste 240
Chalcobombus 206-7, 208, 214, 227-8, 232
group 206
hirsutus 207
humilis 206-7, 208
martialis 208
Chalicodoma 182-3
Chiavone 201
Clavicera 188
Cockerell, T. D. A. 159
cockroaches 255
Colletidae 161-2, 255
copal 160, 217-21, 249
Corent 195-6
Crematogaster 219
Ctenoplectra 172
Ctenoplectrella 172, 173
dentata 172
splendens 255
viridiceps 172, 173
Ctenoplectrinae 172, 173-4
Cyrtapis 163
anomalus 163

Dasypoda 171, 175
depstum 158, 230
Diandrena 170
Dianthidium 179, 180
curvatum 179
tertiarum 180
Dominica 257

266 INDEX

East African copal 160 halictinus 181
Electrapis 206, 225-7, 228-39, 254; see laminarum 181, 182
Protobombus, Roussyana mersatus 181
apoides 226, 227-9, 237 mildredae 181-2
meliponoides 225, 227, 229-31, 233, 237; priscus 182
pl. 3, figs 5, 6 saxosus 182
minuta 257 truncorum 180
tornquisti 231, 237 Hermann, P. 159
Eucera 178-9, 201, 212, 253 Hesperapis rhodoceratus 170
antennata 193 Hukong Valley 160, 223
mortua 178-9, 193 Hylaeus albilabris 188
Eucerini 192-3 Hypotrigona 218, 219-21
Euglossa 237 bottegot 218
Euglossini 206 dominicana 256

eocenica 256
gvibodoi 218-21; pl. 2, figs 5-7
Florissant 161, 163 et seqq.
Formicinae 202
Fraas, O. 158 Klebs, R. 159
Krantz, F. 160
Krottensee 209
Gabbro 201, 249

Gard 193
Glyptapis 172, 173, 174 larval chambers 203-6, 256
fuscula 173 Lerida 155
mirabilis 173-4 Lestyvimelitta 206, 213, 217
neglecta 174 Libellulapis 170, 171
veducta 174 antiquorum 170
veticula 174 wilmattae 170, 171
veticulata 174 Liotrigona bottegoi 218
Lithandrena 165, 169
antiquorum 169
Halictidae 162, 163-5 saxorvrum 169
Halictinae 162-3, 164-6 Lithanthidium 180
Halictoides maurus 170 pertriste 180
Halictus 163, 167-8, 171, 175, 198 Lithoblatta lithophila 255
florissantellus 164 Lithophaga nuda 204
miocenicus 164 Lithurge 176, 177-8
vuissatelensis 164 adamitica 176-8
schemppi 164-5 cornutum 176
scudderiellus 164 Lithurginae 176, 177-8
Haplomelitta 174 Lithurgus 176-7
Hauff, B. 159 adamiticus 177
Hauffapis 239, 243-4 Luke Thomas Collection 160

scharmanni 244, 246
scheevt 244, 247

gallaunt 247 Macrocera 193
vahdei 247 Mainz 159
scheuthlet 243-4, 247 Manning, Dr F. J., obituary 156
seemanni 247-8 Megachile 179-80, 182, 183
zeunert 247-8 amaguensis 182
Heer, O. 161 centuncularis 182
Heriades 180, 181-2 praedicta 183

bowditchi 181 sp. indet. 183

INDEX 267

Megachilidae 172, 175, 176-87 Parschlug 251

gen. et sp. indet. 186 Pelandvena 165, 169, 170
Megachilinae 176, 178, 179-86 veducta 169, 170
Megachilini 180, 181-6 Perdita 170
Megilla pilipes 194 Piski 202
Melectini 192, 199 Podalirius 194, 204
Melipona 213, 224, 254 Polistes 203

togoensts 217 kirbyanus 202-3
Meliponini 206, 213, 214-25 Probombus hirsutus 256
Meliponorytes 216, 223 Prohalictus 163

devictus 222-4 schemppi 163-4

sicula 216, 224 Protobombus 227, 231-2, 233

SUCCINI 216-7, 221 indecisus 231, 232-3
Meliponula 217 tristellus 233

boccandei 217 Protomelecta 199
Melissodes 212 brevipennis 199

Melitta 175
tricincta 175
willardi 175

Melittidae 171, 172-5 Radoboj 159, 195, 253

Melittinae 174-5 Randecker Maar 158, 164, 244, 246-8
Menat 256 Rohwer, S. A. 159

Mesoblattina lithophila 255 Ross, H. 157

Rott am Siebengebirge 157, 178, 184, 189,
200, 202, 240, 242-3, 251
Roussy, L. 160
Roussyana 227, 233, 234-9
palmnickenensis 226, 233-6, 237, 254;
pl. 3, figs 1, 2
proava 228, 236-9; pl. 3, figs 3, 4

Mexico 257
Michener, C. D. 156, 161, 192
Musca lithophila 255

native bee 257
Naucoris carinata 255
Noguetrapis silacea 257

Nomada 167
Nomia 169 Scheele Collection 160
Scheer, EB. 156
Schempp, K. 159
Oeningen 159, 161, 177, 185-6, 190, 196-8, Scheuthle, W. 158
201, 210 Siberia 182
Orthoptera 255 Sicilian amber 158, 224
Osmia 167, 180, 183, 184-6, 202, 205 Simjovel 257
antiqua 183, 184-5 Snodgrass, R. E. 156
bicornis 183 Solnhofen 255
carbonum 184, 253 Sophrobombus 206, 213-4, 215-6
dormitans 251 fatalis 213, 214-6; pl. 2, figs 3, 4
sp. ? dubia 183, 185-7 South Australia 160
immortua 183, 186 Sphecodes 164
kirbyana 183, 202-3 Sphecoidea 155-6
nigra 183, 185-6; pl. 1, figs 5, 6 Spokane 212
primaeva 167, 183 Statz, G. 160
spp. indet. 186, 256 Swinhoe, R. J. C. 160
Synapis 232, 240, 241-3, 249
cuenoti 240, 241
Panurginae 170, 171 henshawi 237, 240-3
Panurgus 170 dormiens 241-2, 251; pl. 4, figs 1-5

Pavrandrena andrenoides 170 kaschkei 240, 243

268 INDEX

Tetrvagona 216, 221, 222-4 sicula 224-5
elongata 221 togoensis 217-8
ividipennis 222-4; pl. 2, figs 8, 9 Trinidad 255
succini 221-2, 224

Tetragonula 222 Uruguay 205-6

Tetralonia 193 auroranormai 205-6

antennata 193
berlandi 192, 193

- wasp 183, 202
duvaucelii 193 Le sage!

ie nest 203
tibial comb, rake, spurs 157 Wheeler, Dr W. M. 159
Timon-David Collection 160 een oe ee Ey aes

Tosi, A. 158
Trigona 214, 216-7, 218-26, 236-7, 254;
see Hypotrigona, Nogueivapis, Tetragona
amalthea 216-7
braunsi 218
bottegoi 218, 220
devictus 222

dominicana 256 aoe
elongata 221 jurinet 189-90

eocenica 256 senilis 189, 190-2; pl. 1, figs 3, 4

evythra 217-8 violacea 189
togoensis 217 Xylocopinae 187-8, 189-92

gvibodoi 218 Xylocopini 188-9, 190-2

ividipennis 222

wing venation 157

Xilocopa 189

Xylocopa 189, 190-2, 210-1
abavus 210
friesei 189
hydrobiae 189

laeviceps 222-4 Zeuner, Professor F. E., obituary 156
Professor F. E. ZEUNER, S. F. Morris

Pha); D.Sse:, FSA; FG.S be -Zs Department of Palaeontology
F. J. Mannine, Ph.D.t BritisH Museum (NATURAL HIsTory)

CROMWELL ROAD
Lonpon SW7 5BD

Accepted for publication 2 December 1975

EKG aa

le, 2.

FIGS 3, 4.

FIG. 5.
Fic. 6.

BiG. 7:

PLATE 1
Anthophora melfordi Cockerell (p. 194)
Holotype, BM(NH) I.9259. Oligocene, Florissant. x 2°5.

Andrena primaeva Cockerell (p. 167)
Holotype, ETH Zirich SN.202. Miocene, Oeningen. x5.

Xylocopa senilis Heer (p. 190)
ETH Zurich 21. Miocene; Oeningen, x2) x22:

Osmia nigra sp. nov. (p. 185)
Holotype, ETH Zirich Pl. I. 719. Head and thorax. Miocene, Oeningen. x 4°5.
Holotype, ETH Zirich Pl. I. 719. Abdomen. Miocene, Oeningen. x 4:5.
Celliforma bedfordi sp. nov. (p. 204)
Holotype, BM(NH) In.31434. Lateral view. Pleistocene or Sub-Recent, Venus

Bay, South Australia. xo-g.

Fic. 8.

Fic. 9.

Holotype, interior view. Horizon and locality as Fig. 7. x 2:0.

Celliforma septata sp. nov. (p. 205; see also Pl. 2, fig. 1)
Holotype, BM(NH) In.31433. Horizon and locality as Fig. 7. x1.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 3

IPILA II 2

Celliforma septata sp. nov. (p. 205; see also Pl. 1, fig. 9)
Fic. 1. Paratype, BM(NH) In.31435. Longitudinal section. Pleistocene or Sub-Recent,
Venus Bay, South Austraha. x1.
Bombus abavus Heer (p. 210)
Pres 2)) Pt EW Zunich Pils sos = Miocene, Oeningen! 543-38

Sophrobombus fatalis Cockerell (p. 214)
Fre. 3. Mamiburg, Scheele Collection 1225, Kocene, Baltictambers ).<7-5-
Fic. 4. Hamburg, Scheele Collection 1225. Eocene, Baltic amber. x 8.
Trigona (Hypotrigona) gribodoi Magretti (p. 218)

Fic. 5. BM(NH) In.38984. Pleistocene copal, East Africa. x Io.
Fic. 6. BM(NH) In.38984. Pleistocene copal, East Africa. x 12.
Fic. 7. BM(NH) 58511. Pleistocene copal, East Africa. x Io.

Trigona (Tetragona) iridipennis Smith (p. 222)

Fic. 8. BM(NH) In.43809. ? Pleistocene, Burmese amber, Hukong Valley, Burma. x1II.
Fic. 9. BM(NH) In.20702. ? Pleistocene, Burmese amber, Hukong Valley, Burma. x 13.

Bull. Br. Mus. nat. Hist. (Geol.) 27,

Y
Z

PLATE 3

Electrapis (Roussyana) palmickenensis (Roussy) (p. 233)
Figs 1, 2. Hamburg, Scheele Collection 218. Eocene, Balticamber. Fig. 1, x11; Fig. 2,
x 5°5-
Electrapis (Roussyana) proava (Menge) (p. 236)
Fic. 3. Lectotype, BM(NH) In.43592. Eocene, Baltic amber. x5.
Fic. 4. Paralectotype, BM(NH) In.18757. Eocene, Baltic amber. x 4:5.
Electrapis (Electrapis) meliponoides (Buttel-Reepen) (p. 229)
Figs 5, 6. Hamburg, Scheele Collection 238. Eocene, Balticamber. x 7:5.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 3 AE

YI

PIV AGNEe)

Apis (Synapis) henshawi dormiens subsp. nov. (p. 241)

Fre. 1. Statz Collection. ©@ligocene, Rott. 9x42
Fic. 2. BM(NH) In.36655. Oligocene, Rott. x5.
Eig. 3. statz Collections) "Oligocene ixout,) a4"
Fic. 4. BM(NH) 59634. Oligocene, Rott. x 2-6.
Fic. 5. Statz Collection. Oligocene, Rott. x 4.

Apis (Apis) mellifera Linnaeus (p. 249)
PiG..6,, SM(NIE) 53570. leistocene copal, ast, aimicar pays

Bull. Br. Mus. nat. Hist. (Geol.) 27, 3

ar

Re:

oe

os

Io.

3

A LIST OF SUPPLEMENTS
TO THE GEOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya.

Pp. 213; 30 Plates; 2 Text-figures. 1965. OUT OF PRINT.

. Et-NacGar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper

Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley,
Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. 11.

. DAVEY, R. J., Downrz, C., SARJEANT, W. A. S. & WiLLiAMs, G. L. Studies on

Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 Plates; 64 Text-
figures. 1966. £8.20.

. APPENDIX. DAVEY, R. J., DOWNIE, C., SARJEANT, W. A. S. & WILLIAMS, G. L.

Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 24.
1969. 95P.

. Erxiotr, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of

the Middle East. Pp. 111 ; 24 Plates; 16 Text-figures. 1968. OUT OF PRINT.

. RHODES, F. H. T., Austin, R. L. & Druce, E. C. British Avonian (Carboni-

ferous) Conodont faunas, and their value in local and continental correlation.
Pp. 313; 31 Plates; 92 Text-figures. 1969. £13.10.

. Cuitps, A. Upper Jurassic Rhynchonellid Brachiopods from Northwestern

Europe. Pp. 119; 12 Plates; 4o Text-figures. 1969. £5.25.

. Goopy, P. C. The relationships of certain Upper Cretaceous Teleosts with

special reference to the Myctophoids. Pp. 255; 102 Text-figures. 1969.
7:79

£7.70.
. OWEN, H.G. Middle Albian Stratigraphy in the Anglo-Paris Basin. Pp. 164 ;

3 Plates; 52 Text-figures. 1971. £7.20.

. SIDDIQUI, Q. A. Early Tertiary Ostracoda of the family Trachyleberididae

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. £9.60.
Forey, P. L. A revision of the elopiform fishes, fossil and Recent. Pp. 222;
92 Text-figures. 1973. £11.35.

Wit.iAms, A. Ordovician Brachiopoda from the Shelve District, Shropshire.
Pp. 163; 28 Plates; 11 Text-figures; 110 Tables. 1974. £12.80.

Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BSg 5NU

MISCELLANEA

Pp 271-345 ; 12 Plates; 12 Text-figures

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
GEOLOGY Vol, 27, No: 4
LONDON: 1976

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

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

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

These papers form Vol. 27, No. 4 of the Geological
(Palaeontological) series. The abbreviated titles of
periodicals cited follow those of the World List of
Scientific Periodicals.

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

ISSN 0007-1471

© Trustees of the British Museum (Natural History), 1976

BRITISH MUSEUM (NATURAL HISTORY)

Issued 28 October, 1976 7 Price £7°95

CONTENTS
Fusella M’Coy 1844, a problematic brachiopod genus from the Lower
Carboniferous. By C. H. C. BRUNTON & A. RISSONE . : :

The Jurassic Chaetetid, Blastochaetetes bathonicus J. C. Fischer, in England.
By G: f. ELLioir ; : : ; : ; : :

Coprolites of Ptlophyllum cuticles from the Middle a of north
Yorkshire. By C. R. Hix : ; :

A new fossil freshwater crab from the Ngorora Formation (Miocene) of
Kenya. By S. F. Morris

Cyclothyris (Cretaceous Brachiopoda) from California. By E. F. OWEN

Cranial material of Oligocene and Miocene flamingos: with a description of
a new species from Africa. By C. J. O. HARRISON & C. A. WALKER

A new fossil pelican from Olduvai. By C. J. O. Harrison & C. A. WALKER

Chelidonocephalus trilobite fauna from the Cambrian of Iran. By R. A.
ForTEY & A. W. A. RUSHTON : ; ; :

Index to Vol. 27 no. 4 .

Page
275
285

289

295
301

395
315

321

341

FUSELLA M’COY 1844, A PROBLEMATIC
BRACHIOPOD GENUS FROM THE
LOWER CARBONIFEROUS

By C. H. C. BRUNTON & A. RISSONE

ABSTRACT

The genus Fusella is redescribed and assigned to the subfamily Strophopleurinae. Related
or similar species are discussed.

INTRODUCTION

THE GENUS name Fusella, first published in 1844 by M’Coy for small transversely
fusiform spiriferide brachiopods, has been poorly known and ill-used for about one
hundred and thirty years. We redescribe the type specimen of the type species,
F. fusiformis (Phillips), and other conspecific and congeneric material in the hope
of establishing the genus Fusella on a more stable basis. Silicified specimens of
F. rhomboidea (Phillips) allow the description of interiors believed to be closely
comparable to those of F. fusiformis. The genus is formally redescribed and we
discuss both its position within the Spiriferacea and those species which, in the past,
have been assigned to Fusella.

SY SLEMATICS

Superfamily SPIRIFERACEA King 1846
Family MUCROSPIRIFERIDAE Pitrat 1965
Subfamily STROPHOPLEURINAE Carter 1974

DIAGNOSIS (emended). Small to medium very transverse Mucrospiriferidae with
simple lateral costae and lirate, subimbricate micro-ornamentation. Fold and
sulcus commonly non-costate, with or without median rib and groove; sulcus
bounding ribs commonly accentuated, as are corresponding dorsal grooves. Ventral
interarea large and denticulate. Dental plates short or buried by shell thickening.
Dorsal sockets small and closely set ; cardinal process commonly medially supported
by short ridge ; shell substance impunctuate.

Discussion. Recently Carter (1974) proposed a classification of the Spiriferidae
which involved the erection of the Strophopleurinae. Into this subfamily he placed
‘Strophopleura Stainbrook, 1947; Alispirifer Campbell, 1961; Acumunothyris
Roberts, 1963 ; Vozseyella Roberts, 1964 (= Amesopleura Carter, 1967) ; ? Eleuthero-
komma Crickmay, 1950; ? Pterospirifer Dunbar, 1955; ? Celstfornix’ Carter
(1974.:677). This subfamily corresponds partially with a new taxon of the authors’
which was in script form at the time of Carter’s 1974 publication. We accept

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4: 275-284 Issued 28 October 1976

276 BRUSELEA M’ COM

Carter’s subfamily but emend its taxonomic position and, importantly for the
purpose of this paper, add the genus F'usella as a firm member of the subfamily.
We believe that Alispirifer Campbell doubtfully belongs here but would suggest
the inclusion of Brachythyrina Frederiks 1929 and Paeckelmanella Likharev 1934.
Roberts (1971) placed Vozseyella in the Mucrospiriferidae but the above grouping
within the Strophopleurinae removes Fusella and Brachythyrina from the
Spiriferidae. Alispirifer, Paeckelmanella and Pterospirifer are removed from the
Licharewiinae and Eleutherokomma from the Acrospiriferinae of the Treatise
(Williams et al. 1965) classification.

Ivanova (1972 : 315) proposed the family Paeckelmanellidae for Paeckelmanella,
Alispirifer, Spiriferinaella and Pierospirifer. The family was placed, with some
reservation, in the Syringothyridacea, within the suborder Spiriferidina, but
neither a full diagnosis nor a discussion was provided. Carter’s subfamily, here
used, partially equates with Ivanova’s family, as can be seen from the generic
constituents. Ivanova, Carter and we ourselves all agree that the taxon containing
Paeckelmanella — and we believe Fusella — belongs within the Spiriferidina, but at
family levels of classification there is no agreement ; Ivanova placed the mucro-
spiriferids within the new suborder Delthyrididina whereas most previous authors
placed them in the Spiriferidina.

Genus FUSELLA M’Coy 1844

TYPE SPECIES. Spirifera fusiformis Phillips 1836: 210; pl. 9, figs 10, 11. By
original designation of M’Coy (1844 : 132).

DIAGNOSIS. Small (commonly less than 30 mm wide) strongly transverse shells
with pointed extremities. High, variably concave, denticulate ventral interarea
extending full width of shell. Lateral profile subcircular. Ventral sulcus bordered
by pair of prominent ribs, dorsal fold variably developed. Lateral ribbing weak
to moderately developed. Dental plates close together within sulcus, subparallel
and with callosity filling apex of delthyrium in large specimens. Crural bases
converge to valve floor posteriorly. Shell impunctate.

Discussion. Although M’Coy (1844) was reasonably precise, for that date, in his
description of Fusella, the name has been ill-used ever since. This is because there
is only one well-known specimen of the type species F. fustformis in existence and
from this it is impossible to learn any detailed information about the internal
morphology. It is this species which M’Coy specified as the type of his new
‘subgenus’ Fusella, characterized as follows.

‘Shell elongate transversely, fusiform, cardinal area wide, much curved; beaks
incurved. This group would embrace these little Spivifers of the mountain
limestone which have a perfectly fusiform outline, the depth being equal to the
length, and the sides cylindrical ; the cardinal area is extremely wide in proportion
to their size and is always hollowed or much curved, thus contrasting with the
narrow, flat area of the typical Spivifers while the strongly incurved beaks
distinguish them from the Cyrtiae. It would include the S. bicarinata, S.
rhomboides, &c. &c.’ (1844 : 132).

L. CARBONIFEROUS BRACHIOPOD ATG]

M’Coy’s description stressed the incurved nature of the ventral umbo, which leads
to the belief that he had not seen Phillips’ actual specimen of F. fusiformis in which
the ventral interarea is almost flat, only being concave medially, close to the umbo.
The umbo projects beyond the hinge line by no more than 1:5 mm; Waterhouse
(1970 : 3) is in error in writing that the ‘umbo extended 2:5 mm beyond the hinge’,
probably because he took his measurement from his published illustrations which
are at almost twice the stated magnification. M’Coy’s description may therefore
have been influenced by specimens from the Cork area of Ireland best assigned to
F. rhomboidea (Phillips). This suggestion is further supported by the size quoted
by M’Coy for F. fustformis, viz. ‘length four lines, width one inch three lines, depth
four lines’ (1844 : 132), a width which is somewhat greater than that of the type
specimen of F. fusiformis.

A review of the confused use of the name Fusella was published by Waterhouse
(1970). He redescribed the type specimen, from the Gilbertson Collection in the
British Museum (Natural History), and compared it with various other species in
an attempt to suggest its affinities. He concluded that Umispirifer Campbell 1957,
with which Fusella has sometimes been synonymized, was distinctive and we agree
with this view. Waterhouse thought that the shell substance of F. fusiformis was
punctate, leading him to discard species such as Spirifer rhomboidea Phillips as being
closely related, but to the conclusion that Fusella was ‘probably related to members
of the Syringothyrididae’ (1970: 6). Both optical and scanning electron-microscope
studies of the type specimen and second undoubted specimen of F. fusiformis show
that the shell is not endopunctate but quite normal for impunctate spiriferaceans
PE) i lies: 20; 2T).

Thomas (1971) is one of the latest of several palaeontologists to say that it seemed
inadvisable to use the generic name Fusella until the type specimen was adequately
known from topotypic material. Had the name fallen from use this would be a
sensible suggestion but in view of its continued appearance in the literature,
commonly quite incorrectly, it is desirable to further Waterhouse’s contention that
it should become a well-known genus in its own right. Following the Russian lead
when Ivanova (1960) placed S. tornacensis de Koninck within Fusella some
palaeontologists, such as Carter (1967), have placed species in Fusella which differ
widely from F. fusiformis. Within their concept the genus is relatively less wide,
very much longer and has a strongly uniplicate anterior commissure. Carter (1971)
described the genus Mirifusella, said to be ‘most similar to Fusella M’Coy’, but in
fact differing considerably in outline and internal features. We have, therefore, a
situation in which some palaeontologists advise the suppression of the name Fusella
and others use this name, at times quite incorrectly. Because of this confusion it
is desirable to correct the use of Fusella to the best of our ability, even if this is done
without resort to additional genuine topotypic material. In the collections of the
British Museum (Natural History) there exists one specimen clearly conspecific with
the type specimen of F. fusiformis. This second specimen is in the Davidson
Collection (B 7379) and came from Dovedale, Derbyshire (Pl. 1, figs 5-7). The
ventral umbo has been broken from this specimen and it is possible to see that the

1 A line or ligne is one twelfth of an inch (=2-1167 mm).

278 PUSELLA NCCOYV

dental plates did not extend anterodorsally to support completely the teeth and
delthyrial margin near the hinge line. Unfortunately searches in the Derbyshire
or Bolland and Clitheroe areas have failed to reveal additional specimens. The
information on internal morphology of Fusella is, therefore, mainly based upon the
silicified Fermanagh specimens collected by Brunton and assigned to F. rhomboidea
(Phillips), a species believed to be closely related to F. fustformis and which originally
also came from Bolland (PI. 1, figs 13-10).

In the last twelve years two genera have been proposed which are probably synony-
mous with each other and also are closely related to Fusella. These are Votseyella
Roberts 1964, with type species Strophopleura anterosa Campbell 1957 from the
Lower Carboniferous of the Werria Basin, New South Wales, Australia, and
Amesopleura Carter 1967, with type species Spirifer novamexicana Miller 1881 from
the Lower Carboniferous (Osagian) of New Mexico. Both these authors thought
that Spirifer mundula Rowley 1893 should probably be placed in their genus.
Subsequently Roberts (1971) has placed Amesofleura into synonymy with Vozseyella.
Having inspected Roberts’ 1964 material from the Greenhills area of New South
Wales and Carter’s 1967 material from the Lake Valley region of New Mexico we
agree with this synonymy. The question then arises as to the relationship of
Voiseyella with Fusella. Neither Roberts nor Carter compared their genera with
Fusella, although the former (1971) discussed the relationship of Fusella with
Umispirifer and the latter placed his new species Jlanoensis (1967) within his concept
of Fusella. These genera are clearly quite closely related ; their dimensions, outlines
and profiles are very similar, as is the form of ribbing. The most important differ-
ence, and that which prevents the synonymy of Vovzseyella within Fusella, is that
the dental plates of Vozseyella diverge from the umbo following the ribs bordering
the sulcus whilst those of Fusella are unusual in remaining subparallel within the
confines of the ventral sulcus (Pl. 1, fig. 19) ; they do not follow the borders of the
sulcus, asin many spiriferaceans. (The dental plates of V. novamexicana also follow
the ribs bordering the ventral sulcus.)

In view of the varied use of the name Fusella, and despite the redescription by
Waterhouse (1970), it seems desirable to provide a description of F. fusiformis
based upon the type specimen and second specimen in the Davidson Collection.

Fusella fusiformis (Phillips)
Pl. 1, figs 5-12
1836 Spirvifera fusiformis Phillips: 217; pl. 9, figs 10, 11.
1849 Spirifera fusiformis Phillips; Brown: 108; pl. 51, figs 4, 5.
1858 Sphirifera fusiformis Phillips; Davidson: 56; pl. 13, figs 15, 15a.
1970 Fusella fusiformis (Phillips) ; Waterhouse : 3; figs 1A-F.

TYPE SPECIMEN. The single specimen described and figured by Phillips (1836)
in the Gilbertson Collection, British Museum (Natural History), B 249, from
Bolland, Yorkshire.

DIMENSIONS. Width (incomplete) 22-6mm. Mid-point to ear of more complete
side 13°3mm. Length 8-2mm. Thickness 8-2mm. Angle of sulcus 33°. An
estimate of the complete width of the shell is about 28 mm.

L. CARBONIFEROUS BRACHIOPOD 279

DiAGNosIs. Small transversely narrowly rhombic shells as long as thick and
about three and a third times as wide as long. High ventral interarea with only
slightly projecting umbo. Ribbing weakly developed and dorsal median fold
prominent only anteriorly ; dorsal umbo medially sulcate. Dental plates short and
subparallel, diverging anteriorly less than the angle of sulcus.

DESCRIPTION. The tip of the right side of the shell is missing and sediment
obscures two thirds of the interarea. The external surface has been deeply exfoliated
in patches, especially on the ventral valve. Elsewhere the shell is somewhat eroded
and on only one small area in the ventral sulcus is primary shell preserved. Thus
details of external ornamentation cannot be given, nor is it possible to count the
number of ribs with certainty other than on the left half of the dorsal valve where
there are ten or eleven simple costae. Within the more deeply exfoliated secondary
fibrous shell of the ventral valve short radially arranged traces of mantle canals are
visible. Growth lines are clearly seen, especially on the dorsal valve (about 4 per mm
at the side of the fold) and indicate a fusiform shape throughout ontogeny ; they
became more prominent anteriorly. The ventral interarea is vertically grooved by
flexures in the secondary fibres (Pl. 1, figs 10, 21) which formed a delicate denticula-
tion of the hinge line at the inner surface. (When covered by primary shell this
feature would not be visible.) The ventral sulcus is prominent but the expected
dorsal fold is virtually lacking although its position is marked by a pair of bordering
ribs more prominent than the others. There is a dorsal median depression which
becomes shallower and less well defined anteriorly from the umbo. The anterior
commissure is uniplicate. No internal structures can be seen in the specimen
although the exfoliation of the ventral umbo shows the secondary fibres to be bent
along lines interpreted as the bases of the dental plates. If this is correct, it can be
said that the dental plates are positioned on the borders of the ventral sulcus at a
distance of 2 mm from the umbo but anteriorly remain subparallel and thus within
the diverging sulcus. (This is the same as is seen in the Fermanagh silicified material
assigned to Fusella.)

The second specimen, from Dovedale (B 7379, Pl. 1, figs 5-7), is smaller than the
holotype and in that it too has lost its right tip and ventral umbo it is less well
preserved (half width 11-3 mm, length c. 6-3 mm, thickness 6: mm). The dorsal
fold is slightly developed anteriorly, resulting in a rather more prominent uniplication
of the anterior commissure than in the holotype. The broken ventral umbo allows
it to be seen that the dental plates are short for they did not reach anteriorly to the
plane of fracture. It is impossible to measure the angle of divergence of the ventral
sulcus but an estimate of the delthyrial angle, based on the remaining dorsal half
of the interarea, is 30°. As in the holotype, the dorsal umbo is unusually shaped
with a shallow V-shaped groove about 1 mm long terminating anteriorly on the
pair of large ribs bordering the fold. Inspection of the visible surfaces shows that
the shell structure of both specimens is fibrous with no sign of endopunctation.

Discussion. It is unfortunate that F. fusiformis is a rare species in rocks of low
to mid Viséan age. If it were not for the second conspecific specimen from Dovedale
we might have considered the type specimen to be a freak. Indeed it is possible

280 FPUSELLA NE COM

that the presence of a reasonably developed dorsal fold is more characteristic ;
this feature is better developed on the Dovedale specimen and it is unusual for a
uniplicate spiriferacean not to have a better developed fold. Growth lines show that
even at a dorsal valve length of 2-5 mm the commissure was uniplicate. The
amplitude of the uniplication is 3-0 mm on the type specimen and 3:2 mm on the
Dovedale specimen. The growth lines also show that these shells grew with a
fusiform shape virtually throughout their postembryonic ontogeny ; a growth line
about 1mm from the dorsal umbo indicates that the valve width at that stage
was 4mm.

1mm

Fic. 1. Camera-lucida drawings of silicified dorsal valve interiors of F. rhomboidea (Phillips)
from the Viséan of Co. Fermanagh, Ireland. A, B, dorsal and posterodorsal views of a
nearly fully developed valve ; C, dorsal view of part of a fully developed valve ; D, dorsal
view of a juvenile valve at twice the magnification. Note the posteriorly narrow sockets
which, in adult shells, remained functional only anteriorly. Cross shading denotes
broken shell material. a.s.— adductor muscle scar; b.c.— broken stump of the crus ;
c.—crus; c.p.—cardinal process; d.i.— dorsal interarea; d.u.—dorsal umbo; mr. —
median ridge (dividing the adductor scars); s.—socket; s.r.—socket ridge, which
merges dorsally with the crural base. (See also Pl. 1, fig. 18.)

L. CARBONIFEROUS BRACHIOPOD 281

Fic. 2. Camera-lucida drawings of a silicified ventral valve of F. rhomboidea (Phillips)
from the Viséan of Co. Fermanagh, Ireland. A, posterior view; B, lateral view;
C, ventrolateral view. a.d.— apex of the delthyrium, filled by secondary shell material ;
d.p.— dental plate; d.r.— dental ridges bordering the interior surfaces of the edges of
the delthyrium. Anteroventrally these ridges are supported by the dental plates which
buttress across the ventral shell cavity. i.v.i. — internal surface of the ventral interarea ;
s.t.— secondary shell thickening between the dental plates; t.—tooth; v.s. — ventral
sulcus. (See also Pl. 1, fig. 19.)

In the British Isles the species most closely resembling F. fusiformis is F.
rhomboidea (Phillips), the type specimen of which also came from Bolland (PI. 1,
figs 13-17). This species is a constituent of the silicified brachiopod faunas being
studied by Brunton (1966, 1968) from Ireland but as yet has not been redescribed
except briefly by Brunton & Champion (1974). Other than F. rhomboidea there are
several species mentioned by Waterhouse (1970), some of which he believed may
belong to Fusella. It was his mistaken belief that FP. fusiformis was endopunctate
which led him to discard species like rhomboidea, convoluta (Phillips) and trigonalis
(Martin) as being closely related. For the same reason Waterhouse related Fusella
to Syringothyris, in particular some specimens believed to be S. cuspidatus (J.
Sowerby). He figured (1970: fig. 2, A-F) a specimen named ‘Spirifer cuspidatus
(Phillips)’ from Treak Cliff, Derbyshire, in the British Museum (Natural History)
collections (BB 40831). This specimen belongs neither to the species cuspidatus
(first described by Martin in 1809 but ascribed to J. Sowerby (1816) by Muir-Wood

282 BUSSE LEA COM

when she selected the lectotype in 1951) nor, since it is impunctate, to Syringothyris
which includes only endopunctate species. The outline, even for a young specimen
of cuspidatus, is too transverse, and furthermore true cuspidatus does not have a
sulcate fold or denticulate ventral interarea such as that of the Treak Cliff specimen.
Of the Gilbertson Collection specimens (B 297) from Bolland, labelled as S. cuspidatus
and discussed by Waterhouse, only the largest three are true endopunctate Syringo-
thyris cuspidatus. The fourth is poorly preserved and of doubtful affinity but the
fifth is impunctate and like the Treak Cliff specimen BB 40831. These specimens
should probably be assigned to the Strophopleurinae but do not accord with any
presently described British species. Their dental plates are more like those of
Voiseyella than Fusella.

Spirifer distans Sowerby is another species discussed by Waterhouse (1970) in the
belief that it is closely related to F. fusiformis. Although some specimens assigned
to distans may resemble F. fusiformis and despite true S. distans being impunctate,
we do not believe it to be a Fusella species. The Sowerby type specimen of S. distans,
from near Dublin, Ireland (B 61009), differs from F. fusiformis in external shape
and a second Irish specimen (B 7664), more complete than the type refigured by
Davidson (1858 : pl. 8, figs 5-8), is illustrated here for comparison (Pl. 1, figs 1-4).
The dental plates of this species diverge, following the borders of the ventral sulcus,
and there is an apically complete arched delthyrial plate.

Another species belonging in the subfamily but remaining difficult to assign to a
genus, through a lack of internal information, is Spivifer roemerianus de Koninck,
from the Tournaisian of Belgium. The same is true for the species Spzirifer
triangularis J. de C. Sowerby, placed in Fusella by Muir-Wood in 1951. This species
is larger than F. rhomboidea and differs in having a high carinate fold and prominent
ventral median rib in the sulcus. Spirifer convoluta Phillips is another extremely
transverse species but it reaches much larger dimensions (at least 80 mm wide) than
F.. fusiformis. Its interior is unknown so the generic assignment is doubtful, but
if it were to be included within Fusella the diagnosis of that genus would require
emendation to include species at this size. Spirifer bicarinata M’Coy was one of the
species M’Coy mentioned originally as being in Fusella. Other than by M’Coy’s
description and incomplete figure, S. bicarinata is virtually unknown; the type
specimen seems to be lost and it is unwise to continue using the name.

In conclusion, therefore, we assign Fusella to the Strophopleurinae and in addition
to the type species, F. fusiformis (Phillips), we include FP’. rhomboidea (Phillips) and
doubtfully FP. trigonalis (Martin), F. triangularis (J. de C. Sowerby), F. roemerianus
(de Koninck) and F’.. convoluta (Phillips) within this genus. Strophopleura probably
evolved from the Mucrospiriferidae in the Upper Devonian and gave rise to the
Tournaisian Voiseyella and Acuminothyris and to the Viséan Fusella ; from this the
Carboniferous to Lower Permian genus Brachythyrina may have evolved, and also
the northern Permian genera Paeckelmanella and Pterospirifer.

REFERENCES

Brown, T. 1849. Iilustrations of the fossil conchology of Great Britain and Ireland, with
descriptions and localities of all the species. viii +273 pp., 117 pls. London.

L. CARBONIFEROUS BRACHIOPOD 283

Brunton, C. H. C. 1966. Silicified productoids from the Viséan of County Fermanagh.
Bull. By. Mus. nat. Hist. (Geol.), London, 12 : 173-243, pls 1-109.
1968. Silicified brachiopods from the Viséan of County Fermanagh (II). Bull. Br. Mus.
nat. Hist. (Geol.), London, 16: 1-70, pls I-9.
— & CHAmpion, C. 1974. A Lower Carboniferous brachiopod fauna from the Manifold
Valley, Staffordshire. Palaeontology, London, 17 (4) : 811-840, pls 107-111.
CaRTER, J. L. 1967. Mississippian brachiopods from the Chappel Limestone of central Texas.
Bull. Am. Paleont., Ithaca, 53 : 253-488, pls 13-45.
1971. New early Mississippian silicified brachiopods from central Iowa. Smithson.
Contr. Paleobiol., Washington, 3 : 245-255, pls 1-2.
1974. New genera of spiriferid and brachythyridid brachiopods. J. Paleont., Tulsa,
48 (4) : 674-696, pls 1-4.
Davipson, T. 1858. A monograph of British Carboniferous brachiopods, 2 (5) 1. Palaeontogr.
Soc. (Monogr.), London: 1-48, pls 1-8.
Ivanova, E. A. 1960. Spiriferida. Jn Orlov, Y. A. (ed.). Principles of Palaeontology, (7)
Bryozoa and Brachiopoda : 264-280. Moscow (Akad. Nauk SSSR). [In Russian.]
1972. Main features of Spiriferid evolution (Brachiopoda). Paleont. J., Washington,
6 (3) : 309-320. [Translated from the Russian.]
M’Coy, F. 1844. A synopsis of the characters of the Carboniferous Limestone fossils of Ireland.
207 pp., 29 pls. Dublin.
Muir-Woop, H. M. 1951. The Brachiopoda of Martin’s ‘‘Petrificata Derbiensia’’. Ann.
Mag. nat. Hist., London, (12) 4: 97-118, pls 3-6.
PuHILuips, J. 1836. Tilustvations of the geology of Yorkshive. Part II. The Mountain Lime-
stone district. 253 pp., 25 pls. London.
RoBerts, J. 1964. Lower Carboniferous brachiopods from Greenhills, New South Wales.
J. geol. Soc. Aust., Adelaide, 11 (2) : 173-194, pls 1-6.
1971. Devonian and Carboniferous brachiopods from the Bonaparte Gulf Basin, North-
western Australia. Bull. Bur. Miner. Resour. Geol. Geophys. Aust., Melbourne, 122 : 1-399,
pls 1-59.
Tuomas, G. A. 1971. Carboniferous and early Permian brachiopods from Western and
Northern Australia. Bull. Bur. Miner. Resour. Geol. Geophys. Aust., Melbourne, 56: 1-272,
pls 1-31.
WATERHOUSE, J. B. 1970. The Lower Carboniferous brachiopod genus Fusella M’Coy 1844.
Life Sci. Occ. Pap. R. Ont. Mus., Toronto, 15: 1-12, pls 1, 2.
Wittiiams, A. et al. 1965. Brachiopoda. Jn Moore, R. C. (ed.). Tveatise on Invertebrate
Paleontology, H. 927 pp., 746 figs. Kansas.

Co, C. BRUNTON, Ph.D:

A. RIssongé

Department of Palaeontology

British Museum (NATURAL HIsToRY)

CROMWELL Roap

Lonpon SW7 5BD Accepted for publication 9 February 1976

PLATE 1

The specimens are all housed in the BM(NH) and those in Figs 1-19 were coated with
ammonium chloride sublimate before being photographed. The prints are not retouched.

Spirifer distans J. de C. Sowerby

Fics 1-4. Viséan of Millicent, Co. Kildare, Ireland. Viewed anteriorly, posteriorly, dorsally
and from the left side. x1. B 7664.

Fusella fusiformis (Phillips)
Fics 5-7. Viséan of Dovedale, Derbyshire. Viewed anteriorly, posteriorly and dorsally.
The arrow on Fig. 7 indicates the region illustrated in Fig. 20. x2. B 7379. Davidson
Collection.

Fics 8-12. Holotype, figured by Phillips, from Bolland, Yorkshire. Viewed anteriorly,
dorsally, posteriorly (the arrow indicates the region enlarged in Fig. 21), ventrally and from
the left side. x1-:5. B249. Gilbertson Collection.

Fic. 20. Exfoliated standard secondary fibres from the position marked on Fig. 7. The
posterior dorsal margin is to the top and the mid-line to the right. Scanning electron
micrograph, x 210.

Fic. 21. Exfoliated secondary fibres of the ventral interarea of the holotype (see Fig. 10)
showing the flexures resulting in a fine denticulation at the commissure. Scanning electron
micrograph, x 140.

Fusella rhomboidea (Phillips)

Fics 13-17. Lectotype, figured by Phillips, from Bolland, Yorkshire. Viewed ventrally,
anteriorly, posteriorly, dorsally and from the right side. x1; Figs 15,17 x2. B 236.
Gilbertson Collection.

Fics 18-19. Silicified specimens from the Upper Viséan of the Sillees river, Co. Fermanagh,
Ireland (see Brunton 1966 for locality details). Fig. 18, mature dorsal valve interior (see
also Fig. 1A-D, p. 280), x4. BB 61611. Fig. 19, young ventral valve interior looking
posteriorly (see Fig. 2A-C, p. 281), x3. BB 61612.

Bull. By. Mus. nat. Hist. (Geol.) 27, 4 IPIG AMIE, a

‘

mo

iF;
i

i

wb

i

’ i
- : / ad | 4
1 f
PAW 7 oo
; ab i i. *
Ae — ~
Fe RAN 5 ot
+f bag ot Fg
‘ a Ve
a) : :
L a
7 4? gil?
i]
pm
o,
7 -
i = yt _) 7
¢

THE JURASSIC CHAETETID, BLASTOCHAETETES
BATHONICUS J. C. FISCHER, IN ENGLAND

By GRAHAM FRANCIS ELLIOTT

ABSTRACT

Three English occurrences are listed of Blastochaetetes bathonicus Fischer, a species abundant
in the French Bathonian east of the Paris Basin. It is suggested that their location, rarity and
growth-forms may be due to migration, south of the Ardennes—London landmass, into less
favourable conditions in the Cotswold area of southern England.

THE POST-PALAEOZOIC chaetetids were monographed by J. C. Fischer (1970). He
discussed the structures and classification of these somewhat problematic calcified
radial-tubular colonial organisms in detail, and compared them with bryozoa,
tabulates and stromatoporoids, concluding that they were referable to the Hydrozoa
(Coelenterata ; Cnidaria). His species blastochaetetes bathonicus (Fischer 1965,
1970) occurs commonly in the Middle Bathonian of the Aisne and Ardennes in
northern France, and is also recorded from Haute-Marne and the Céte-d’Or; all
these areas are in the Jurassic outcrop east of the Paris Basin. In Aisne and
Ardennes the fossil is abundant in the ‘Calcaires pseudo-oolithiques supérieurs’, in
which it forms characteristic banks or beds at some levels. Commonly it is
accompanied by algal oncolites in a coarse-grained calcareous rock, suggesting that
it flourished during life in agitated waters (Fischer 1969a, 1969Qb).

In England it appears to be a rare fossil, and only three occurrences are known to
me, all Bathonian ; two are in Gloucestershire, and one near Bath (Elliott 1972,
1974). These records are now given in detail.

1. Great Oolite, probably near top of White Limestone. Tiltup’s End, Nailsworth,
Glos. For description of this old quarry, now overgrown, see Witchell (1886)
and Woodward (1894 : 270). The specimen is in the British Museum (Natural
History) Dept. of Palaeontology, registered number R 49314, Witchell Collection.

2. Great Oolite, probably near top of White Limestone. Limekiln Pit, Royal
Agricultural College, Cirencester, Glos. Another old exposure (Woodward
1894: 282). BM(NH) Dept. Palaeont. reg. no. R 49313, Hinde Collection
(? ex Tomes Collection). Plate 2, figs 1-3.

3. Great Oolite, Upper Rags, Bed 3 of Elliott (1974). Bathampton Down, east of
Bath, Avon (formerly Somerset). BM(NH) Dept. Palaeont. reg. no. R 49603,
G. F. Elliott Collection. Plate 2, fig. 4.

In thin section all three specimens show the characteristic internal structures
and approximate dimensions as given in full detail by Fischer (1970: 179); the
amber-coloured tube-walls with fibrous ‘jet d’eau’ or feather-structure are beautifully
preserved. The external growth forms, however, are very different. The Tiltup’s

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 : 285-288 Issued 28 October 1976

286 BLASTOCHAETETES BATHONICUS

End specimen is a compact hemispherical growth exactly like that given as typical
for the species by Fischer (1970 : fig. 20a, b), but is a small example (diameter 30 mm,
as opposed to an average of 40mm, maximum 120 mm, in the French material),
No matrix remains. The Cirencester example is larger (approximate diameter
55 mm), higher-domed, and in section is seen to be grown round a coral nucleus
of 15 mm diameter. It is interesting that this nucleus is almost completely altered
to clear crystalline calcite, traces of identifying coral-structure surviving only at the
margins, in contrast to the excellent preservation of the enveloping Blastochaetetes.
The whole specimen is in a white pseudoolitic rock. Finally the Bathampton
example occurs as thin irregular spreads, possibly crusts, of less than 10 mm thick-
ness ; they are in an ochreous organic-rubble limestone with brachiopods, sponges,
molluscs, echinoderms and bryozoa, whole and broken. The rock is interpreted asa
current-accumulated deposit (Elliott 1974). This ‘spreading’ Blastochaetetes is
regarded as a growth form only, for the Tiltup’s End and Cirencester examples both
contain old external surfaces heavily overgrown by later developments of the same
organism.

The obvious and shortest migration-route in Bathonian times, from the colonies
of the Aisne—Ardennes where the species flourished, was westwards in suitable
facies south of the Ardennes—London landmass, and so to the clear-water calcareous
site of the present Cotswolds in England. The areas to both north and south of
this were in different facies. The species did not become abundant in England, and
presumably did not survive into the changed conditions indicated by the overlying
Forest Marble facies (Upper Bathonian). The recent ‘Tentative reconstruction
of the palaeogeography of part of southern England at the end of White Limestone
times’ (Palmer & Jenkins 1975: fig. 6) indicates the Cotswold area as offshore-
current marine. The numerous small facies-changes seen there in the actual rocks
at about this level, as well as the rare varied growth forms of B. bathonicus, may well
be evidence of less favourable conditions for the species than those which existed
in France.

Note. Hartman & Goreau (1972), after a comparison with living Sclerospongiae,
regarded the Chaetetida as extinct sclerosponges.

REFERENCES

Exuiott, G. F. 1972. Cretacicrusta gen. nov., a possible alga from the English Cretaceous.
Palaeontology, London, 15 : 501-503, pls Ioo, Ior.
1974. Note on the palaeoecology of a Great Oolite fossil-bed at Bath (English Jurassic).
Proc. Geol. Ass., London, 85: 43-48.
FISCHER, J. C. 1965. Découverte d’un niveau-repere a “‘Chaetetidae’’ dans le Bathonien
moyen ardennais. C.r. hebd. Séanc. Acad. Sci., Paris 260 : 6641-6643.
1969a. Géologie, paléontologie et paléoécologie du Bathonien au sud-ouest du Massif
ardennais. Mém. Mus. natn. Hist. nat. Paris (n.s.C) 20. 320 pp., 20 pls.
1969b. Sur l’existence d’un cycle biostratigraphique dans le Bathonien au sud-ouest du
Massif ardennais (Note préliminaire). Bull. Soc. géol. Fr., Paris, (7) 10 : 303-307.
1970. Revision et essai de classification des Chaetetidae (Cnidaria) post-Paléozoiques.
Annls Paléont. (Invert.), Paris, 56 : 149-233, 6 pls.

JURASSIC CHAETETID 287

Hartman, W. D. & GorEAu, T. F. 1972. Ceratoporella (Porifera: Sclerospongiae) and the
Chaetetid “‘Corals’’. Tvans. Conn. Acad. Arts Sci., New Haven, 44: 131-148, 26 figs.
PALMER, T. J. & JENKYNS, H. C. 1975. A carbonate island barrier from the Great Oolite

(Middle Jurassic) of central England. Sedimentology, Oxford &c., 22 : 125-135.
WITCHELL, E. 1886. On the Forest Marble and Upper Beds of the Great Oolite, between
Nailsworth and Wotton-under-Edge. Proc. Cotteswold Nat. Fld Club, Gloucester, 8 : 267-
280, pls 4, 5.
Woopwarp, H. B. 1894. The Jurassic Rocks of Britain, IV. The Lower Oolitic rocks of
England (Yorkshire excepted). Mem. geol. Surv. U.K., London, 628 pp.

G BeErriors:, :Sc., Ph.D., F:G.S., F.L:S.

Department of Palaeontology

British Museum (NATURAL HIsTorRy)

CROMWELL Roap

Lonpon SW7 5BD Accepted for publication 9 February 1976

22

PAE 2

Blastochaetetes bathonicus J. C. Fischer

Fic. 1. Solid specimen, ‘cauliflower-head’ growth embedded in matrix, x2. M. Jurassic,
Great Oolite (Bathonian) ; Limekiln Pit, Royal Agricultural College, Cirencester, Glos.
BM(NH) Dept. Palaeont., reg. no. R 49313.

Fic. 2. The same specimen; thin section, x 15, showing radial growth with old outer surface
overgrown by a differently orientated second growth.

Fic. 3. The same specimen; portion of outer surface, x 15.
Fic. 4. Thin section of a spreading or laminar growth, x15. M. Jurassic, Great Oolite

(Bathonian) ; Bathampton Down, near Bath, Avon. BM(NH) Dept. Palaeont., reg. no.
R 49603.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 PLATE 2

COPROEMES OF PIrILOPHYLLUM CUTICLES
FROM THE MIDDLE JURASSIC OF
NORTH YORKSHIRE

By CHRISTOPHER RICHARD HILL

ABSTRACT

Large coprolites from the Roseberry Topping plant bed (Middle Jurassic) consist mainly
of Ptilophyllum fragments. Their significance is discussed.

DESCRIPTION

THE SPECIMENS discussed here were collected over 50 years ago by Dr H. Hamshaw
Thomas, though so far as I know he never described them. They were found in
siltstones of Lower Deltaic age which form the upper layers of the Roseberry Topping
plant bed (Thomas 1913). Besides being much the largest of the half dozen or so
coprolites now known from the Yorkshire Middle Jurassic they have yielded an
entirely different flora from those previously investigated.

The material is in the Collection of the Palaeontology Dept. of the British
Museum (Natural History), and consists of six blocks labelled V 58510 and an
additional one labelled V 58510a. Those labelled V 58510 look as if they once
formed part of a single larger block which broke up in collecting, and judged on the
six fragments this original block must have been large, having more than 250 pellets
covering an area of about rm”. As the pellets occur on several bedding planes,
occupying about 5 mm thickness of the cross-bedded matrix, I assume that their
burial was rapid. Otherwise I would expect them to have become widely dispersed,
whereas they are in fact somewhat clustered into random groups, very like fresh
droppings of a modern herbivore such as a rabbit. Indeed, much like freshly
dropped rabbit pellets the round or oval outlines of the fossil ones sometimes show a
constriction to a point on the side from which they were extruded, and fine striations,
presumably caused by the anal sphincter, may be seen radiating from this point
onto the impression surfaces. The clarity of these markings in the fossil doubtless
indicates that the fresh pellets were of a firm consistency and also of a uniform,
possibly gelatinous, texture. ®

The diameter of the pellets ranges from 8 to 18mm, though they are mostly
10-13 mm. Where present the substance, which is composed of compressed cuticle
fragments and finely divided organic detritus, is a dull black colour. It is about
0-5 mm thick, thinning to the edges, and the lack of compression border suggests
that the pellets were perhaps somewhat flattened before their final burial (com-
pression theory of Walton 1936). The fact that adjacent pellets sometimes overlap
one another may equally indicate that they were originally rather flattened (Harris

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 : 289-294 Issued 28 October 1976

COPROLITES OF

PETEORM VEL Vit CuUricres 291

1946). However they could well overlap if, as in the pelleted form of sheep dung,
they were sometimes produced adhering together in masses.

V 58510a may have come from a different horizon of the Roseberry Topping
siltstone as the colour of the matrix is rather darker. The pellets have a similar
form to those of V 58510 but they are uniformly smaller, about half the size, and
many more of them overlap their neighbours. As, however, they yield an almost
identical flora to those of V 58510 I imagine that they may well have been produced
by the same kind of animal, but possibly an individual having a smaller anus; by
analogy a young rabbit produces much smaller pellets than a fully grown one.

Of the abundant cuticles yielded by Schulze-Ammonia maceration over 99%
of the determinable ones are fragments from leaves and petioles of Ptilophyllum
pectinoides (Phillips) Phillips. Only a few scraps from other species were seen,
chiefly a specifically indeterminable Solenites or Czekanowskia and a Bennettitalean
cuticle resembling Nulssomopteris vittata (Brongn.) Florin (though possibly also
referable to certain species of Anomozamites or Pterophyllum). There are in addition
several well-preserved micropylar and interseminal scale cuticles identical with those
of Williamsoma Inldae Harris and W. leckenbyi Nathorst.

The size of the cuticle fragments reaches about I mm x 3 mm, though it is usually
I mm? or smaller. The substance of the pellets, however, shows marked fracturing
into rectangular blocks and this, which presumably occurred after deposition, may
well have reduced the size of some of the cuticle fragments.

DISCUSSION

Coprolites from the Yorkshire Deltaic beds are rare and informative fossils.
Other than occasional insects, fish skeletons and reptilian footprints they provide the
only evidence so far known for the activity of vertebrate animals in the delta. The
examples previously described, collected from the Gristhorpe Bed and at Beast
Cliff, have been discussed by Harris (1946, 1951, 1956, 1964) and Harris e¢ al. (1974).
None of them, however, has pellets even as much as half the size of those in the present
specimens and they have yielded entirely different floras, chiefly composed of
comparatively delicate leaf and seed cuticles or pollen: mainly Caytonia, Solenites,
Androstrobus, and thin Bennettitalean cuticles. In contrast to these the leaf of the
Roseberry specimens, Ptilophyllum pectinoides, had a robust cuticle. Frequently
also this leaf was resinous, as is indicated both in hand specimens and in the coprolite
by the internal casts of mesophyll cells sometimes seen adhering in patches to the
upper cuticle (Pl. 3, figs 1, 2, 3; Harris 1949). I conclude that the Roseberry
herbivore was a rather larger animal than those which provided Harris’s specimens,
perhaps the size of a sheep or large rabbit. Indeed the present pellets are remarkably
similar to the pelleted form of sheep’s dung. The animal clearly could utilize the
rather harsh resinous leaves of Piilophyllum, though it did not digest the resin to
any noticeable extent.

The nature of the herbivore is, of course, unknown, though presumably it was a
reptile, such as an ornithopod dinosaur, or possibly a mammal. Ifa mammal then
the pellets seem larger than would be expected from any mammals of Middle

292 COPROLITES OF

Jurassic age so far described in the literature. Gut dimensions, however, though
generally related to size, do not display a direct mathematical proportionality to it,
and the analogy to the size of a sheep is approximate.

The intimate association, within the coprolite, of the flower Willamsonza with the
leaf Ptilophyllum is striking. These organs are also seen associated together as
hand specimens in the normal flora at Roseberry and on this basis, with additional
evidence from morphologically intermediate organs, Harris (1969) has attributed
them to the same plant. In this respect it seems that the coprolite provides a
remarkably faithful representation of the normal Roseberry flora. Assuming that
the animal fed on living plants the evidence thus reinforces Harris’s restoration.

The overall floristic composition of the dung is also like that of the normal flora
at certain layers in the upper part of the siltstone, and this is interesting in view of
the complex factors, such as fragmentation and sorting, which normally operate
during the deposition, as at Roseberry, of terrestrial plants in stream channel
deposits. As it is natural to assume that the herbivore fed on living vegetation the
facts suggest that the fragmentary debris which constitutes the normal fossil flora
might sometimes be a remarkably straightforward representation of plant
communities then growing in the vicinity. They may also indicate that the plants
grew fairly close to Roseberry, for if the animal had strayed far one might expect
the dung to yield at least a few species not normally found there.

There is, however, a simpler explanation. The animal may possibly have been a
scavenger, browsing on the Ptilophyllum debris which was being transported in the
deltaic streams and ultimately deposited in great abundance at Roseberry and other
localities. If this is right the dung would be expected to show fragments of species
we already know from the normal Roseberry flora, but it would not necessarily
indicate anything about plant communities in the living vegetation other than that
their debris could be palatable to an unknown animal.

Unfortunately there is insufficient evidence to make a certain choice between these
explanations, though so far as I know there are no modern vertebrates which would
normally prefer a diet of rotting leaves to fresh ones. For this reason the first
explanation seems perhaps nearer the truth.

REFERENCES

Harris, T.M. 1946. Notes on the Jurassic Flora of Yorkshire, 19-21. Ann. Mag. nat. Hist.,
London, (11) 12 : 357-378.
1949. Notes on the Jurassic Flora of Yorkshire, 43-45. Ann. Mag. nat. Hist., London,
(12) 2: 561-585.
1951. Notes on the Jurassic Flora of Yorkshire, 49-51. Ann. Mag. nat. Hist., London,
(12) 4: 915-937.
1956. The investigation of a fossil plant. Pvoc. R. Instn Gt Br., London, 36 (163) : I-11.
1964. The Yorkshire Jurassic Flora, 2: Caytoniales, Cycadales and Pteridosperms. viii+
I9I pp., 7 pls. Brit. Mus. (Nat. Hist.), London.
1969. The Yorkshire Jurassic Flora, 3: Bennettitales. vi+186 pp., 7 pls. Brit. Mus.
(Nat. Hist.), London. ’
—— MIL.inetTon, W. & MILLER, J. 1974. The Yorkshive Jurassic Flora, 4: Ginkgoales and
Czekanowskiales. vilit+150 pp., 8 pls. Brit. Mus. (Nat. Hist.), London.

PIPTELOPH VLE OM CULICLES 293

Tuomas, H. H. 1913. The Jurassic plant beds of Roseberry Topping. Naturalist, London,
1913 : 198-200.

Watton, J. 1936. On the factors which influence the external form of fossil plants; with
descriptions of the foliage of some species of the Palaeozoic Equisetalean genus Anunularia
Sternberg. Phil. Trans. R. Soc., London, (B) 226 (535) : 219-237.

Ca. Hirt, 1B.Se.-PheD.

Department of Palaeontology

BriTIsH Museum (NATURAL HIsToRY)

CROMWELL Roap .

Lonpon SW7 5BD Accepted for publication 9 February 1976

PLATE 3
Ptilophyllum pectinoides (Phillips) Phillips
Fics 1, 2, 3. Upper cuticle prepared by Schulze maceration of a hand specimen found at
Hasty Bank. Fig. 1, V 58591, light micrograph, x160. Fig. 2, inner surface of cuticle ;
scanning electron micrograph showing resinous casts inside the ‘cutinized’ walls of palisade
mesophyll cells, x 336. Fig. 3, same, x 1040.

Fic. 4. Coprolite, chiefly of Ptilophyllum pectinoides, from Roseberry Topping. V 58510, x TI.
Photo: J. V. Brown.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

A NEW FOSSIL FRESHWATER CRAB FROM
THE NGORORA FORMATION (MIOCENE)
OF KENYA

By SAMUEL FRANCIS MORRIS

ABSTRACT

A new species of freshwater crab, Potamonautes (Lirrangopotamonautes) tugenensis, is de-
scribed from the Miocene of the Tugen Hills, Kenya. It is compared with Recent species of
P. (Orthopotamonautes) and P. (Lirrangopotamonautes).

MATERIAL

DR MARTIN PICKFORD of Bedford College, London, whilst collecting fossil mammals
from the Lake Baringo area of northern Kenya, collected a number of claystone
nodules from the northern end of the Tugen Hills, about 20 miles (32 km) north-west
of Lake Baringo. They came from the two neighbouring localities of Kamwina,
Kabarsero area (Pickford locality 2/72) and Kalimale area (Pickford locality 2/68).
The Kamwina locality is a lake shoreline deposit with diatomaceous shales and the
Kalimale section is a relict river channel. Kamwina is Member C and Kalimale
Member D of the Ngorora Formation of Miocene age, approximately II-12 x 108
years (Bishop & Chapman 1970, Bishop, Chapman, Hill & Miller 1971). The
Kalimale locality has yielded only isolated limb segments. For the latest assessment
of the chronostratigraphy of the region, Berggren & van Couvering (1974) should
be consulted.

The nodules collected by Dr Pickford contain freshwater crabs and ostracods.
The brittle nodules are not laminated in any way so that attempts to break them
open by hammer resulted in the shattering of the nodule. An attempt to develop
the nodules by acid treatment was unsuccessful. It remained to split each one with
a rock splitter, which was less damaging than the hammer but nevertheless reduced
it to a number of fragments that had to be glued back together. If there was any
preferred plane of splitting it was through the abdomen and along the smoother
subhepatic and pterygostomial areas. Even though the specimens are in nodules
they arecrushed. Fortunately Dr Pickford had collected one specimen, the holotype,
which had weathered free of matrix and was not badly crushed, and also some
isolated segments of limbs. Altogether about 50 nodules were collected that
contained crab remains. The limb fragments from Kalimale are not sufficiently
different to suggest that a second species is present, but limb fragments are not very
diagnostic. A modern freshwater crab species would probably not occupy both
such diverse habitats as river channels and lake shores, since this would involve
different water speeds, temperature and bottom conditions. Hence it is possible

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 : 295-300 Issued 28 October 1976

296 KENYA MIOCENE

that a second species is present, but that problem can only be resolved by more and
better material from Kalimale.

SNES AIM IIE 'S)

Section BRACHYRHYNCHA Borradaile 1907
Superfamily XANTHOIDEA Dana 1851
Family POTAMIDAE Ortmann 1896
Genus POTAMONAUTES Macleay 1838

Type SPECIES. Thelphusa perlata H. Milne-Edwards 1837, by subsequent
designation of Rathbun 1905 : 159.

Subgenus LIRRANGOPOTAMONAUTES Bott 1955

TYPE SPECIES. Potamon (Potamonautes) lirrangensis Rathbun 1904, by original
designation of Bott 1955 : 264.

Potamonautes (Lirrangopotamonautes) tugenensis sp. nov.
Pl. 4, figs 1-6

Diacnosis. A Lirrangopotamonautes with strongly produced front and medially
divided urogastric area, single spines on the carpus and on the otherwise smooth
merus.

MATERIAL. One more or less complete carapace, 50 distorted and fragmentary
carapaces and some isolated chelipeds. The holotype and most of the paratypes
are in the National Museum of Kenya, Nairobi. The remaining paratypes (nos
In 61517—24) are in the British Museum (Natural History), London. The material
comes from two Miocene localities in the Ngorora Formation, at Kamwina and
Kalimale about 20 miles (32 km) NW of Lake Baringo, Tugen Hills, Kenya.

DESCRIPTION. Carapace one quarter wider than long at its widest point about
one third back. The front is moderately long and strongly produced with only a
slight ventral deflection ; it is smooth and occupies about one third of the anterior
carapace width. The anterior margin is smooth and strongly sinuous with a
thickened edge. On the holotype one ocular peduncle is preserved. The post-
frontal crest is strong, overhangs the front and is nearly straight with a slight
deflection towards the front at the mid-line ; it meets the anterolateral margin at
an obtuse angle. The postfrontal crest and anterolateral margins are smooth or
slightly beaded. The margins of the orbits are entire, thickened and smooth.
The single extraorbital spine is well developed. The median furrows join at or only
just posterior to the postfrontal crest and pass forward on to the frontal area,
cutting the frontal area into two. The lateral margins are convex anteriorly and
may be slightly concave posteriorly.

The anterolateral marginal ridge passes up on to the dorsal carapace just behind
the widest part of the carapace. The posterolateral marginal ridge starts below the

FRESHWATER CRAB 297

point at which the anterolateral marginal passes on to the dorsal carapace, and goes
in an anteroventral direction.

The hepatic areas are inflated and traversed by two rows of pits. The line of the
pits curves proximally across the branchial area to join the cervical groove opposite
the widest part of the mesogastric area. The urogastric area is clearly delimited by
the cervical groove and the furrow between the uro- and metagastric areas. The
urogastric area is divided by a longitudinal furrow. The proximal walls of the
cervical groove are deeply hachured on the internal mould (and probably also on
the shell itself). The cervical groove is deep medially but becomes shallower
anterolaterally. The rhomboidal branchial areas are inflated and covered with
feeble rugosities.

The ventral surface is poorly preserved — usually only the first four thoracic
sternites or hemisternites are preserved. No abdomina have been found except for
one poorly preserved female, of which the segments appear simple and not fused.
The third maxilliped is reasonably well preserved although no merus has yet been
found. The ischium is broad and subrectangular ; it bears a groove parallel to the
inner margin and about one third in. The exopodite is about a third of the width
of the ischium and tapers anteriorly. The posterior inner angles of the ischia close
against a knob-like anterior extension of the first thoracic sternite, which has the
shape of a ‘parson’s nose’?. The lateral flanges of the anterior extension have two
small swellings, sometimes extended into ridges, one on each flange. Although
corresponding depressions have not been found on the underside of the ischia, they
appear to have a press-stud effect presumably to locate accurately the third
maxilliped when it is closed. The fingers of the larger, right, cheliped are approxi-
mately the same length as the palm and are gently curved towards the mid-line.
The fingers are grooved laterally and bear large teeth. There is virtually no gape
between the fingers, which are opposed throughout their length. The smaller
cheliped is similar to the larger one except that the palm is narrower and the teeth
of the fingers fine with an occasional larger tooth. The smooth carpus bears a
single spine at the distal end of the inner marginal ridge. The ridge is sharp and
strongly elevated. The merus also bears a single spine at the distal end of its inner
marginal ridge but the outer marginal ridge is strong and smooth or may possibly
have a few coarse granules.

DISCUSSION

Very few species of freshwater crabs have been found fossil and this is the first that
can be assigned to a Recent genus. The form of the postfrontal crest and the shape
and position of the epibranchial angle clearly indicate that it belongs to the Recent
African genus Potamonautes. Potamonautes was split by Bott (1955) into fifteen
subgenera, to a large extent on the form of the gonopods. However there are
sufficient carapace differences to allow the fossils to be compared with the two
subgenera Lirrangopotamonautes and Orthopotamonautes. Bott (1955 : 310) erected
four new genera for fossil species of potamid crabs, but stated that on the characters

1 Pygostyle of domestic fowl.

298 KENYA MIOCENE

of the carapace, which alone is usually preserved, the fossil species could be compared
with Recent species. The closest species to the present fossils are the living P.
(Orthopotamonautes) depressus (Krauss) and P. (Lirrangopotamonautes) lirrangensis
(Rathbun). Both these are restricted geographically to central and east Africa at
the present time. O. depressus has a weakly produced or even deflected front
whereas L. lirvrangensis has a strongly produced front similar to L. tugenensts.
However L. livrangensis has a denticulated anterolateral border and three rows of
denticles on the merus of the cheliped. The widest part of the carapace of O.
depressus lies just in front of the anterior third, whilst that of L. twgenensis lies just
behind.

The many species of Potamonautes are distributed through central and southern
Africa, especially in the tropical belts and the Rift Valley system. The majority
have discontinuous distributions at the present day in the cooler water of the
montane regions where they are found in many habitats from rapid-flowing rivers
to shallow trickles or even standing water (Williams 1968). But some species groups,
e.g. the O. johnstont group (Williams 1968) to which L. tugenensis probably belongs,
prefer the slow-moving warmer waters of the lakes and especially the coastal belts.
The Kamwina fossils appear to come from lacustrine deposits in which the water
could have been warm and slow-moving. The discontinuous occurrences of the
Recent species appear to reflect a residuum of previous land forms. The single
fossil species does not allow any reconstruction of previous drainage patterns, but a
number of occurrences might give some clue both to the evolution of the drainage
pattern and to the interrelationship of the many species of freshwater crabs from
east Africa. Dr Keith Bannister is at present trying to unravel the drainage
pattern of east Africa by studying the interrelationships of various species of the
large fish Barbus which are also distributed through many separated bodies of water
(personal communication).

Although L. tugenensis is the first fossil freshwater crab from Africa that can be
assigned with any certainty to a modern genus, Joleaud & Hsu (1935 : 108 ; fig. 11)
published a specimen, described as ‘Potamonidé gen. nov.’, from Cenomanian
terrestrial deposits of Tanout, Niger (Furon 1935 : 35), which is difficult to assess
from the outline figure. Unfortunately the figured specimen cannot be found at
present in Paris. In the form of the postfrontal crest it could certainly belong to
the Potamidae near Potamonautes but the two lateral crests do not conform to any
Recent genus.

REFERENCES

BERGGREN, W. A. & VAN COUVERING, J. A. 1974. The Late Neogene: Biostratigraphy,
geochronology and paleoclimatology of the last 15 million years in marine and continental
sequences. Palaeogeogy. Palaeoclimat. Palaeoecol., Amsterdam, 16: 1-216.

Bisuop, W. W. & CHAPMAN, G. R. 1970. Early Pliocene sediments and fossils from the
northern Kenya Rift Valley. Nature, Lond. 226: 914-918.

, Hitt, A. & Mitter, J. A. 1971. Succession of Cainozoic vertebrate assemblages
from the northern Kenya Rift Valley. Nature, Lond. 228 : 389-394.

BoRRADAILE, L. A. 1907. On the classification of the Decapoda. Ann. Mag. nat. Hist.,
London, (7) 19: 457-486.

FRESHWATER CRAB 299

Bott, R. 1955. Die Siisswasserkrabben von Afrika und ihre Stammesgeschichte. Annis
Mus. r. Congo Belge, Terverun, C. Zoologie (III, III) 1 (3) : 209-352, 30 pls, 103 figs.
Dana, J. D. 1851. On the Classification of the Cancroidea. Am. J. Sci., New Haven, (2)
12 : 121-131.

Furon, R. 1935. Le Cretacé et le Tertiaire du Sahara soudonais. Avchs Mus. natn. Hist.
nat. Paris (6) 13 : 1-96, pls 1-7.

Joreaup, L. & Hsu T.-Y. 1935. Crustacés décapodes du Cretacé de Tanout (Damergou
Niger francais). Avchs Mus. natn. Hist. nat. Paris (6) 13: 99-110, 11 figs.

MactEay, W.S. 1838. Illustrations of the Zoology of South Africa; Annulosa. 75 pp., 4 pls.
London.

MitNE-Epwarps, H. 1837. Histoive Naturelle des Crustacés 2. 532 pp. Paris.

OrTMANN, A. E. 1896. Das System der Decapoden-Krebse. Zool. Jb., Jena, 9: 409-453.

RaTHBUN, M. J. 1904-5. Les Crabes d’Eau douce. Avchs Mus. nain. Hist. nat. Paris (4)
6: 224-312; 7: 159-321; 18 pls.

Witiiams, T. R. 1968. The taxonomy of the East African River-Crabs and their association
with the Simulium neavei complex. Trans. R. Soc. trop. Med. Hyg., London, 62 : 29-34.

S. F. Morris

Department of Palaeontology

British Museum (NATURAL HIsTorRy)

CROMWELL Roap

Lonpon SW7 5BD Accepted for publication 9 February 1976

PLATE 4

Potamonautes (Lirrangopotamonautes) tugenensis sp. nov.

Member C, Ngorora Formation, Miocene ; Kamwina, Kenya

Fic. 1. Holotype, Kenya National Museum. Dorsal carapace. x 1I°5.

Fic. 3. Paratype, Kenya National Museum. Sternum and ischium. x 3.

Fic. 4. Paratype, Kenya National Museum. x 1-2.

Fic. 5. Paratype, Kenya National Museum. Mesogastric and urogastric areas. x 3.
Fic. 6. Paratype, Kenya National Museum. Sternal ‘knob’. x3.

Member D, Ngorora Formation, Miocene; Kalimale, Kenya

Fic. 2. Kenya National Museum. Merus and carpus of left cheliped showing single carpal
spine. XI-75.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 PLATE 4

CYCLOTHYRIS (CRETACEOUS BRACHIOPODA)
FROM CALIFORNIA

By ELLIS F. OWEN

ABSTRACT

Serial sections confirm the occurrence of Cyclothyris densleonis in the Upper Cretaceous of
North America; the known distribution of the genus is extended to California.

Mesozoic brachiopods have not been extensively described within the
North American continent, and it is not surprising that little interest has
been shown by palaeontologists in the few species of Rhynchonellidae and
Terebratulidae already established in the literature. If, on the other hand,
some of these records are further investigated, they can provide useful informa-
tion regarding the ubiquitous nature of some genera and species of little-known
distribution.

Among early records of brachiopods from beds of known Cretaceous age are those
of Whiteaves (1876-1903) and Anderson (1902, 1958) who described rhynchonelloid
brachiopod species which they referred broadly to the genus Rhynchonella. More
recently Imlay (1937) described a species from the Lower Cretaceous of Mexico
which he referred to Cyclothyris ? subtrigonalis. This was the first citation, albeit
tentative, of the genus Cyclothyris from the American continent, but the accompany-
ing transverse serial sections of the species do not support its assignment to this
genus.

It was not until 1955 that Cooper, in a description of some brachiopods from the
Cretaceous of Arizona, first accurately recorded the genus Cyclothyris from North
America, describing and figuring a species as C. americana from the Mural Limestone
(Middle Albian) of the Bisbee Quadrangle. Until this positive record, the genus
was considered to have been confined to the Cretaceous of western Europe
and England where it is represented by species from the Aptian to Upper
Cenomanian.

Anderson (1902) described and figured a rhynchonellid as Rhynchonella densleonts
from the Middle Cretaceous of Shasta County (Horsetown Group), northern
California. Later, in a description of Cretaceous beds of the Pacific coast (Anderson
1958), he considered that beds within the Horsetown Group yielding R. densleonts
were no younger than Lower Cenomanian in age and that their probable range was
from Upper Albian to Lower Cenomanian. A specimen of R. densleonis Anderson
has recently been obtained from the type section and is described here. From the
internal structures seen in transverse serial sections of a specimen (Fig. 4), it can be
positively assigned to Cyclothyris.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 : 301-304 Issued 28 October 1976

302 CRETACEOUS CYCLOTAHYVRIS

x9)

WEP REY

ee

Fic. 4. <A series of fifteen transverse serial sections through the umbo of a duplicate
specimen of Cyclothyris densleonis Anderson from the Horsetown Group (Albian—Ceno-
manian), Shasta County, California. Geological Survey of America Locality No. 1051.
The encircling deltidial plates around the foramen, the comparatively short dental lamellae

and general morphology of the hinge-plates are typical of the genus. The numerals denote
the distance in millimetres between each section. x 3.

SYSTEMATIC DESCRIPTION
Family RHYNCHONELLIDAE Gray 1848
Subfamily CYCLOTHYRIDINAE Makridin 1955
Genus CYCLOTHYRIS M’Coy 1844

Cyclothyris densleonis (Anderson)
Fig. 4; Pl. 5, figs 1-6

1902 Rhynchonella densleonis Anderson: 72; pl. 7, figs 157, 158.
1958 hynchonella densleonis Anderson ; Anderson: 86; pl. 1, figs 3, 4.

DEscRIPTION. External characters. Acutely biconvex Cyclothyris, 17-2 mm long,
20mm wide, 11:°9mm thick. Broadly triangular in general outline. The brachial

FROM CALIFORNIA 303

valve is slightly inflated ; the pedicle valve has a broad, shallow sulcus originating
from about the mid-line and widening anteriorly with a fairly extensive trapezoidal
linguiform extension. Shell ornament consists of numerous radiating fine costae
or costellae which are affected by a series of marked marginal plications creating a
distinctive ornament ; similar ornament is characteristic of Cyclothyris antidichotoma
(Buvignier) from the Upper Aptian and Lower Albian of France and England. The
umbo is slightly produced and the sharp beak suberect. Distinct beak-ridges
border an extensive interarea with exposed deltidial plates encircling a medium-
sized foramen.

Internal characters. The series of transverse serial sections given here (Fig. 4)
for C. densleonis have been compared to those of Cyclothyris latissima (J. de C.
Sowerby), the type-species from Faringdon, Berkshire, as figured by Owen (1962 : 46).
The two series of sections appear almost identical in every detail. The character-
istic deltidial plates encircling the foramen are shown to perfection, as well as the
comparatively short dental lamellae in the pedicle valve. The shape, length and
angle of deflection of the hinge-plates also agree, as do the depth of insertion and
general quadrate shape of the hinge-teeth.

REMARKS. There is an unmistakable similarity between C. densleoms (Anderson)
and C. antidichotoma (Buvignier), but it differs from the latter species in having finer
and more numerous costellae, a more acutely triangular general outline, a narrower
fold and sulcus, a more inflated brachial valve and considerably smaller overall
dimensions. The marginal plicae in C. densleonts appear to be more marked within
the sulcus and on the faint or incipient brachial fold, where there are usually three
or four.

In general outline it is more readily comparable to Cyclothyris mirabilis (Walker)
from the Lower Albian of Leighton Buzzard, Bedfordshire, but differs in having
more, but less acutely developed, marginal plicae, and a slightly more produced
umbo ; it is less inflated than C. mirabilis.

A species which shows a closer morphological affinity with the American form was
described and figured by Panow (1969) as ? Cyclothyris antidichotoma (Buvignier),
and occurs in Lower Cenomanian beds in the Cracow district of Poland. The
specimen figured by Panow (1969: pl. 109, fig. 9) has a wider anterior sulcus, a
slightly less extensive linguiform extension, a less produced umbo and a less acutely
inflated brachial valve.

The importance of the present record is that it confirms that of Cooper (1955) in
establishing Cyclothyris within the Albian and Cenomanian of North America and
extends the distribution of that genus to include California. It may be possible in
time to investigate the records of other workers, such as Whiteaves (1876-1903), so
that a clearer view of the distribution of Cyclothyris and similar brachiopod genera
may be obtained.

REFERENCES

ANDERSON, F. M. 1902. Cretaceous deposits of the Pacific coast. Pyvoc. Calif. Acad. Sci.,
San Francisco, (3) 2 1: 1-129, 12 pls.

23

304 CRETACEOUS .CYCEGHAY RIS

ANDERSON, F. M. 1958. Upper Cretaceous of the Pacific coast. Mem. geol. Soc. Am., New
York, 71 : 1-378, 75 pls.

Cooper, G. A. 1955. New Cretaceous Brachiopoda from Arizona. Smithson. misc. Collins,
Washington, 131 4: 1-18, 4 pls.

ImMLAy, R. 1937. Lower Neocomian fossils from the Miquihuana region, Mexico. J. Paleont.,
Chicago, 11 7 : 552-574, 14 pls.

Owen, E. F. 1962. The brachiopod genus Cyclothyris. Bull. Br. Mus. nat. Hist. (Geol.),
London, 7 2 : 39-63, 2 pls.

Panow, E. 1969. Contribution to the knowledge of the brachiopods from the Upper Cre-
taceous of the Krakow district. Roczn. pol. Tow. geol., Krakow, 39 4 : 555-608, 4 pls.
WHITEAVES, J. E. 1876-1903. Mesozoic Fossils 1: 1-415, 51 pls. Ottawa (Geol. Surv.

Canada).

E. F. Owen, M.Phil., M.I.Biol., F.L.S.

Department of Palaeontology

British Museum (Natura History)

CROMWELL ROAD

Lonpon SW7 5BD Accepted for publication 9 February 1976

eo 7 al =
a
fe
j P

re os eee a

; AS out ‘ y rh. a ae oi 7 i ae | . re oe 7
rd, itera an 7 ioe ©. 7

7 Diets | pe fey" 7

prtaglanhy 02 peel 6 0: | —:.
SIPS «ATS :

PLATE 5

Prefix BM(NH) — British Museum (Natural History), London
MHNG —- Muséum d’Histoire Naturelle, Geneva, Switzerland

Cyclothyris densleonis (Anderson)

Fic. 1a, b,c. Cretaceous, Horsetown Group, Shasta County, northern California.

BB 76200. xX2.
Fic. 4a, b,c. Same locality and horizon. BM(NH) BB 76201. x2.

Cyclothyris antidichotoma (Buvignier)

Fic. 2a,b,c. Lower Albian, Shenley Hill, Leighton Buzzard, Bedfordshire.

BB 41495. XI.

BM(NH)

BM(NH)

Fic. 2d. Enlargement of margin of shell of the same specimen as above, showing the typical

antidichotoma ornament. x2.

Fic. 3a, b. Albian, Morteau, Doubs, France. Internal mould of a young specimen.

CB 4740." x2.

MHNG

PLATE 5

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

CRANIAL MATERIAL OF OLIGOCENE AND
MIOCENE FLAMINGOS : WITH A DESCRIPTION
OF-n NEW SPECIES FROMMAFRICA

By C. J.O. HARRISON & C. A. WALKER

ABSTRACT

Cranial material is available for two mid-Tertiary flamingos, Phoenicopterus croiseti from the
Oligocene and a new species from the Miocene, and this is compared with skulls of the three
Recent genera. The skull of P. cvoiseti from the Upper Oligocene of France is partly recon-
structed, using additional material. It is sufficiently distinct from the type species of Phoeni-
copterus, the Recent P. ruber, to warrant its designation as the type species of a new genus,
Gervaisia. Its bill is slender and less decurved than in Phoenicopterus, the mandibular sym-
physis is shorter and the upper jaw is broad to the tip. New flamingo material from the
Lower Miocene of Rusinga Island, Kenya, is referred to Phoenicopterus on the basis of palate
structure. It is specifically distinct from P. ruber, the bill-shape being different and the size
much smaller, and it is therefore described as a new species, P. aethiopicus.

INTRODUCTION

THE ReEcENT flamingos comprise four species separated into three genera,
Phoenicopterus, Phoeniconatas and Phoenicoparrus, the last including both P. jamest
and P. andinus. Phoenicopterus ruber has three separate populations, P. 1.
antiquorum in Eurasia and Africa, P. 7. yuber in the Caribbean and Central America
and P. 7. chilensis in southern South America. These are sometimes treated as
separate species, but their bones differ only in size. All three forms have been used
here for comparison but the skull of the last, Phoenicopterus ruber chilensis, has been
chosen for the figures.

Skulls representing the three genera were available for examination and are shown
in lateral view (Fig. 5B-D). These are Phoenicopterus ruber, Phoeniconaias minor
and Phoenicoparrus jamesi respectively. The three, in that order, show a progressive
modification of the bill, probably linked with different feeding methods. In the
first two we know of such differences, Phoenicopterus ruber being a more generalized
feeder, on minute crustacea for the most part, and Phoeniconaias minor a more
specialized feeder on blue-green algae. The bill-structure of Phoentcoparrus suggests
a feeding method involving fine filtering similar to that of Phoentconazas.

The most generalized form, Phoentcopterus, has a longer and less angled bill. In
the other two the lower mandible is stouter and shorter and is deflected downwards
more vertically ; the bony structure is more laterally inflated and the mandibular
symphysis longer. At the same time the upper jaw is laterally narrow (Fig. 6), and
more sharply angled (Fig. 5), while the palatal ridge is a narrower and deeper wedge
(Fig. 7). Itis apparent that for the three Recent genera the palate and bill structure
are diagnostic features, and cranial material of fossil species can be compared with

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4: 305-314 Issued 28 October 1976

306 OLIGOCENE AND MIOCENE

Fic. 5. Lateral views of skulls of : A, Gervaisia croiseti (after Gervais) ; B, Phoenicopterus
vuber chilensis ; C, Phoeniconaias minor; D, Phoenicoparrus jamesi. The small arrows
indicate the posterior margin of the mandibular symphysis ; the shaded portions of the
upper mandible indicate the area of the palate visible laterally.

FLAMINGOS oe7

rT
on

erie
ery rereeT ULE
ree

ee

z —-o 0
ae 2
o

o

ae

we 5

Fic. 6. Anterior views of upper mandibles of: A, Gervaisia croiseti (after Gervais) ;
B, Phoenicopterus ruber chilensis ; C, Phoeniconaias minor; D, Phoenicoparrus james.

A B LS

Fic. 7. Transverse section of palate, level with anterior end of nares, of : A, Phoenico-
pterus ruber ; B, Phoeniconaias minor ; C, Phoenicoparrus jamest.

these. Cranial material of fossil flamingos is available for only two forms,
Phoenicopterus crotseti and a new form from Africa.

GERVALISTIA

Phoenicopterus croiseti Gervais 1852
Figs 5A, 6A

P. crotsett was described by Gervais from parts of two skulls and a tarsometatarsus
from Clermont-Ferrand and Gergovie, Puy-de-Déme, France. The deposits at
these localities are now thought by Thenius (1959) to be Upper Oligocene and not
Lower Miocene as originally stated. More recently a specimen in the collection
of the British Museum (Natural History) has been prepared ; it is no. A 2665 (ex

308 OLIGOCENE AND MIOCENE

27827), from the Croiset Collection purchased in 1848, and though said to be collected
at Pérignat, France, it is inscribed ‘Gergovie’. The preparation has exposed most
of the anterior part of a lower mandible and a separate small portion of the posterior
end of the mandible in the region of the coronoid process. The mandibular
symphysis has undergone some lateral crushing. In shape, size and age the specimen
is referable to the present species. With the help of it, using Gervais’s plates but
not his original specimens, it has been possible to reconstruct part of the skull of
P. crotsett (Fig. 5A).

The cranial region is too incomplete to be useful, but the bill shows distinctive
characters (Figs 5A, 6A). It is longer, narrower and less abruptly decurved than the
bills of Recent species. The BM(NH) specimen appears deeper towards the tip of
the lower mandible, but the broad tip of the upper jaw, present in Gervais’s specimen,
may cause the rami of the lower mandible to diverge more sharply dorsally and
appear shallower in lateral view. As shown by the BM(NH) specimen the lower
mandible is not only narrower and straighter than those of Recent species, with a
shorter symphysis as indicated by the small arrows in Fig. 5, but the bony structure
is also less inflated. That of Phoenicopterus shows distinct lateral inflation of the
bone, especially in the deep mid-section of the mandible, and this inflation increases
in Phoenicoparrus and Phoeniconaias. The posterior fragment of lower mandible
(A 2665), which includes the coronoid process, has been crushed and provides little
information other than that the process appears to have been well developed and
that the foramen a little below the process was present.

The palatal ridge of P. crozseti is shallow and the upper jaw is down-curved but
not angled. The end of the upper jaw differs from those of other flamingos in being
broad right to the tip, those of the series of three Recent species listed above showing
increased lateral constriction (Fig. 6).

From the shape of the bill P. crotseti would appear to have had a more generalized
method of feeding than living flamingos. It differs more from the three extant
genera than they differ between themselves. Ifthe others are separated into different
genera then crovseti requires a genus of its own. Apart from the cranial material we
have little information on it; the limb bones referred to this species by Gervais
(1852), Milne-Edwards (1867-71) and Lydekker (1891) are similar in size and
structure to those of Phoenicopterus ruber and are not outstandingly diagnostic.
The following new genus is proposed.

GERVAISIA gen. nov.

ETyMoLocy. The genus is named after Paul Gervais who first described the
species. It is feminine in gender.

TYPE SPECIES. Phoenicopterus crotsett Gervais 1852 : 233-234; pl. 2, figs 4-5.

Dracnosis. Bill long and slightly decurved but not angled. Palatal ridge
shallow and apparently absent from anterior third of upper jaw. Upper jaw broad

and dorsoventrally flattened towards the tip, tapering only gradually. Similar in
size to Phoenicopterus ruber.

FLAMINGOS 309

OccCURRENCE. Transitional Oligocene/Miocene deposits of France and Germany
(Brodkorb 1963 : 272).

A NEW LOWER MIOCENE AFRICAN FLAMINGO

The collection of the British Museum (Natural History) contains some fragmentary
avian material from the Lower Miocene of Rusinga Island in the Kenyan waters of
Lake Victoria. A number of these fragments, mostly ends of long bones, appear
referable to a single species of small flamingo, similar in size to the Recent
Phoeniconaias minor. They consist of two mandibular fragments, one proximal and
three distal ends of humeri, one distal end of a femur, three distal ends of tibiotarsi,
and five proximal and two distal ends of tarsometatarsi. The limb bones, which are
mostly broken or crushed to some degree, show little difference from those of
Phoeniconaias minor save that an almost complete proximal end of a tarsometatarsus
shows the calcaneal ridges of the hypotarsus to be similar in proximal view to those
of Recent species but only half to two-thirds their length.

The jaw fragments, however, differ more from Phoeniconaias minor in their
structure. The larger fragment (Figs 8A, 9A) is the posterior part of the upper jaw
with most of the narial apertures, embedded in matrix at its posterior end. Its
most significant feature is the poorly-developed palatal projection with the double
ridge along its centre. In this respect it resembles Phoenicopterus ruber so closely
that it seems reasonable to include it in that genus (Fig. gA—B).

The other fragment is from the anterior end of the lower jaw. It is slender,
slightly decurved, and heavily grooved in a manner more similar to that of
Phoeniconaias. It tapers more gradually than does the bill of Phoenicopterus ruber
in this region and may therefore have been longer and more slender ; it appears
less deeply hollowed ventrally towards the posterior and in comparison with P. ruber
the bill may have been shallower in the region of the symphysis. The most marked

D

Fic. 8. Dorsal views of basal area of upper mandibles of: A, Phoenicopterus aethiopicus
(holotype) ; B, Phoenicopterus ruber chilensis; C, Phoeniconaias minor ; D, Phoenico-
parrus jamest.

310 OLIGOCENE AND MIOCENE

Fic. 9. Lateral (right) view of posterior part of upper jaw and transverse section of palatal
ridges of: A, Phoenicopterus aethiopicus ; B, P. ruber chilensis.

A B

Fic. 10. Transverse section of lower mandible, near tip, of: A, Phoenicopterus ruber
chilensis ; B, Phoenicopterus aethiopicus (holotype) partly concealed by matrix.

difference however is in the shape of the tomial edge of this region of the mandible.
Recent species of flamingo, and as far as can be determined this applies also to
Gervaisia croisett, have a flattened dorsal edge to the tomium on which the flat edge
of the upper jaw rests, although the latter is much reduced in Phoemcoparrus and
Phoeniconaias. Phoenicopterus ruber has slender, incurving tomia on the lower jaw
(Fig. 10A). The new species has stout tomia slanting inwards on the inner surface,
with a sharp dorsal ridge, and the outer surface bordering this ridge is laterally
compressed to produce a shallow hollow (Fig. 10B). The general appearance
suggests that the anterior end of the upper jaw, at present unknown, may have
overlapped the lower jaw. If this were so it would constitute a unique feature within
the Phoenicopteridae and further taxonomic separation of this species might be
justified. For the present, in view of the palatal similarity and the lack of confirma-
tion of possible differences, we propose to treat this as a new species of the genus
Phoenicopterus.

FLAMINGOS 311

Phoenicopterus aethiopicus sp. nov.
Figs 8A, 9A, 108; Pl..6

EtymoLocy. The species is named after the continent in which it occurs.

Dracnosis. Small, slightly smaller than Phoeniconaias minor. Palatal ridge
shallow and double, very similar to that of Phoenicopterus ruber. Tip of lower jaw
narrow, slender and heavily grooved and with stout tomia compressed along the
external side to produce a shallow groove, the appearance of which suggests that
it might accommodate tomia of upper jaw. Tarsometatarsus with calcaneal ridge
of hypotarsus short.

MATERIAL. Holotype: the posterior part of an upper jaw, BM(NH) Pal. Dept.
No. A 4382. Paratypes: BM(NH) Nos A 4383-4398, comprising the distal part of a
lower jaw, two distal and five proximal ends of tarsometatarsi, three distal ends of
tibiotarsi, one distal end of a femur, and three distal and one proximal end of humeri.
Most material partly damaged or crushed. Collected and presented by the late
Del. S. B. Leakey in 1935-

OCCURRENCE. Lower Miocene; locality Rs1z, Rusinga Island, NE Lake
Victoria, Kenya. :

DEscrRIPTION. The holotype is the posterior portion of an upper jaw broken
short proximal to the anterior end of the nares, and with some shattering of the
internasal bridge. The posterior end terminates at the frontonasal hinge, where it
is embedded in matrix. The bone is slender, widening dorsally at the posterior
end. The culmen shows only slight curvature and is a little flattened where it
terminates posteriorly. The nares are elongated and of fairly even width, rising
slightly at the posterior end. The palate is flat posteriorly, with shallow paired
ridges present along most of the length and increasing in depth anteriorly.

The anterior end of the lower jaw (A 4384) is within the region of the symphysis
and lacks the tip. It is slender, rounded and slightly decurved. Proximally it is
flattened and slightly hollowed ventrally. Both ventral surface and sides have many
deep longitudinal grooves. The internal surfaces are concealed by matrix, but the
interior is deeply hollowed, with incurving tomia. The latter are thicker than those
of P. ruber and project dorsally, the edges forming dorsal ridges. The sides of the
jaw bordering the tomia are laterally compressed to form on each side a shallow
hollow with a prominent lower margin, its appearance suggesting that the tomia
of the upper jaw may have overlapped it.

On the tarsometatarsus the hypotarsus is very short proximodistally, differing
in this respect from those of Recent species and from the similar-sized P. minutus
Howard 1955 of the Upper Pleistocene of California. The distal end of the tarso-
metatarsus is like that of Phoeniconaias, differing from that of P. novaehollandiae
A. H. Miller 1963 of the Australian Miocene in the greater spread of trochleae and
less clearly indicated articulation for digit 1, from P. floridanus Brodkorb 1953 of the
Florida Lower Pliocene in the proximally shorter trochlea for digit 2, and
Phoeniconaias gracilis A. H. Miller 1963 of the Australian Early Pleistocene in the
rounder and broader trochlea for digit 2 and more extensive articular surface on the

gi2 OLIGOCENE AND MIOCENE

trochlea for digit 4. The present species also differs from P. floridanus in having
the distal end of the tibiotarsus narrower posteriorly, and from P. stocki L. Miller
1944 of the Mexican Middle Pliocene in being proportionally narrower across the
anterior intercondylar groove of this bone. The other known fossil flamingos,
Phoeniconotius eyrensis A. H. Miller 1963 of the Australian Miocene and Phoenico-
pterus copet Shufeldt 1891 (see Shufeldt 1892) of the American Middle Pleistocene, are
both considerably larger species. Except in the matter of size, however, the known
differences between these various species are slight, and where critical cranial
material is absent it is impossible to assess the true affinities adequately.

MEASUREMENTS. Holotype, A 4382: dorsal length 27-2mm; ventral length
28:Imm; depth at posterior end 17-0 mm, at posterior end of nares to tomium
11-5 mm, to palate 14-3 mm, depth at anterior end to tomium 8-5 mm, to palate
II‘4mm; width of nares 5-7 mm ; width of posterior nasal strut 5-3 mm.

Anterior portion of lower jaw, A 4384: length 26-I mm ; posterior width 12-4 mm,
depth 9‘8 mm; width between tomia posteriorly 5-:o mm; anterior width 7-8 mm,
depth 745mm; width between tomia anteriorly 4.2mm; depth of lateral
compression bordering tomia 2:2 mm.

Proximal end of tarsometatarsus, A 4383: width at proximal end 14:0mm;
width of hypotarsus 7-9 mm ; anteroposterior depth of proximal end to hypotarsus
158mm; depth of cotyla 86mm; proximodistal length of external calcaneal
ridge 8-8 mm, of internal ridge 8-6 mm.

ACKNOWLEDGEMENTS

We wish to thank Miss M. L. Holloway of the British Museum (Natural History)
for preparing the line drawings.

REFERENCES

BRODKORB, P. 1953. A Pliocene flamingo from Florida. Chicago Acad. Sci. nat. Hist. Misc.
124 : 1-4, 2 figs.
1963. Catalogue of fossil birds, part 1. Bull. Fla St. Mus. biol. Sci., Gainesville, 7

4 + 179-293.

GerRvaIs, P. 1848-52. Zoologie et Paléontologie Francaises,1. 271 pp. Paris.

Howarp, H. 1955. Fossil birds from Manix Lake, California. Prof. Pap. U.S. geol. Surv.,
Washington, 264-J : 199-206, pl. 50.

LYDEKKER, R. 1891. Catalogue of the fossil birds in the British Museum (Nat. Hist.). xxvii+
368 pp., 75 figs. London.

MILLER, A.H. 1963. The fossil flamingos of Australia. Condor, Santa Clara, Calif., 65 : 289-
299, 6 figs.

MILLER, L. 1944. A Pliocene flamingo from Mexico. Wilson Bull., Sioux City, Iowa, 56:
77-82, 2 figs.

MILNE-Epwarps, A. 1867-71. Recherches anatomiques et paléontologiques pour servir a@
Vhistotve des oiseaux fossiles de la France. 474+632 pp., atlas 200 pls (2 vols). Paris.
SHUFELDT, R. W. 1892. A study of the fossil avifauna of the Equus Beds of the Oregon

desert. J. Acad. nat. Sci. Philad. (2) 9: 389-425, pls 15-17.
THENIUS, E. 1959. Handbuch der stratigraphischen Geologie, 3, Tertiar. Teil 2, Wirbeltier-
faunen. xi+328 pp. Stuttgart.

FLAMINGOS 313

Dr C. J. O. Harrison, Ph.D. C. A. WALKER

Subdepartment of Ornithology Department of Palaeontology

British Museum (NATURAL History) BritisH Museum (NATURAL History)
TRING CROMWELL ROAD

HERTS LonpDon SW7 5BD

Accepted for publication 9 February 1976

PLATE 6
Phoenicopterus aethiopicus sp. nov.

Holotype, BM(NH) A 4382. Views of the posterior portion of the upper jaw, x2. A, right
lateral. B, left lateral. C, dorsal. OD, palatal.

Paratype, BM(NH) A 4383. Views of the proximal end of a right tarsometatarsus, x 2.
E, external. F, internal. G, proximal.

PA ES 6

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

A NEW FOSSIL PELICAN FROM OLDUVAI
By C. J. 0. HARRISON & C. A. WALKER

ABSTRACT

A new species of fossil pelican, Pelecanus aethiopicus, based on a tarsometatarsus and a scapula
(both incomplete), is described from the Middle Pleistocene of Olduvai, Tanzania. It is com-
pared with Recent and fossil forms.

IN THE COLLECTION of the Department of Palaeontology of the British Museum
(Natural History) are several fragments of bird bones collected from the Middle
Pleistocene deposits at Olduvai, Tanzania. All the specimens are of birds associated
with an aquatic environment, and are referable to Recent genera, but some may be
sufficiently different from present-day forms to make it necessary to consider them as
separate species. One such is the proximal part of a tarsometatarsus of a pelican
which might have been a forerunner of one or more Recent species within the genus
concerned.

Order PELECANIFORMES Sharpe
Suborder PELECANI Sharpe
Family PELECANIDAE Vigors
Genus PELECANUS Linnaeus
Pelecanus aethiopicus sp. nov.
Plate 7

DiaGnosis. Proximal end of tarsometatarsus having internal edge of anterior
surface with posterior curvature and with smooth, not abrupt slope to lip of internal
cotyla. On internal anterior edge, the ridge marking proximal edge of ligamental
furrow curves posteriorly only a little distal to level of hypotarsus. Distal edge of
posterior aperture of inner proximal foramen only just proximal to level of distal
edge of hypotarsus. Hypotarsus posteriorly prominent on internal side, with sharp
slope towards internal side. Internal calcaneal ridge stout, with head only slightly
prominent distally, and with slight ridge towards internal edge. Triangular bridge,
tapering distally, over second tendinal groove formed by heads of next two calcaneal
ridges. Internally the edge of this triangular head begins only a little anterior to
edge of head of internal calcaneal ridge, and slopes externally, with some anterior
curvature. External tendinal groove accommodated laterally, on the external side
of hypotarsus, rather than on posterior surface (Fig. II).

MATERIAL. Holotype, proximal end of a right tarsometatarsus, British
Museum (Natural History) Palaeontology Dept. No. A42g91. Paratype, proximal
end of a right scapula, No. A 4292 in the same collection.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4: 315-320 Issued 28 October 1976

FOSSIL PELICAN

316

“SNDJOAI
-ouo'd ‘q ‘16ehy ‘on (HN)Wa “Aou ‘ds snaidoiyjan ‘gq ‘9 { suaosa{na ‘gq ‘q ‘ sndsisa
‘dT ‘WV ‘isrepezyourosie} FYSII Jo pus yewrxoid jo SMOIA [eUIOJUI pue IONIOJUW “II ‘DIY

oR)

FROM OLDUVAI 317

Locality AND Horizon. Bed I, site FLKS, Olduvai Series (Middle Pleistocene),
Olduvai Gorge, Tanzania (Leakey 1965: 102). Collected and presented by the late
Drs S$: B: Leakey, 1935.

DescriPTION. The holotype is the proximal end of a right tarsometatarsus
(Fig. 11C) in good condition, but with the tendinal canals and anterior fossa filled
with matrix and some crushing of the shaft distally. The external surface is broad
and flat, terminating posteriorly in a distinct ridge, and anteriorly forming a more
rounded ridge bordering the anterior fossa. The anterior surface has a distinct
median metatarsal groove deepening towards the occluded fossa. At the distal edge
of the fossa there is a small tibialis anticus tubercle on the inner side and a smaller
ancillary pit next to it on the external side.

The internal edge is more smoothly rounded than the external edge and shows
some posterior curvature distally. At the proximal end it widens to accommodate
a broad, short hollow just below the lip of the cotyla, presumably for the attachment
of the proximal ligament ; the inner edge of this projects inwards across the proximal
edge of the fossa. The more distal part of the internal edge is smoothed away where
a ligamental band would have crossed it, this surface terminating proximally in a
small ridge a little below the level of the tibialis anticus scar and curving posteriorly
across the internal side (Fig. 11C).

The internal surface is slightly curved and narrow proximally, to widen again
just below the lip of the internal cotyla, which is broken in the present specimen.
The proximal surface is roughly rectangular. The intercotylar prominence is
elevated and bluntly rounded. Viewed anteriorly, it shows some external
deflection ; viewed laterally it shows slight anterior prominence, continuing the line
of the inner edge of the proximal ligamental attachment hollow, and dorsally slopes
gradually towards the hypotarsus. The internal cotyla is narrower laterally and
concave with a slight anterior tilt. The external cotyla is narrow and slopes
anteriorly, its anterior edge curving down towards the anterior fossa. There is a
narrow slanting intercotyla depression at the posterior inner edge of the internal
cotyla, and a narrow, shallow groove crosses the posterior edge of this cotyla and
extends to the opening of the second tendinal canal.

The hypotarsus is large and stout. The main structure is the internal calcaneal
ridge which is thick and prominent, although extending less far proximodistally
than those of some Pelecanus species. Its posterior surface is flattened, projecting a
little at the edges and sloping externally with a slight longitudinal ridge towards the
internal side. Distally, below this overhanging surface the calcaneal ridge extends
distally, decreasing in height. The internal surface of the ridge is irregular with small
ridges and foramina. The posterior aperture of the inner proximal foramen, at the
base of the inner calcaneal ridge just below the projecting surface, is large and
irregular in shape. The anterior surface of the inner calcaneal ridge is concave,
with a narrow projecting external edge.

On the external side of the large inner calcaneal ridge there are two tendinal
canals, one above the other, only one showing between the edge of the posterior
surface of the ridge and the edge of the surface roofing the next tendinal canal. The
second lateral tendinal canal, between the middle and outer calcaneal ridges, is

24

318 FOSSIL PELICAN

roofed over. The middle calcaneal ridge is much shorter and slimmer than the
inner, and only half as long. The roof of these forms a triangular-shaped surface,
tapering distointernally and with a hollow above the tendinal canal, and has an
external slant. The posterior opening of the outer proximal foramen is just distal
and external to the posterior opening of the second tendinal groove. On the external
side of the external calcaneal ridge there is a shallow lateral groove, a second groove
diverging anteriorly from the distal end of this and passing across the proximal
posterior corner of the external surface. Both these latter grooves are small and
poorly defined, with a slight ridge between them which is eroded in the specimen.

The paratype is the proximal end of a right scapula with furcular articulation
broken off short, at a level with the eroded coracoid articulation. The shaft shows
some crushing. The coracoid articulation is a prominent rounded facet projecting
beyond the line of the anterior edge. The glenoid facet, which slants proximo-
internally, is oval and elongate, with a small prominent lip along its outer edge, and a
concavity in the bone on the dorsal surface above it.

MEASUREMENTS (in millimetres). (a). Holotype: proximal end of a right
tarsometarsus, BM(NH) No. A 4201.

Maximum length as preserved 53°9
Maximum width across cotyla 24°5
Minimum measurable width 16-2
Depth of hypotarsus 29°5
Depth at internal cotyla I5°5

External depth of distal end to central posterior ridge 19°3
Internal depth of distal end to central posterior ridge 15°6

Depth of external side at level of tibialis anticus scar 13°4
Depth of internal side at level of tibialis anticus scar 9:9
Length of cap of internal calcaneal ridge 20°7
Maximum width of cap of internal calcaneal ridge ray
Length of second calcaneal ridge 16-0
(b) Paratype: proximal end of a right scapula, BM(NH) No. A 4292.
Maximum length of scapula as preserved 45'1
Maximum width 26-2
Width to internal edge of coracoid articulation 24:0
Width of coracoid articulation 1O2
Maximum thickness of shaft 5°9
Width of shaft I1I°3
Width of glenoid facet 8-6
Length of glenoid facet 14°0

COMPARISON AND COMMENTS. The following Recent specimens were available for
examination. Pelecanus crispus 5, P. rufescens 1, P. conspicillatus 2, P. onocrotalus 8,
P. erythrorhynchus 1, P. occidentalis 5. Of these, crisbus (Fig. 11A), rufescens
(Fig. 11B) and conspicillatus differ from aethiopicus in having on the external side of
the large internal calcaneal ridge a series of three tendinal grooves of which the outer
two are open, and in having a more posteriorly prominent calcaneal ridge. P.
erythrorhynchus differs from aethiopicus in having the three tendinal grooves and a
proximodistally long calcaneal ridge like that of onocrotalus but less abrupt distally.
Its internal calcaneal ridge is, however, stout and only projects a short distance.

FROM OLDUVAI 319

P. onocrotalus differs from aethiopicus in its stouter, less projecting, and proximo-
distally longer internal calcaneal ridge, and in its broader and less projecting
hypotarsus. PP. occidentalis is peculiar in that two discrete populations have different
hypotarsi. The North and Central American population differs from aethiopicus in
having the hypotarsus narrow and prominent, with two open tendinal canals
posterior to a closed one, the more anterior of the two being partly enclosed. The
southern population, P. 0. thagus of Chile and Peru, is a larger form in which the
hypotarsus is much stouter and projects less. The external edge of the internal
calcaneal ridge is extended externally, projecting beyond the tendinal canals. The
two open posterior canals become displaced anteriorly and the more anterior of the
two is completely enclosed. The general effect is as though the hypotarsus had been
compressed onto the tarsometatarsal shaft with accompanying distortion. In view
of this divergence in what is otherwise a specifically constant character there may be
grounds for re-examining all characters of the two forms of occidentalis to determine
whether their earlier separation as two species may be preferable. Both forms
differ from aethiopicus in the characters mentioned.

Hypotarsal material is available for only a few fossil pelicans. P. gracilis Milne-
Edwards 1867, P. intermedius Fraas 1870 and P. odessanus Widhalm 1886 (listed by
Lambrecht 1933 : 279) all have the narrower and more projecting hypotarsus typical
of the cvispus group; on the last two the two open posterior tendinal grooves are
apparent. The first two of these species are from the Miocene, the last from the
Lower Pliocene. Of the three Pleistocene species already known P. halieus Wetmore
1933 from North America, described from a radius, is very small and osteologically
similar to erythrorhynchus, the Recent North American species. P. grandiceps
De Vis 1906 and P. proavus De Vis 1892 (p. 444), found in Australia and described as
similar to conspicillatus, the only Recent species occurring in that region, were
described from various bones among which the proximal ends of tarsometatarsi
were not represented. Present information suggests aethiopicus differs from all
of these.

ACKNOWLEDGEMENTS. We would like to thank Miss M. L. Holloway for making
the line drawings, and Mr C. Keates for taking the photographs.

REFERENCES

De Vis, C. W. 1892. Residue of the extinct birds of Queensland as yet detected. Proc. Linn.
Soc. N.S.W., Sydney, (2) 6: 437-456; pls 23-24.

1906. A contribution to the knowledge of the extinct avifauna of Australia. Ann. Qd
Mus., Brisbane, 6: 1-25, pls I-9.

Fraas, O. 1870. Die Fauna von Steinheim. Jh. Ver. vaterl. Naturk. Wiirtt., Stuttgart,
26 : 281-283; pl. 13, figs 3a-c, 4.

LAMBRECHT, K. 1933. Handbuch der Palaeornithologie. xix+1022 pp., 4 pls, 209 figs. Berlin.

LEAKEY, L. S. B. 1965. Olduvai Gorge 1951-1961, 1. xiv+118 pp., 97 pls, 2 figs, I map.
Cambridge.

MILNE-Epwarps, A. 1867. Recherches anatomiques et paléontologiques pour servir a V histoire
des oiseaux fossiles de la France, 1: 250-254; atlas 1: pls 38-39. Paris.

WEtTmoRE, A. 1933. Pliocene bird remains from Idaho. Smithson. misc. Collns, Washington,
87, 20: I-12, 8 figs.

320 FOSSIL PELICAN

WIDHALM, J. 1886. Die Fossilen Vogel-knochen der Odessaer Steppen-Kalk Stein-briiche
der Neuen Slobodka bei Odessa. Zap. novoross. Obshch. Estest., Odessa, 10, Beilage :

4-6:; .ph5.
Dr C. J. O. Harrison, Ph.D. C. A. WALKER
Subdepartment of Ornithology Department of Palaeontology
British Museum (NATURAL History) British Museum (NATURAL HIsTory)
TRING CROMWELL ROAD
HERTS Lonpon SW7 5BD

Accepted for publication 9 February 1976

PLATE 7

Pelecanus aethiopicus sp. nov.

Holotype. Views of the proximal part of a right tarsometatarsus, BM(NH) No. A q2or1.
A, internal; B, anterior; C, external; D, posterior; E, proximal. xr.

PLATE 7

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

CHELIDONOCEPHALUS TRILOBITE FAUNA
FROM THE CAMBRIAN OF IRAN

By R. A. FORTEY & A.W. A. RUSHTON

ABS LRACT

A trilobite assemblage from the Mila Formation in northern Iran contains Chelidonocephalus
preannulatus sp. nov., Hadvagnostus edax sp. nov., Koldiniella mitella Sivov, Peronopsis fallax
aff. minor (Brégger), Dorypyge sp., Parakoldinia ? sp. and Tsinania? sp. It is of earliest
Upper Cambrian age. The type species of Chelidonocephalus King (C. alifrons) and Ivanoleesia
King (I. pistformis) are redescribed ; Ivanoleesia falconi (King) is transferred to Anomocarella.
Some problems in the classification of ptychoparioid trilobites are discussed.

INTRODUCTION

IN NORTHERN IRAN Cambrian rocks crop out in the east-west trending Alborz
Mountain belt south of the Caspian Sea. In western Iran further Cambrian outcrops
are known along a belt parallel to the Zagros Thrust north-west of Shiraz (Fig. 12).
Fossils from the latter area have been described by King (1937), and more recently
Kushan (1973) described the faunas and zones of the Mila Formation (Stécklin
et al. 1964 : 20) in the Alborz Mountains.

While surveying the Qazvin quadrangle north-west of Tehran (Annells e¢ al. 1975),
Dr R. G. Davies collected fossiliferous samples from within c. 10 m of the local base
of the Mila Formation near Sanghabad (field locality RD 569), at lat. 36° 06’ 43” N,
long. 50° 37’ 37” E. These samples yielded an assemblage of trilobites some-
what ambiguous in age when compared with the faunas described by Kushan
(1973). The assemblage includes well-preserved agnostids previously undescribed
from Iran, and as a new species of Chelidonocephalus is present we have taken the
opportunity to reassess Chelidonocephalus and associated genera and species known
only from Iran, the affinities of which were obscure.

The trilobites are preserved, with fragments of horny brachiopods, in a grey
sparry limestone cut by many minor joints. Several, but not all, show slight
tectonic deformation. Although generally well preserved, the trilobites are all
disarticulated and many were broken before fossilization.

Reference is made to specimens in the Palaeontology Department, British
Museum (Natural History) (It and In), the Geological Survey of India (GSI) and
the Institute of Geological Sciences, London.

We are grateful to Dr R. G. Davies for discussing the field occurrence of the
fossils with us and to Mr S. F. Morris for technical assistance. A. W. A. Rushton
publishes by permission of the Director, Institute of Geological Sciences.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 : 321-340 Issued 28 October 1976

322 CHELIDONOCEPHALUS FAUNA

50 55

Caspian

Fic. 12. Sketch-map of part of Iran. Structural lines from Stécklin, 1968. Localities
1-5 are from Kushan 1973: 1 = Mila Kuh, 2 = Shahmirzad, 3 = Haanakdar, 4 = Ab-
har, 5 = Qanli-Chapoghlu. Locality 6 = Sanghabad (R. G. Davies field Locality 569).
Localities 7-10 are from King, 1937: 7 = Chal-i-Sheh, 8 = Darreh Shu, 9 = Tang-i-
Tehbud Bezuft, to = Ma’dan. Chelidonocephalus alifrons is recorded from localities
I, 3, 4, 5, 7, 8, 9, Lvanoleesia pisiformis from 1, 3, 4, 5, 10 and Koldimella mitella from

3, 4, 5; 6.

AGE AND CORRELATION OF THE FAUNA
Kushan (1973 : 125) recognized the following zones in the Mila Formation.

Tremadoc Saukia Zone
Alborsella Zone
Kaolishania Zone
Prochuangia Zone
Drepanura Zone
Dorypyge Zone
Ivanoleesia Zone

Upper Cambrian

Middle Cambrian

Ivanoleesia is confined to its zone and is accompanied by Anomocarella etc.
According to Kushan the several Dorypyge species are confined to the Dorypyge
Zone and some are associated with agnostids and Chelidonocephalus alifrons or C. sp.
The Drepanura Zone is characterized by Drepanura (Spinopanura) or by Koldiniella,

CAMBRIAN OF IRAN 323

or both, and also contains C. alifrons, C. sp. and agnostids. Kushan did not allow
for an overlap of the ranges of Dorypyge and Koldiniella, such as occurs in the
present assemblage.

We refer the present fauna to the Drepanura Zone because there are several
specimens of Koldimella mitella in the collection, and the presence of this species,
of agnostids and of a species of Chelidonocephalus are typical of this zone. There is,
on the other hand, only one specimen of Dorypyge and that is a different species from
any recorded by Kushan from the Dorypyge Zone. On a priori grounds Kushan
made a good case for the Middle Cambrian age of the Dorypyge Zone and the
Upper Cambrian age of the Drepanura Zone, but as the new record implies an Upper
Cambrian age for a Dorypyge species (a genus hitherto restricted to the Middle
Cambrian) we here examine the age of the fauna in relation to the Middle-
Upper Cambrian boundary. We accept the boundary as being between the Lejopyge
laevigata Zone and the Agnostus pisiformis Zone, as reviewed by Daily & Jago
(1975). Chelidonocephalus is restricted geographically and is of no help in wider
correlation, nor is the Tsinania ? sp. because it cannot be convincingly compared
with other described forms. Peronopsis fallax aff. minor resembles a subspecies
known from the late Middle Cambrian of Scandinavia and P. fallax cf. minor from
the A. pisiformis Zone in the Nuneaton district, England (Taylor & Rushton
1972: 19). Hadragnostus edax sp. nov. is comparable with H. Jas Opik from the
eretes and quasivespa Zones of the Mindyallan Stage in Queensland: Opik regarded
these horizons as Upper Cambrian but Daily & Jago (1975 : 538) suggested that the
eretes Zone and part of the guasivespa Zone should be correlated with the top of the
laevigata Zone, thus making H. las a topmost Middle Cambrian species. A third
species, Hadragnostus modestus (Lochman), occurs in the Crepicephalus Zone of
Montana (Lochman 1944) in beds equivalent to part of the A. piszformis Zone.
Thus these agnostids do not give a precise age as they are related to forms occurring
both above and below the Middle-Upper Cambrian boundary.

The only species of the present fauna found beyond the Mila Formation is
Koldimella mitella Sivov, which occurs in the lowest Upper Cambrian at Salair,
western Siberia. Rozova (1968: table 4) correlated this occurrence with the
Kulyumbe ‘superhorizon’ of north-eastern Siberia which is referable to the
Pedinocephalina-Toxotis (2) Zone, and this in turn was correlated with a level
above the base of the A. pisiformis Zone by Daily & Jago (1975: 546). K. muitella
seems therefore to indicate an Upper Cambrian age, and with it goes the rest of the
fauna, including the Dorypyge sp. We furthermore conclude that Kushan’s choice
of the Middle-Upper Cambrian boundary in the Mila Formation was apposite.

SYSTEMATIC DESCRIPTIONS

CLASSIFICATION PROBLEMS. One of the reasons palaeontologists find the mass of
‘ptychoparioid’ trilobites difficult to classify is that some persist in forcing genera
into higher taxonomic groups on the basis of a single character in defiance of other
characters which contradict such an assignment. Blindness and effacement of
external features are thought to have been polyphyletically derived. Yet the blind

324 CHELIDONOCEPHALUS FAUNA

Conocoryphidae even now have the company of such distinct genera as Hartshillia
and Meneviella thrust upon them despite Lake’s observations that this ignores the
characters of their thoraces and pygidia (1940: 277). Similarly, Kobayashi, who
recognized the polyphyletic character of effaced trilobites (1935 : 303), grouped
nine varied genera in the Tsinaniidae (1960 : 397) but he himself regarded five of
these as doubtful, with which we agree. We would add a sixth, Koldimia. This
leaves the family with three closely related genera — Tsinania, Dictyites and perhaps
the poorly known Dictyella : if Kobayashi is correct in regarding all these as effaced
asaphiscids is there any need for a family Tsinaniidae at all?

We have considered the genera likely to be related to Chelidonocephalus,
Ivanoleesia and Anomocarella below without regard to the (apparently haphazard)
arrangement of the families in which they have been included. This has not led to
a satisfactory understanding of the families concerned, and we have therefore been
forced to use the category ‘family uncertain’.

Family AGNOSTIDAE McCoy 1849
Genus HADRAGNOSTUS Opik 1967

TYPE SPECIES. Original designation, H. las Opik 1967.

Discussion. Hadrvagnostus has a distinctive cephalon with a long glabella and
short preglabellar field. The front of the cephalon is transverse and straight, or
even slightly emarginate in H. las (Opik 1967: pl. 58, fig. 7) and H. modestus
(Lochman 1944: pl. 5, fig. 10) ; the new species H. edax, described below, has a
straight or convex anterior margin.

The pygidium of Hadrognastus differs from that of Peronopsis in that it has a
terminal node. Baltagnostus has a terminal node but the pygidial border between
the posterolateral spines is widened medially (i.e. slightly crescent-shaped) or has a
pygidial ‘collar’. Opik (1967 : 103) described the articulating device of H. las as
simple, but the holotype of H. edax sp. nov. has an agnostoid articulation (Opik
1967 : 72) which supports reference of Hadragnostus to the Agnostinae.

Hadragnostus edax sp. nov.
Pl. 11, figs 3-15
DERIVATION OF NAME. Latin, edax, greedy, referring to the swollen pygidial axis.
HoLotypPe. A well-preserved pygidium, It 13463 (PI. 11, figs 7, 8, 12).

TYPE LOCALITY AND HORIZON. Mila Formation, Sanghabad, Taleghan Range,
Alborz Mountains, northern Iran.

FIGURED MATERIAL. Cephala It 13468 (Pl. 11, fig. 3), It 13483a (Pl. 11, figs 9,
10), It 13483b (Pl. 11, figs 14, 15), It 14023 (Pir, fig. 11). ) Pygidia i 1463
(holotype, Pl. 11, figs. 7, 8; 12), dtergagn «(Pl Sem, she. 3), Pe o47 2 ee
figs 4-6).

CAMBRIAN OF IRAN 325

DESCRIPTION. Cephalic outline tends to be subquadrate, length about four fifths
of width. Glabella three quarters of cephalic length, bilobed; anterior lobe
relatively small, nearly twice as wide as long, narrower than posterior lobe and one
third of its length or less. Posterior lobe parallel-sided, rounded or slightly pointed
behind, without lateral furrows or a median node, but some specimens have a faint
posterior node (Pl. 11, fig. 15). Basal lobes triangular, inflated, connected behind
glabella. Each cheek (excluding border) as wide as posterior glabellar lobe, tapering
strongly forwards ; in front of the glabella the cheeks are narrower than the length
of the anterior glabellar lobe and are faintly separated by a partial preglabellar
median furrow which forms a shallow depression just in front of the glabella and
shallows or fades out forwards. Laterally and frontally the cheeks curve down
steeply to the border which is mainly flat and horizontal but has a narrow outer
rim and is curved under at the edge.

Thorax unknown.

Pygidium strongly convex, transverse, length about three quarters of width.
Axis large, extending back to overhang the posterior border furrow. At anterior
end axis is just less than half the total pygidial width, and widens slightly backwards.
The articulating furrow has a depression either side of the median line, giving a
weakly diplagnostid character. A median tubercle at the highest point has its
posterior end just in front of the mid-length of the axis. Irregularities in the axial
furrow indicate the merest traces of lateral furrows opposite the front and back of
the median tubercle. A small terminal node lies well above the border furrow.
Each flank about two fifths as wide as axis in the larger specimens but proportionally
wider — nearly half the axial width — in a small specimen about 1:1 mmlong. Flanks
slope steeply down to the border, narrow backwards, separated behind axis. Border
broad and thick, of even width between the small posterolateral marginal spines.
Surface smooth.

Discussion. The present form is very similar to Hadvagnostus las Opik
(1967 : pl. 58, figs 6-10) but has a slightly shorter and narrower anterior glabellar
lobe. In H. las this lobe is more than one third of the length of the posterior lobe
but in the present form it is less than a third. WH. las has a stronger and wider
preglabellar median furrow. The pygidial flanks are narrower, less than half the
width of the axis, whereas in H. Jas they are more than half, and in Opik’s reconstruc-
tion (1967 : 103, text-fig. 24) the posterolateral marginal spines are shown as larger
than those of H. edax.

One figured cephalon of Hadragnostus modestus (Lochman 1944: pl. 5, fig. 10) is
typical of the genus: the anterior glabellar lobe is narrower than the posterior lobe
but it is longer than in H. edax. The preglabellar median furrow is narrower than
that of H. Jas and stronger than in H. edax. The basal lobes of H. modestus are
smaller than in either of the other species. Lochman’s other figured cephalon has
a more rounded outline, the posterior glabellar lobe is more tapered and the
preglabellar field is longer; it may represent another genus. The holotype and
paratype pygidia of H. modestus are meraspids with the median tubercle
at the mid-length of the pygidium. Full comparison with H. edax cannot be
made.

326 CHELIDONOCEPHALUS FAUNA

Family QUADRAGNOSTIDAE Howell 1935
Genus PERONOPSIS Hawle & Corda 1847

TYPE SPECIES. By monotypy, Battus integer Beyrich 1845.

Peronopsis fallax (Linnarsson 1869) aff. minor (Brogger 1878)
Pl. 12, figs 1-14

FIGURED MATERIAL. Cephala It 13480a—b (Pl. 12, figs 1, 2), It 13480c (Pl. 12,
figs 3, 4), It 13480d (Ph 12, fig..14), It 14024 4Pl. 12, fies-o, 11); lta 4026 (Eig:
fig. 10). Pygidia It 13467 (Pl. 12, fig. 6), It 13469 (PI. 12, figs 12, 13), It 13485
(Pl. 12, figs 5, 8), It 14025 (PI. 12, fig. 7). Some twenty unfigured cephala and
pygidia also present.

DESCRIPTION. Cephalon rounded in front, nearly as long as wide. Glabella
about 0-7 of cephalic length, bilobed; anterior lobe rounded in front, slightly
narrower than, and less than two fifths the length of, the posterior lobe. Trans-
glabellar furrow narrow at external surface, broader in exfoliated specimens.
Posterior glabellar lobe has faint lateral furrows at two fifths of its length from the
front and a faint median node, generally elongate, at or in front of its mid-length ;
bluntly rounded behind. Basal lobes triangular but with a forward extension in
some specimens which indents the sides of the glabella, somewhat as in Ptychagnostus
atavus (Tullberg) ; this feature is not seen in exfoliated specimens. Cheeks about
as wide as glabella, narrowing forwards slightly, confluent in front. A pit
resembling an incipient preglabellar median furrow is seen in front of the glabella in
some specimens. Cheeks slope steeply down to border in small or exfoliated
specimens, less steeply in the largest specimens. Border furrow is a curve where
the cheeks flatten out to form a border, and is broad in large specimens. Border is
a narrow convex rim.

Thorax unknown.

Pygidium convex, subquadrate, length about 0-9 of width. Axis about two fifths
of total width, parallel-sided with a tendency to narrow slightly at the anterior
third. Posterior end generally bluntly pointed but may be rounded, especially in
large specimens, not reaching border. Median tubercle at anterior two fifths ;
traces of lateral furrows opposite anterior and posterior ends of tubercle. Articu-
lating device simple, basic (Opik 1967 : 72). Flanks (pleural lobes excluding border)
about half the width of axis but proportionally narrower in specimens more than
2mm long, and always confluent behind axis although the connection is narrow in
some specimens. Border furrow broad and shallow. Border broad, of even width
between the strong posterolateral marginal spines. Surface smooth.

Discussion. In most features the present form falls within the wide range shown
by Westergard’s (1946: pl. 3, figs 3-7) figures of P. fallax minor. It differs from
many Peronopsis species in that the pygidial flanks are confluent behind the axis,
and resembles P. fallax minor but not P. fallax ferox (Tullberg) in the character of
the cephalic border (Westergard 1946 : 38). The cephala agree with Westergard’s
figures although the glabella is generally longer than in his fig. 4 and the anterior

CAMBRIAN OF IRAN 327

glabellar lobe is proportionally longer than in his fig. 3; neither figure shows a
preglabellar depression or elongated Ptychagnostus-like basal lobes. A specimen of
P. fallax ct. minor (PI. 12, fig. 15) from the A. pisiformis Zone of central England
(Taylor & Rushton 1972 : 19) has a weak preglabellar depression but differs in having
a longer anterior glabellar lobe and smaller basal lobes ; the pygidial axis is evenly
rounded behind. Probably compression has caused the lateral margins of the
pygidium to converge backwards slightly.

Some of the pygidia of ‘Agnostus’ simplexiformis Rozova (1964 : 24) resemble that
of the present form. Rozova figured no cephala of ‘A.’ semplexiformis but Lazarenko
& Nikiforov (1968 : pl. 3, figs 1-4) showed that both cephalon and pygidium have
constricted acrolobes (Opik 1967 : 68). Judging from these figures we would transfer
‘A.’ simplexiformis to Agnostoglossa Opik (1967 : 145).

Family DORYPYGIDAE Kobayashi 1935
Genus DORYPYGE Dames 1883

TYPE SPECIES. Original designation, Dorypyge richthofeni Dames 1883.

Discussion. About forty specific names have been applied to Dorypyge, sensu
stricto, but according to Kobayashi’s revision (1960: 347) many of these are
synonyms. Kushan (1973) has discussed six forms from Iran.

Dorypyge sp.
Pioai, figs 1, 2

MATERIAL. One fragmentary pygidium about 7 mm long, It 13461.

DESCRIPTION. Outline, excluding spines, trapezoidal, width nearly 1-5 times
length. Axis only slightly tapered backwards, composed of three well-defined
rings, a fourth less well defined, and a rounded bulbous terminal part. Flanks
slightly narrower than axis anteriorly, marked by three pairs of narrow pleural
furrows but no interpleural grooves. Border broad, the anterior pleural furrows
crossing it indistinctly. Medially, the border forms a narrow transverse connection
behind the axis. Convex parts of surface covered with sparse coarse granules. The
marginal spines are distinctive although not all are preserved. The pygidium
follows the common plan of Dorypyge species in having four pairs of main marginal
spines anterior to the largest pair, but the present form is unusual, and perhaps
unique, in showing a small subsidiary spine anterior to each of the first three of the
‘main’ spines ; their bases make the outline of the pygidium jagged. Whereas the
thickness of the ‘main’ spines increases from the first to fourth pair, the subsidiary
spines are progressively thinner from the first to third pair. The fifth pair of
marginal spines is represented only on the left by a scar where the spine is broken
off ; it was twice as thick as the fourth spine. The outline of the posterior part of
the pygidium suggests the former presence of a substantial sixth pair of spines. The
doublure, which is partly exposed, is a smooth band parallel to the margin and
convex ventrally.

328 CHELIDONOCEPHALUS FAUNA

Discussion. The present form differs from all species so far described in having
subsidiary pygidial spines. However Dr P. Jell has told us of certain Kootenia
species which also show ‘subsidiary’ marginal spines which point more nearly
ventrally and are correspondingly difficult to observe. Their presence suggests
that similar spines may be found in other Dorypygidae. In other respects the
present form resembles the subspecies D. richthofeni latwuensis Kobayashi, which
lacks interpleural grooves, and the form figured by Sun (1924: pl. 2, figs 3 c, d)
which Kobayashi (1960 : 348) referred to D. 7. laevis. Sun’s fig. 3d shows that the
marginal spines increase in size backwards and also shows sparse granulation, but
it differs in having slightly wider flanks and only three axial rings. The present
specimen is distinguished from D. ivanensis Kushan and D. khademi khademi Kushan
by the granular surface, and from D. khademi papillosa Kushan by the absence of
interpleural grooves. The fragment illustrated by Kushan (1973: pl. 28, fig. 8) as
D. richthofeni n. subsp. aff. richthofeni differs slightly in having wider axial ring
furrows and pleural grooves, and a wider (sag.) post-axial part of the border.

Family ASAPHISCIDAE Raymond 1924
Genus IRANOLEESIA King 1955
(Pro Irania King 1937, non De Filippi 1863)

DiaGnosis. Asaphiscid trilobites with glabella slightly tapering, truncate
anteriorly. Posterior two pairs of glabellar furrows deep, inner half of 1P bifurcate,
posterior branch directed obliquely backwards. Occipital ring may be subdivided
with lateral lobes. Postocular fixed cheeks small (sag., trans.).

TYPE SPECIES. Original designation, Ivania pisiformis King 1937.

Discussion. The type species (and, with the removal from Ivanoleesia of Irania
falcom King 1937, the only species) of Ivanoleesia is redescribed below. Kobayashi
(1967 : 439) indicated that Ivanoleesia may be a junior synonym of Hundwarella
Reed 1934, from the Cambrian of Kashmir. The two specimens on which the
type and only species of Hundwarella is based, both cranidia, show apparently
bifurcate inner ends of the 1P glabellar furrows (Reed 1934: pl. II, figs 5, 6) which
connect across the middle of the glabella. This is not the case in Ivanoleesia
pisiformis, but the Kashmir specimens are slightly crushed, and the apparent
transverse furrows may be the result of compressing bifurcate furrows of Ivanoleesia
type. A similar effect may be observed on flattened specimens of Hypermecaspis
(compare Fortey 1974: pl. 13, fig. 1 with Lake 1913: pl. 7, fig. 3). Even if this is
accepted Ivanoleesia pisiformis differs from Hundwarella in having a transversely
truncate front margin of the glabella, the preglabellar field only half as wide (sag.),
and a strongly punctate surface sculpture. Since the significance of such characters
in discriminating asaphiscid genera is still imperfectly appreciated, Ivanoleesia is
therefore retained here pending a revision of the whole group. For example, the
genus Iniotoma Opik 1967, from the Mindyallan, has apparently similar glabellar
furrows (Opik 1967 : 232, fig. 81) to those of Ivanoleesia, but they are less impressed
and the eye ridges are less distinct (1967 : pl. 11, figs 1-3). While it seems probable

CAMBRIAN OF IRAN 329

that these genera are separated more by geography than by morphology, a revision
of the genera is beyond the scope of the present account. The genus Protohedinia
Endo 1937 (the type species of which has been illustrated by Chang 1963: pl. 1,
fig. 9) has glabellar furrows similar to those of Ivanoleesia and has palpebral lobes
similarly placed, eye ridges of like prominence, and border and preglabellar fields
of similar proportions. It is difficult to see why it has been included in a separate
family Tengfengiidae by Chang (1963 : 458).

Iranoleesia pisiformis (King 1937)
Pl. 9, figs 6, 8-10, 12
1937 Ivania pisiformis King : 12-13; pl. II, figs 6a-c

Diacnosis. Ivanoleesia species with pitted “surface sculpture. Anterolateral
corners of glabella angulate. Anterior border furrow without plectrum.

HototyPe. Although not named as such, it is clear from the plate description
and measurements given by King (1937:13; pl. II, fig. 6) that he intended the
incomplete cranidium, GSI 16305, as holotype of this species. We figure a cast of
this original on Pl. 9, fig. 6.

TYPE LOCALITY AND HORIZON. Ma’dan, locality 12 of King (1937). The exact
stratigraphic relationship of this species to the new Chelidonocephalus fauna is not
known for certain but Kushan recorded I. 1szformis only from horizons below those
with Koldimella mitella, suggesting that the Chelidonocephalus fauna is slightly the
younger.

FIGURED MATERIAL. Cranidia In 36890 (cast of holotype, Pl. 9, fig. 6),
In 36910 (Pl. 9, fig. 8), In 36911 (PI. 9, fig. 10), In 36912 (PI. 9, figs 9, 12).

DeEscRIPTION. Only cranidia known with certainty, and available material
largely decorticated, but otherwise well preserved in a limestone matrix. General
cranidial outline subquadrate, with anterior margin not greatly protruded, glabella
but slightly elevated above fixed cheeks, sloping downwards anteriorly into pre-
glabellar field. Glabella (with occipital ring) almost rectangular, extending to
three quarters cranidial length, tapering very gently forwards, so that the axial
furrows enclose an angle of about 15°. Anterolateral corners of the glabella are
obtusely angulate, resulting in a truncate anterior glabellar margin. Four pairs of
lateral glabellar furrows, of which the anterior two pairs are shallow and probably
would not be visible on the dorsal surface of the cuticle. 1P has its outer end
opposite the midpoint of the palpebral lobe, its outer third running transversely or
slightly posteriorly, at which point it bifurcates into a short, slightly anteriorly-
directed branch, and a longer, strongly backward-directed posterior branch, the
inner end of which terminates at about one third glabellar length. 2P, with its outer
end almost opposite the front of the palpebral lobe, is arcuate, the inner end curving
backwards in line with the inner end of rP. 3P and 4P opposite the eye ridge, the
former isolated within the glabella, transverse, the latter short, slightly forward-
inclined, placed laterally to 3P. Occipital furrow deep, forward-curved medially,

330 CHELIDONOCEPHALUS FAUNA

and shallowing laterally ; occipital ring with indistinctly defined lateral lobes and
prominent median tubercle. Axial furrows distinct, as is the preglabellar furrow on
the smaller cranidium (PI. 9, fig. 8), although on larger specimens this furrow becomes
shallower to almost effaced medially.

Intraocular fixed cheeks about half width (trans.) of adjacent glabella, horizontal.
Preglabellar field about one sixth length of glabella (sag.), scarcely downsloping
medially ; that part of the fixed cheeks in front of the eye ridge slopes downwards
more steeply and is about twice the length (exsag.) of the preglabellar field (seen in
dorsal view). The eye ridges are strong on internal moulds, converging forwards
at about 65° to the sagittal line. Palpebral lobes of about same width as the eye
ridges, in a posterior position such that the transverse line connecting their anterior
limits crosses the glabella at two thirds its length ; they are of length (exsag.) one
third, or slightly less, that of glabella. Anterior border furrow defining change in
slope from preglabellar field to broad, backward-sloping anterior border, the width
of which (sag., exsag.) is similar to that of the preglabellar field. Postocular cheeks
poorly shown by present material.

Genus ANOMOCARELLA Walcott 1905

Diacnosis. Asaphiscid trilobites with glabellar furrows faint or absent ;
preglabellar field less than one fifth glabellar length ; anterior border furrow with
narrow (trans.) plectrum; anterior border flat. Pygidium with distinct flat
border.

TYPE SPECIES. Original designation, Anomocarella chinensis Walcott 1905.

Discussion. The species described below as Anomocarella falcont was originally
assigned to Ivama by King (1937), and hence subsequently to Ivanoleesza.
Kobayashi (1967 : 493) preferred to regard the species as belonging to Grandioculus
Cossmann. The type species of that genus, G. megalurus (Dames) (Dames 1883 : 20 ;
pl. 1, figs 7, 8, 10. Walcott 1913:192; pl. 18, figs ga-f) has large, somewhat
posteriorly placed palpebral lobes, weakly defined glabellar furrows and a flat
border like that of the Iranian specimens, but the border furrow lacks a plectrum,
which is also the case in Ivanoleesia pisiformis. The type species of Anomocarella
is A. chinensis Walcott 1905, the lectotype of which was carefully identified by
Endo & Resser (1937 : 164-165). This cranidium is only one of the specimens used
by Walcott (1913: pl. 20, figs 3, 3a—d, 4, 4a) to illustrate A. chinensis, and all the others
are from a different locality from the lectotype (pl. 20, fig. 3c) and include more than
one species (Endo & Resser 1937: 165). It is the other material (the specimens
of Walcott 1913: pl. 20, figs 3, 3e) that is used to illustrate the type species of
Anomocarella in the Treatise (Howell 7m Harrington ef al. 1959: 292). Endo &
Resser emphasized the presence of a plectrum as ‘the most distinctive generic
feature’ of Anomocarella, with which we agree. The major point of difference
between the Iranian material and the type species is the shorter preglabellar field
of the latter ; this is not regarded as of generic importance, as it is a variable feature
in other species assigned to Anomocarella by Endo & Resser (1937). Further back-
ward migration of the plectrum (or further shortening of the preglabellar field)

CAMBRIAN OF IRAN 331

results in the broadly backward-deflected border furrows seen in Mapania (Opik
1961 : fig. 53).

Anomocarella and allied genera are sometimes referred to a separate family
Anomocarellidae Hupé 1955. Since the characters pertinent to the division of
genera within the Asaphiscidae and Anomocarella-group are unclear at best, and
there is no phyletic concept on which to base separate families, Anomocarella is here
retained doubtfully within the Asaphiscidae.

Anomocarella falconi (King 1937)
ribo. ne -1?> Pl, 10} figs 1=5, 8
1937 Ivania falconit King : 13-14; pl. II, figs 7a-f.

Diacnosis. An Anomocarella species with transverse width of plectrum half to
two thirds that of anterior margin of glabella ; preglabellar field of moderate width ;
dorsal surface of cuticle minutely granulose.

HoLotype. An external mould of a cranidium, a cast from which was figured
by King (1037: pl. II, fig. 7a), GSI 16306. This cast is here refigured on
Pie xo, fig. 1.

TYPE LOCALITY AND HORIZON. Mila Formation at Darreh Shu (lowest beds),
locality 8 of King (1937), whence Chelidonocephalus alifrons is also recorded (King
1937 : 17).

FIGURED MATERIAL. Cranidia In 36891 (cast of holotype, Pl. ro, fig. 1), In 36908
(Pl. ro, fig. 3), In 36909 (PI. 10, fig. 2), In 36909b (PI. Io, figs 4, 5, 8). Doubtfully
assigned pygidium In 36892 (cast, Pl. 9, fig. 11).

DESCRIPTION. Cranidium of low convexity, downward-sloping in front of eye
ridges, glabella not greatly vaulted above intraocular cheeks. Maximum cranidial
width at posterior margin about one and a half times transverse width at anterior
border. Glabella extends to about three quarters cranidial length, and tapers
uniformly forwards, the axial furrows enclosing an angle between 15° and 20°.
Anterolateral corners of glabella rounded ; axial furrow deeper than preglabellar
furrow, and both better defined on internal moulds. Occipital and glabellar
furrows faint, best seen as smooth, slightly depressed areas on specimen retaining
exoskeleton (Pl. ro, fig. 8). 1P to 3P almost touch axial furrows; distance (exsag.)
between 1P and 2P equal to that between IP and the occipital furrow, but less
than that between 2P and 3P; 3P is level with the point at which the eye ridges
touch the glabella, transverse; 1P and 2P slope slightly backwards. 4P is very
close to 3P but external to it, continuing in line with the furrow defining the posterior
of the eye ridge. Occipital ring widest (sag.) medially, defined by broad, shallow
furrow, of which only the transverse, median section is visible on the internal mould
(Pl. ro, fig. 3); outer ends of this section deepened into internal muscle scars.
External surface shows narrow parafrontal band in front of midpart of frontal
glabellar lobe ; this band extends further laterally on the internal mould, but neither
on dorsal nor on ventral surfaces is the connection with the eye ridge displayed.

332 CHELIDONOCEPHALUS FAUNA

Palpebral lobes of half length of glabella, anterior limit opposite outer ends of 3P
furrows ; palpebral rims broad, defined by shallow palpebral furrows which are
outward-bowed medially. Maximum transverse width of intraocular cheek half
width of adjacent glabella. Eye ridges not visible abaxially on dorsal surface of
exoskeleton, but their confluence with the frontal lobe of the glabella is clearly
shown by anterior shallowing of the axial furrows. The eye ridges, where seen, are
only slightly oblique, virtually transverse on the smallest cranidium. Postocular
fixed cheeks narrow, width (exsag.) less than that of occipital ring, bisected by
strong border furrow. Narrow posterior border carries small articulating socket (?)
at about mid-width, maximum transverse width of border less than that of occipital
ring. Preglabellar field of length (sag.) between 0-14 and 0-17 times glabellar length
on available material. Anterior border furrow marks an abrupt change in slope
anterolaterally, where it is gently bowed forwards; medially there is a distinct
plectrum, of transverse width about half to two thirds that of frontal lobe of glabella.
Laterally the wide anterior border is horizontal to slightly declined, but in front of
the plectrum more nearly carries forwards the downward slope of the preglabellar
field. There are faint indications of caeca crossing the border furrow. The dorsal
surface of the exoskeleton is covered with minute granules.

An incomplete pygidium (Pl. 9, fig. 11) was assigned to this species by King
(1937: pl. II, fig. 7b). It has a convex rhachis showing three or four axial rings and
a broad flattened border. It also resembles the posterior part of the pygidium
here tentatively assigned to Chelidonocephalus preannulatus sp. nov. (p. 338), and
it must be admitted as a possibility that the pygidium is that of C. alifrons. Our
material is inadequate to resolve the question.

Discussion. Anomocarella falcom differs from the type species A. chinensis
(Walcott 1913: pl. 20, fig. 3c. Endo & Resser 1937: pl. 34, fig. 6) in having less
oblique eye ridges (due to an anterior forward limit of the palpebral lobe) and a
longer (sag., exsag.) preglabellar field; the surface sculpture of A. chinensis is
described as punctate. Of the many species assigned by Walcott (1913 ; some
re-illustrated in Lu et al. 1965 : pls 59-61) to Anomocarella only the type species has a
plectrum, and the others should be excluded from the genus. One of the species
from the Mapan Formation, A. concava Endo & Resser (1937: 167; pl. 35, fig. 8),
approaches A. falconi closely in the proportions of preglabellar field and plectrum.
It differs only in having a glabella that hardly tapers forwards, and anterior limits
of the palpebral lobes that approach the glabella more closely.

Genus KOLDINIELLA Sivov 1955

TYPE SPECIES. Original designation, K. mitella Sivov 1955.

Discussion. This genus is characterized by a broad, flat cephalic border and
effacement of the other cephalic characters. Traces of the axial furrow suggest that
the glabella extends to or nearly to the anterior border (Kushan 1973: pl. 31, fig. 3) ;
this feature, the more backward eyes and the transverse pygidium distinguish
Koldimella from the more effaced forms of Maryvillia and Blountia (Rasetti 1965 :

CAMBRIAN OF IRAN 333

pls 9, 10). Koldiniella is similar to but more effaced than the asaphiscid Pe:shania
Resser & Endo from the late Middle Cambrian ; it may even have been derived from
a form like P. lubrica Chang (1957: pl. 1, fig. 2) which has a comparable frontal
border but has the axial furrows and occipital furrow more distinct. Peishania
(and Liopeishama Chang 1963, to which P. lubrica was transferred in Lu et al. 1965)
have longer pygidia than Koldiniella mitella. The dorsal surface of the cranidium
of Liopeishania spannensis Palmer & Gatehouse is more effaced than other species
and approaches Koldimella in this respect. It differs in having a shorter anterior
border which is curled down and under in front (Palmer & Gatehouse 1972: pl. 4,
fig. 0).

Koldimella has been variously grouped with the families Illaenidae, Illaenuridae
and Tsinaniidae on account of its effaced cephalon, but there are no characters
uniting it convincingly with any of these families and the frontal border is a feature
distinguishing it from all (and from Kzingstonia). It perhaps comes closest to the
Illaenuridae because Jl/aenurus has a transverse pygidium and in Macelloura the
glabella reaches to the frontal border.

Koldiniella mitella Sivov 1955
PLO sso, 7, 9,10, 15, 13 2) 14°, 16"?

FIGURED MATERIAL. Cranidium It 14021 (Pl. io, figs 6, 7); pygidia It 13477
(Pl. ro, figs 9, 10), It 14022 (Pl. 10, fig. 15); hypostome assigned to the species
It 13506 (PI. ro, fig. 16) ; cranidium doubtfully assigned, It 13476 (PI. ro, figs 13, 14).
Ten unfigured specimens also present.

Discussion. Kushan (1973: 149) redescribed this species using material from
Iran. Material in the present collection is identical. A hypostome tentatively
referred to K. mitella (Pl. io, fig. 16) has a gently convex median body divided by a
faint transverse furrow. The lateral border widens out posterolaterally but is
narrow posteromedially. The surface is smooth except posterolaterally where it is
striated parallel to the margin. This hypostome cannot be that of Dorypyge
because it differs from that referred to Dorypyge aenigma (Westergard 1948 : pl. 2,
fig. 7) ; the smooth surface and striated margin recall Koldimella mitella rather than
the associated Chelidonocephalus preannulatus.

K. mitella is distinguished from K. convexa Lazarenko by the longer and flatter
anterior border and by the weaker axial furrows at the base of the glabella (Rozova
1964 : pl. 15, figs I-5).

K. proixa Lazarenko (in Lazarenko & Nikiforov 1968: pl. 1, figs 16-19) has a
longer, more salient frontal border and a much longer pygidium, nearly as long as
wide.

One small cranidium 1-7 mm long (PI. 10, figs 13, 14) has a much narrower (sag.)
frontal border, less than a tenth of the length of the cranidium. It may possibly bea
juvenile form of K. mitella because among larger specimens the border seems to have
grown more rapidly than the rest of the cranidium, but the border is convex instead
of being flat or even slightly concave as it is in specimens of K. mitella 3 mm or more
in length. The small cranidium may therefore be referable to Parakoldinia Rozova

25

334 CHELIDONOCEPHALUS FAUNA

1960. It is most closely comparable with P. lana Rozova (1960: pl. 7, figs 1-5), a
broad and not especially convex species with a distinct frontal border, horizontal in
front view, which does not show strong striae. Unfortunately the present specimen
is too different in size for an exact comparison to be made, but Rozova’s fig. 3 shows
the border furrow more sharply cut and the border dropping abruptly downwards in
front, whereas in the present specimen the border curves down more gradually in
front.

Genus TSINANIA Walcott 1914

TYPE SPECIES. Original designation, I/laenurus canens Walcott 1905.

Discussion. Kobayashi (1935 : 303) has suggested that Tsinmanza is an effaced
member of the Asaphiscidae.

Tsinania ? sp.
Pl. ‘ro, figs ‘11; 12, 17—"9

FIGURED MATERIAL. Cranidia It 13493 (Pl. 10, figs 11, 12), It 13458 (Pl. ro,
figs 17-10).

DESCRIPTION. Cranidium convex. Axial furrows effaced at external surface but
a trace of their ends seen on the exfoliated specimen. The base of the glabella is a
little more than half as wide as the cranidium. No trace of occipital or glabellar
furrows. In palpebral view the middle of the eyes lies just behind the mid-length.
Palpebral furrow very faint. Preocular sutures diverge forwards slightly but curve
inwards across the anterior border. Postocular sutures diverge backwards obliquely,
straight. A very faint border furrow separates off a narrow, barely differentiated
frontal border. Pleuroccipital furrow weak. Surface smooth.

Discuss1on. These specimens agree closely with Tsinamia canens (Walcott 1913 :
pl. 23, fig. 2) except that the eyes seem farther forward, the base of the glabella is
wider and there is a very faintly marked border. Kobayashi (1952) revised Tsinama
canens and showed that the glabella is less than half as wide as the cranidium. No
published figures show a frontal border. Tsimania? sp. also resembles Pletho-
metodus obtusus Rasetti (Longacre 1970 : pl. 6, figs 14, 15) except that the latter has
a well-marked occipital furrow and curved postocular sutures. Plethometopus
convergens (Raymond) has only a weak occipital furrow but the preocular sutures
converge forwards (Longacre 1970: pl. 3, figs 11, 12). Plethopeltoides lepidus
Lazarenko (im Lazarenko & Nikiforov 1968: pl. 9, figs 1-7) has a better-marked
occipital furrow and is less convex. Wanwanoglobus convexus (Kobayashi 1966 : 265
and text-fig. 4) has a much more convex cranidium. We know of no Cambrian
trilobite which matches the present cranidium in all details.

Family UNCERTAIN
Genus CHELIDONOCEPHALUS King 1937

DiAGnosis. Cranidium opisthoparian, “ptychoparioid’, with glabellar furrows
(other than occipital) not incised. Broad false border furrow delimiting flattened

CAMBRIAN OF IRAN 335

or gently convex anterior cranidial border. True border furrow crosses preocular
cheek, cut by preocular suture at sutural mid-length (exsag.), inner ends of border
furrow curve backwards to define the plectrum. Narrow (sag.), transverse inflated
band in front of glabella. Broad palpebral lobes between one third and half
glabellar length, posterior limits opposite outer parts of occipital furrow defining
narrow (exsag.) postocular cheek. Width of intraocular cheek less than that of
adjacent glabella. Eye ridges subdued.

TYPE SPECIES. Original designation, C. alifrons King 1937.

Discussion. Chelidonocephalus was placed in the Alokistocaridae in the Treatise
(Howell 7m Harrington e¢ al. 1959 : 238), but other possibilities must now be con-
sidered. Chang (1959: 223) allied Chelidonocephalus with a number of Middle
Cambrian genera from China, including Poshania Chang 1957 (type species P.
poshanensis Chang 1957), Peichishania Chang 1957 (type species Eymekops rect-
angularis Endo & Resser 1937), Inowyella (type species I. peiensis Endo & Resser
1937), Ordosia (type species O. fimbricauda Lu 1954) and Tattzuia (type species
T. insueta Endo & Resser 1937). All these genera share with Chelidonocephalus the
backward-curved border furrow (termed below the ‘true’ border furrow), although
the development of the border itself is variable. In the Treatise (Harrington e¢ al.
1959) Chelidonocephalus, Ordosia, Inouyella and Taitzuia are assigned to four different
superfamilies, while Chang (1959 : 223) indicates that they may be included within
a single family Namanoiidae Lermontova 1951, based on the Lower Cambrian
genus Namanoia from eastern Siberia, which has a similar disposition of furrows
on the cranidial margin. We believe that the form of the border furrows is assuredly
significant at generic level, but that its significance at family level is more contentious.
The backward-curving border furrows are present, for example, on Llanoaspis
walcottt Resser (Rasetti 1965 : pl. 8, fig. 14) from the Upper Cambrian of Tennessee,
which is certainly unrelated to Chelidonocephalus and the other genera listed above,
having distinctively different facial sutures. Similarly Namanoza, with its quadrate
glabella and lack of parafrontal band, is unlikely to be closely related to Chelidono-
cephalus, which therefore cannot be assigned to the Namanoiidae.

We consider that Chelidonocephalus is most nearly related to Poshania, which is
identical in the combination of true border furrow and parafrontal band (Chang
1959: pl. 2, figs 4, 5), and differs significantly only in lacking a distinct false border
furrow. The resemblance is perhaps more apparent when comparing small cranidia
of Chelidonocephalus (Pl. 9, fig. 5) with adult Poshania as the whole border region
is then of similar proportions. Peichishania, Inouwyella, Ordosia and Taitzuia are
probably unrelated to these two genera ; Peichishania was assigned by Opik (1967 :
221) to the Auritamidae, Ovdosia by Lu (1954) to the Leiostegiidae (and Taztzuza
closely resembles Ordosia). Inouyella, with its tapering glabella with narrow,
backward-directed furrows (Endo & Resser 1937: pl. 46, figs 9, 14), is probably
unrelated to any of the foregoing.

The Alokistocaridae have been included within the concept of the Papyriaspididae
by Opik (1961: 149). He (Opik 1967 : 298) includes Poshania (misspelt Poshauia)
in the family Mapaniidae of his superfamily Rhyssometopacea, which might therefore

336 CHELIDONOCEPHALUS FAUNA

be expected to include Chelidonocephalus also. We briefly discuss below (p. 337) some
reasons for treating a rhyssometopacean affinity with caution.

This discussion shows to what an extent the familial (and superfamilial) concepts
are capable of varied interpretations in these late Middle to early Upper Cambrian
forms. For the time being, therefore, we have to leave Chelidonocephalus as of
uncertain family.

Chelidonocephalus alifrons King 1937
Pl. 8, figs 1-5

Diacnosis. Chelidonocephalus species with punctate surface sculpture. Anterior
border of cranidium gently convex. Palpebral lobes farther from glabella, and
convexity lower than that of C. preannulatus nov. sp. (p. 338).

LECTOTYPE. King (1937) did not specify a holotype for this species, but of his
original syntypic series that figured in his pl. II, fig. 8A corresponds with the measure-
ments cited in the text, and was probably intended as the type. We therefore
select this specimen, GSI 16311, as lectotype.

TYPE LOCALITY AND HORIZON. The lectotype is associated with Anomocarella
falcont (King) from a limestone 2300 feet (700 m) below the local top of the Cambrian,
at Chal-i-Sheh (loc. 3 of King 1937: text-fig. 1). The evidence from Kushan’s
(1973) faunal lists with Chelidonocephalus is that a late Middle Cambrian or an
earliest Upper Cambrian age is most probable for these beds.

FIGURED MATERIAL. Cranidia In 36893 (cast of lectotype, Pl. 8, fig. 1), In 36895
(Pl. 8, figs 2, 3, 5), In 36806 (PI. 8, fig. 4).

DESCRIPTION. Cranidia only identified with certainty, these being well pre-
served in limestone. Convexity is low, the whole surface of the cranidium sloping
gently downwards and forwards from the occipital ring. The largest specimen, a
fragmentary cranidium, probably had a cephalic length of about 20mm. Maximum
transverse width, across posterior margin, about one and a half times the width
across the anterior cranidial margin. Glabella tapering gently forwards, axial
furrows enclosing an angle of about 30°, extending to about o-7 of the cranidial
length (sag.) ; front margin truncate. Glabellar furrows absent, but muscle in-
sertion areas are represented by smooth areas lacking the pitted sculpture which
covers the rest of the glabella. There is a pair of large smooth areas centred on the
lateral parts of the occipital furrow, and three (? four) pairs of ill-defined smooth
areas on the flanks of the glabella. The occipital furrow is shallow except where
deepened into pits near its lateral extremities. Postocular and preocular fixed
cheeks downsloping ; intraocular cheek horizontal. Palpebral lobes about one
third glabellar length, with posterior limits slightly in front of points at which
the occipital furrow meets the axial furrows. Transverse width of the intraocular
cheek at anterior limit of palpebral lobe is about 0-7 of the width of glabella at same
section. Palpebral furrows are faint even on internal moulds. Eye ridges are

CAMBRIAN OF IRAN 53%

faint on the external surface of the cuticle, but clearly visible on internal moulds
(Pl. 8, fig. 4) ; their adaxial limits indent the course of the axial furrows. Posterior
border furrow deep and transverse ; convex posterior border of width (exsag.) sub-
equal to that of furrow. In front of the glabella there is a narrow (sag., exsag.)
transverse inflated band, defined posteriorly by the preglabellar furrow and an-
teriorly by the true anterior cranidial border furrow. This latter follows a trans-
verse, gently curved course a little anterior to the front margin of the glabella,
medially deflected sharply backwards when approaching the anterolateral corners
of the glabella, to form the median plectrum (Opik 1967 : 58), this area being inflated
above the genal areas adjacent to it. The anterior border furrow is cut by the
anterior branch of the facial suture at, or slightly in front of, the anterior sutural
mid-length. In front of the plectrum the preglabellar field continues a gently
downward slope into a broad, shallow false border furrow. The gently convex
anterior border widens medially, where it occupies one third to one quarter of the
total preglabellar length. Internal moulds are strongly caecate on frontal part of
cranidium ; caeca pass over both border furrows, but are not visible behind eye
ridge. Dorsal surface of cranidium is finely pitted except in furrows and over
muscle insertion areas. In front of the true border furrow, the pits are replaced by
more densely crowded, minute granules.

Discussion. The description of ‘true’ and ‘false’ border furrows used above
needs clarification. It is contended that the posterior of the two transverse furrows
crossing the preglabellar area represents the border furrow as understood in other
trilobites, and that the anterior furrow in Chelidonocephalus is independently
generated. Homologically, all that area anterior to the true border furrow should
be termed ‘anterior border’, but for convenience the area is divided into preglabellar
field, ‘false’ border furrow and anterior border in line with other trilobites. The
identification of the posterior of the two furrows with the anterior border furrow of
other trilobites is based on the fact that this furrow has a median plectrum, and
that the plectrum is invariably developed on the anterior border furrow. Since
plectra are developed in a number of groups which are, at least on present classi-
fications, only distantly related, its presence may be taken as showing the
position of the true anterior border furrow. The important defining character
of Chelidonocephalus is the presence of two anterior furrows. It is interesting
to note that some forms comparable in this respect are to be found in the
superfamily Rhyssometopacea Opik (1967 : 272) described from the Mindyallan of
Australia, for example Rhyssometopus princeps Opik (1967: pl. 27, fig. 1-3; text-
fig. 96), if we can accept Opik’s ‘frontal wrinkle’ (ibid. : text-fig. 93) as
homologous to our true border furrow. However, rhyssometopaceans show impor-
tant differences from Chelidonocephalus, notably in having large palpebral lobes
close to the glabella, and we are unable to classify our species with the Australian
superfamily.

A comparative description of Chelidonocephalus preannulatus sp. nov. is given
below, which extends the concept of the genus to include forms with larger palpebral
lobes, closer to the glabella. They do not, however, approach rhyssometopaceans
in this respect.

338 CHELIDONOCEPHALUS FAUNA

Chelidonocephalus preannulatus sp. nov.
Pl. 8, figs 6=8 5) Pio, Tes 1125, 38) 4a e

Diacnosis. Chelidonocephalus species with granulate surface sculpture. An-
terior border broad and flat. Palpebral lobes larger, placed closer to the glabella
than in C. alifrons, and convexity and furrow incision generally greater.

DERIVATION OF NAME. Latin — ‘with a ring in front’ — referring to the inflated,
transverse preglabellar band.

HorotyPeE. BM(NH) It 13503, incomplete cranidium partly preserving exo-
skeleton, figured on PI. 9, fig. 2.

TYPE LOCALITY AND HORIZON. Mila Formation, Sanghabad, Taleghan Range,
Alborz Mountains, northern Iran.

FIGURED MATERIAL. Cranidia It 13478 (PI. 8, figs 6-8), It 13504 (Pl. 9, fig 5) ;
free cheek It 13499 (Pl. 9, fig. 1) ; hypostoma, tentatively assigned, It 14020 (Pl. 9,
figs 3, 4); pygidium, tentatively assigned, It 13460 (Pl. 9, fig. 7). Six unfigured
fragments also present.

DISCUSSION. Since this species is similar to the type species C. alifrons in most
proportional characters, a detailed description need not be repeated here. C. pre-
annulatus differs from the type species in the following characters.

(i) Convexity (sag., exsag.) is greater, this being due to the increase in the down-
ward deflexion of the fixed cheek in front of the palpebral lobes.

(ii) Anterior border is flat, not gently convex, forming a broad, horizontal rim
around the cranidial margin. The border occupies about half (sag.) the preglabellar
width, although the false border furrow, except on small cranidia, is ill defined.

(iii) Anterior branches of the facial sutures appear more divergent in dorsal view,
this being due to the greater downward inclination of the preocular cheeks.

(iv) Length of palpebral lobes about half length of glabella. They are also closer
to the glabella, such that the transverse width of the intraocular cheek at the anterior
limit of the palpebral lobe is about half width of glabella at same section (0-7 times
in C. alifrons).

(v) Furrows more deeply incised.

(vi) Surface sculpture of fine granules on glabella. Smooth areas on the glabella
of one specimen (Pl. 8, figs 6-8) display the muscle insertion areas rather better
than C. alifrons. Lateral occipital impression deep, producing a lateral narrowing
of the occipital ring. 1P subcircular, indistinct, isolated within glabella near mid-
line ; 2P of similar size and form adjacent to axial furrow opposite mid-part of
palpebral lobe ; 3P opposite anterior end of palpebral lobe, narrow and transverse ;
4P of similar form, slightly anteriorly directed.

A free cheek (Pl. 9, fig. 1) shows a prominent eye socle, faint caeca radiate from
the eye and there is a lateral continuation of the true and false border furrows
which describe a ridge on the border. The counterpart shows the base of a strong
genal spine.

A small hypostoma (Pl. 9, figs 3, 4) may belong to C. preannulatus. It is
different from that assigned to Koldiniella, and from that of Dorypyge, so that

CAMBRIAN OF IRAN 339

Chelidonocephalus is the most likely polymerid known from the present fauna to
which it could belong. A distinctive feature is a prominent anterior hump on
the middle body.

An incomplete associated pygidium is tentatively referred to this species because it
shows traces of similar granular sculpture. The axis is wider than the flanks, tapers
back, is rounded behind but tends to pass into a postaxial ridge ; it is composed of
4-5 rings and a terminal piece. Flanks show two distinct and one less clear pleural
furrows. Nointerpleuralfurrows. Thereisa suggestion ofa flattened lateral border.

REFERENCES

ANNELLS, R. N., ARTHURTON, R. S., BazLey, R. A. B., & Davies, R.G. 1976. Explanatory
text of the Qazvin and Rasht Quadrangles Map. Geological Quadrangles Nos E3 and E4.
Geological Survey of Iran, Teheran.

CHANG W.-T. 1957. Preliminary note on the Lower and Middle Cambrian stratigraphy of

Poshan, Central Shantung. Acta palaeont. sin., Peking, 5: 29-31, 2 pls.

1959. New trilobites from the Middle Cambrian of North China. Acta palaeont. sin.,
Peking, 7 : 215-236, 4 pls.

1963. A classification of the Lower and Middle Cambrian trilobites from North and
Northeastern China, with description of new families and new genera. Acta palaeont. sin.,
Peking, 11 : 475-487, 2 pls.

Dairy, B. & Jaco, J. B. 1975. The trilobite Lejopyge Hawle and Corda and the Middle —
Upper Cambrian boundary. Palaeontology, London, 18 : 527-550, pls 62-63.

Dames, W. 1883. Cambrische Trilobiten von Laiu-Tung. Ju RICHTHOFEN, F. von, China
A: 3-33, Pls 1,2. Berlin.

Enpo, R. & Rresser, C. E. 1937. The Sinian and Cambrian formations and fossils of southern
Manchoukuo. Bull. Manchur. Sci. Mus., Mukden, 1. ix+474 pp., 73 pls.

Fortey, R. A. 1974. The Ordovician trilobites of Spitsbergen. I. Olenidae. Shr. norsk
Polarinst., Oslo, 160. 81 pp., 24 pls.

HarRINGTON, H. J. e¢ al. 1959. Systematic descriptions, Class Trilobita. Jn Moore, R. C.
(ed.), Treatise on Invertebrate Palaeontology,O. xix+560 pp., 415 figs. Lawrence, Kansas.

Kine, W. B. R. 1937. Cambrian trilobites from Iran (Persia). Mem. geol. Surv. India
Palaeont. Indica, Calcutta, 22 (5). 22 pp., 2 pls.

1955. JIvanoleesia, new trilobite name. Geol. Mag., Hertford, 92: 86.

KOBAYASHI, T. 1935. The Cambro-Ordovician formations and faunas of South Chosen.

Palaeontology, part 3. J. Fac. Sci. Tokyo Univ. sec. 2, 4: 49-344, pls 1-24.

1952. On Mansuyia and the Tsinanidae. Tvans. Proc. palaeont. Soc. Japan, Tokyo,
N.S. 5: 145-154, pl. 13.

1960. The Cambro-Ordovician formations and faunas of South Korea. Part VII, Palae-
ontology VI. J. Fac. Sci. Tokyo Univ. sec. 2, 12: 329-420, pls 19-21.

1966-67. The Cambro-Ordovician formations and faunas of South Korea. Part X,
Stratigraphy of the Chosen Group in Korea and South Manchuria and its relation to the
Cambro-Ordovician formations of other areas. Section B. The Chosen Group of North
Korea and Northeast China. J. Fac. Sci. Tokyo Univ. sec. 2, 16: 209-311. Section C.
The Cambrian of Eastern Asia and other parts of the Continent. Jbid.: 381-535.
KusuHAn, B. 1973. Stratigraphie und Trilobitenfauna in der Mila-Formation (Mittelkambrium-

Tremadoc) im Alborz-Gebirge (N-Iran). Palaeontographica, Stuttgart, A 144 (4-6) : 113-
165, pls 26-34.

Lake, P. 1906-46. British Cambrian trilobites. Palaeontogr. Soc. (Monogy.), London. 350
PP.» 47 pls.

LAZARENKO, N. P. & NikiForov, N.I. 1968. [Assemblages of trilobites from Upper Cambrian
strata on the River Kulyumbe (northwestern Siberian Platform).] Uchen. Zap. nauchno-
issled. Inst. geol. arkt., Leningrad, Pal. Biostrat. 23 : 20-80, pls 1-15. [In Russian.]

340 CHELIDONOCEPHALUS FAUNA

LermMontova, E. V. 1951. [Lower Cambrian trilobites and brachiopods from Eastern Siberia.]
222 pp., 21 pls. Moscow (Vses. nauchno.-issled. geol. Inst. (VSEGEI)). [In Russian.]
LocuMan,C. 1944. Jn LocHmMan, C. & Duncan, D. Early Upper Cambrian faunas of Central

Montana. Spec. Pap. geol. Soc. Am., Baltimore, 54. 181 pp., 19 pls.
LonGacrE, S. A. 1970. Trilobites of the Upper Cambrian Ptychaspid Biomere, Wilberns
Formation, Central Texas. Rep. Invest. Bur. econ. Geol. Univ. Texas, Austin, 66: I-70.
Lu Y.-H. 1954. Two new trilobite genera of the Kushan Formation. Acta palaeont. sin.,
Peking, 3 : 425-438, I pl.

et al. 1965. [Fossils of each group of China: Chinese Trilobites.| 766 pp., 134 pls,
2vols. Peking. [In Chinese.]

Opix, A. A. 1961. The geology and palaeontology of the headwaters of the Burke River,
Queensland. Bull. Bur. Miner. Resour. Geol. Geophys. Aust., Melbourne, 53. 249 pp.,
24 pls.

1967. The Mindyallan fauna of north-western Queensland. Bull. Bur. Miner. Resour.
Geol. Geophys. Aust., Canberra, 74. (1) 404 pp.; (2) 167 pp., 67 pls.

PALMER, A. R. & GATEHOUSE, C.G. 1972. Early and Middle Cambrian trilobites from Antarc-
tica. Prof. Pap. U.S. geol. Surv., Washington, 456-D. 37 pp., 6 pls.

Rasetti, F. 1965. Upper Cambrian trilobite faunas of northeastern Tennessee. Smithson.
misc. Collns, Washington, 148 (3). 127 pp., 21 pls.

REED, F. R. C. 1934. Cambrian and Ordovician fossils from Kashmir. Mem. geol. Surv.
India Palaeont. Indica, Calcutta, 21 (2). 38 pp., 2 pls.

RESSER, C.F. & ENDO, Re 1037. i Endo, Nh. .& essen, ©. Gay,

Rozova, A. V. 1960. [Upper Cambrian trilobites of the Salayr.] Tvudy Inst. Geol. Geofiz.
sib. Otdel., Novosibirsk, 5. 116 pp., 8 pls. [In Russian.]

1964. [Biostratigvaphy and description of trilobites of the Middle and Upper Cambrian of
the north-west Siberian Platform.] 148 pp., 19 pls. Moscow, Akad. Nauk SSSR, Sibir.
otdel. Inst. Geol. Geofiz. (‘Nauka’). [In Russian.]

1968. [Biostratigraphy and trilobites of the Upper Cambrian and Lower Ordovician of
the north-west Siberian Platform.] Tvudy Inst. Geol. Geofiz. sib. Otdel., Moscow, 36.
196 pp., 17 pls, 5 tab. [In Russian.]

Stvov, A.G. 1955. Jn Kuatrina, L. L. [Adlas of guide forms in the fossil faunas and floras
of western Siberia.| 327 pp. Moscow. [In Russian.] (fzde Kushan, not seen.)

STOCKLIN, J. 1968. Structural history and tectonics of Iran: a review. Bull. Am. Ass.
Petrol. Geol., Tulsa, 52 : 1229-1258.

—, Ruttner, A. & NasBavi, M. 1964. New data on the Lower Paleozoic and Pre-Cambrian
of North Iran. ep. geol. Surv. Ivan, Teheran, 1. 29 pp., 1 pl.

Sun Y.-C. 1924. Contributions to the Cambrian faunas of North China. Palaeont. sin.,
Peking, B1 (4). 109 pp., 5 pls.

TayYLor, K. & Rusuton, A. W. A. 1972 [1971 on cover]. The pre-Westphalian geology of

the Warwickshire Coalfield.... Bull. geol. Surv. Gt By., London, 35. 152 pp., 12 pls.
Wa tcott, C. D. 1905. Cambrian faunas of China. Pvoc. U.S. natn. Mus., Washington,
26 : 1-106.

1913. The Cambrian faunas of China. Jn WILLIs, B. et al., Research in China 3: 1-276,
pls 1-24. Carnegie Inst., Washington (publ. 54).
WESTERGARD, A. H. 1946. Agnostidea of the Middle Cambrian of Sweden. Sver. geol.
Unders. Avh., Stockholm, C 477. 140 pp., 16 pls.

1948. Non-Agnostidean trilobites of the Middle Cambrian of Sweden. I. Sver. geol.
Unders. Avh., Stockholm, C 498. 32 pp., 4 pls.

R. A. Fortry, Ph.D. A. W. A. Rusuton, B.A., Ph.D.
Department of Palaeontology INSTITUTE OF GEOLOGICAL SCIENCES
BritisH Museum (NaTurRAL History) EXHIBITION Roap

CROMWELL Roap Lonpon SW7 2DE

Rese oe) Accepted for publication 9 February 1976

PLATE 8

Chelidonocephalus alifrons King 1937 (p. 336)
Mila Formation, Chal-i-Sheh, Loc. 3 of King 1937
I'ic. 1. Plaster cast of lectotype (selected herein), incomplete cranidium, x4. Specimen
figured by King 1937: pl. 2, fig. 8a, GSI 16311 (cast In 36893).
FIGs 2, 3, 5. Cranidium retaining cuticle, dorsal and lateral views x5; Fig. 5, detail of right

side of glabella showing surface sculpture of fine pitting, x 12:5. In 36895.

Fic. 4. Large exfoliated cranidium, showing caeca and form of facial suture, dorsal view, x 3.
In 36896.
Chelidonocephalus preannulatus sp. nov. (p. 338; see also Pl. 9, figs 1-5, 7)

Mila Formation, near Sanghabad, Alborz Mountains

Fics 6-8. Cranidium retaining cuticle, oblique lateral and dorsal views, x 4; Fig. 8, detail of
right side of glabella to show sculptural difference from C. alifrons, x12. It 13478.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

PLATE 9

Chelidonocephalus preannulatus sp. nov. (p. 338; see also Pl. 8, figs 6-8)
Mila Formation, near Sanghabad, Alborz Mountains

Fig. 1. Fragmentary free cheek, x8. It 13499.
Fic. 2. Holotype, partially exfoliated cranidium, x8. It 13503.

Fics 3, 4. Hypostoma attributed to this species, ventral view, and lateral view to show
anterior hump, x10. It 14020.

Hic. 5. Smallest cranidium, xi5. It 13504.

Fic. 7. Pygidium probably attributable to this species, fragmentary, x4. It 13460.
Tranoleesia pisiformis (King 1937) (p. 329)
Mila Formation, Ma’dan, Loc. 12 of King 1937

Fic. 6. Cast of holotype, incomplete cranidium, x5. Original specimen figured by King
1937: pl. 2, fig. 6a, GSI 16305 (cast In 36890).

Fia. 8. Two cranidia, one retaining part of the cuticle and showing punctate surface sculpture,
x6. In 36910.

Fies 9, 12. Internal mould of incomplete exfoliated cranidium, dorsal and lateral views, x Io.
Note punctation on internal surface. In 36912.

Fic. 10. Internal mould of cranidium, x5. In 36911.
Anomocarella falconi (King 1937) ? (p. 331)
Mila Formation, Darreh Shu, Loc. 8 of King 1937

Fic. 11. Cast of incomplete pygidium, x6; original figured by King 1937: pl. 2, fig. 7b, and
there attributed to A. falconi, but possibly belonging to Chelidonocephalus alifyons (p. 332).
GSI 16307 (cast In 36892).

PLATE 9

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

PLATE to

Anomocarella falconi (King 1937) (p. 331; see also Pl. 9, fig 11)
Mila Formation, Darreh Shu, Loc. 8 of King 1937

Fic. 1. Latex impression taken from a cast of the holotype, external mould of cranidium, x 4.
Original of King 1937: fig. 7a, GSI 16306 (cast In 36801).

Fic. 2. Small well-preserved cranidium retaining exoskeleton, dorsal view, x10. In 36g09a.
Same rock fragment as cranidium, Fig. 4.

Fic. 3. Internal mould of large cranidium, x5. In 36908.

Fics 4, 5, 8. Cranidium with exoskeleton, incomplete over posterior part of glabella. Dorsal
and oblique lateral views, x10; Fig. 8, detail of right side to show fine granulation, x 20.
In 36909b. Same rock fragment as cranidium, Fig. 2.

Koldiniella mitella Sivov 1955 (p. 333)
Mila Formation, near Sanghabad, Alborz Mountains
Fics 6, 7. Well-preserved cranidium in palpebral and right lateral views, x8. It 14021.

Figs 9,10. Pygidium, dorsal and posterior views, x 10, showing terrace lines confined to border.
It 13477.

Fic. 15. Pygidium, relatively longer than previous specimen, x10. It 14022.
Fic. 16. Hypostome tentatively assigned to this species, ventral view, x10. It 13506.
Fics 13, 14. Small cranidium, possibly a small growth stage of Koldiniella nmutella but may be
referable to Parakoldinia (see text, p. 333), x10. It 13476.
Tsinania ? sp. (p. 334)
Mila Formation, near Sanghabad, Alborz Mountains

Fics 11, 12. Incomplete cranidium, exfoliated, anterior and approximate palpebral views,
One lit 13493:

Fics 17-19. Small cranidium, retaining exoskeleton, anterior and lateral views, x7; Fig. 19,
palpebral view, x1o. It 13458.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4 PLATE tro

IPILAN IOS ipa

Dorypyge sp. (p. 327)
Mila Formation, near Sanghabad, Alborz Mountains

Fics 1, 2. Pygidium retaining most of exoskeleton, dorsal and oblique lateral views, x6.
Note impression of part of doublure posteriorly and thin subsidiary spine below third ‘main’
marginal spine on left. It 13461.

Hadragnostus edax sp. nov. (p. 324)
Mila Formation, near Sanghabad, Alborz Mountains
Fic. 3. Cephalon, glabella broken, x10. It 13468.

Fias 4-6. Pygidium, tectonically shortened, dorsal, posterior and left lateral views, x Io.
Note terminal node (Fig. 5). It 13472.

Fics 7, 8,12. Holotype pygidium, dorsal, posterior and left lateral views, x 10. Note articu-
lating half-ring. It 13463.

Fics 9, 10. Cephalon cut by joints, dorsal and lateral views, x10. It 13483a, on same block
as Figs 14, 15.

Fic. 11. Cephalon, showing rounded glabellar rear, x10. It 14023.
Bie. 13: “Small pygidium;, «20,1 13470,

Fics 14, 15. Cephalon, glabella damaged, dorsal and posterior views, x15. Note faint node
at posterior end of glabella. It 13483b, on same block as Figs 9, Io.

Bull. By. Mus. nat. Hist. (Geol.) 27, 4

IPE NIE IS, aR
Peronopsis fallax (Linnarsson 1869) aff. minor (Brégger 1878) (p. 326)

Mila Formation, near Sanghabad, Alborz Mountains

Fics 1, 2. Large cephalon, latex cast of external mould, dorsal and lateral views, x 10. Note
narrow axial furrow. It 13480b, on same block as Fig. 14.

Fics 3, 4. Cephalon, slightly compressed laterally, exfoliated in region of axial furrow which
appears wide, dorsal and lateral views, x10. It 13480c.

Fics 5, 8. Pygidium retaining exoskeleton, slightly deformed obliquely, dorsal and lateral
views, X10. It 13485.

Fic. 6. Large pygidium, partly exfoliated, x10. It 13467.
Hie. 7. Pyridium, <10.) lt 14025:

Fics 9, 11. Cephalon, laterally compressed, exfoliated in region of axial furrow, lateral and
dorsal views, x 10. Note elongate glabellar node and preglabellar depression. It 14024.

Fic. 10. Small cephalon exfoliated in region of axial furrow, x1Io. Note preglabellar de-
pression. It 14026.

Fics 12, 13. Pygidium, longitudinally compressed and partly exfoliated, dorsal and lateral
views, X10. It 13469.

Fic. 14. Cephalon retaining exoskeleton, showing Ptychagnostus-like basal lobes, xio. It
13480d, on same block as Figs 1, 2.

Peronopsis fallax (Linnarsson 1869) cf. minor (Brogger 1878) (p. 327)
Mancetter Grits and Shales, Merevale No. 3. Borehole, near Nuneaton, Warwickshire, England

Fic. 15. Damaged internal mould, x10. Institute of Geological Sciences, BDA 1781.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 4

INDEX
to Vol. 27 No. 4 (Miscellanea)

New taxonomic names and the page numbers of the principal references are printed in bold

type. An asterisk (*) denotes a figure.

Abhar 322
Acrospiriferinae 276
Acuminothyris 275, 282
Africa 309-12
Agnostidae, Agnostinae 322-5
Agnostoglossa simplexiformis 327
Agnostus pisiformis Zone 323, 327
simplexiformis 327
Albian 301
Alborsella Zone 322
Alborz Mountains 321, 324, 338
algae 285
Alispirifer 275-6
Alokistocaridae 335
America 312, 319
Amesopleura 275
novamexicana 278
Androstrobus 291
Anomocarella 322, 324, 330-1, 332
chinensis 330, 332
concava 332
falconi 321, 328, 330, 331-2, 336; pl. 9,
fie 11; pl. to, figs 1-5, 8
Anomocarellidae 331
Anomozamites 291
Ardennes—London landmass 285-6
Arizona 301
Asaphiscidae 324, 328-34
Auritamidae 335
Australia 311-2, 319, 337; see Mindyallan
Stage
Avon 285

Baltagnostus 324

Bannister, Dr K. 2098

Barbus 298

Baringo, Lake 295-6

Bath 285

Bathampton Down 285-6

Bathonian 285-6

Battus integer 326

Beast Cliff 291

Belgium 282

Bennettitaleans 291

Bisbee Quadrangle 301

Blastochaetetes bathonicus 285-6, 287-8;
pl. 2

Blountia 332

Bolland 278, 281-2

27

border furrows, true and false 337
brachiopods 275-84, 286, 301-4, 321
Brachyrhyncha 296-7
Brachythyrina 276, 282

Brunton, C. H.C. 275-84

bryozoa 285-6

California 301-4, 311
Cambrian 321-40
Carboniferous, Lower 275-84
Caspian Sea 321
Caytonia 201
Celsifornix 275
Chaetetida, Chaetetids 285-8
Chal-i-Sheh 322, 336
Chelidonocephalus 321-33, 334-6, 337-40
alifrons 321-3, 331-2, 335, 336-7, 338;
pl. 8, figs 1-5; pl. 9, fig. 11
preannulatus 321, 332-3, 336-7,
338-9 ; pl. 8, figs 6-8; pl. 9, figs 1-5, 7
SP. 322-3
Chile 319
Cirencester 285-6
Clitheroe 278
Cnidaria 285
Conocoryphidae 324
coprolites 289-94, 290* ; pl. 3, fig. 4
coral 286
Cotswolds 285-6
crabs, freshwater 295-300
Cracow 303
Crepicephalus Zone 323
Cretaceous 301-4
Croiset collection 308
Cyclothyridinae 302-3
Cyclothyris 301-4
americana 301
antidichotoma 303; pl. 5, figs 2-3
densleonis 301, 302*, 302-3; pl. 5, figs 1,
4
latissima 303
mirabilis 303
subtrigonalis 301
Cyrtia 276
Czekanowshia 291

Darreh Shu 322, 331
Davidson collection 277-8
Davies, Dr R.G. 321-2

342 INDEX

Deltaic, Lower 289 Grandioculus falconi 330
Delthyrididina 276 megalurus 330
Devonian, Upper 282 Gristhorpe Bed 291
Dictyella 324
Dictyites 324 Haanakdar 322
Dorypyge 322-3, 327, 333, 338; Zone 322-3 Hadragnostus 324
aenigma 333 edax 321, 323, 324-5; pl. 11, figs 3-15
ivanensis 328 las 323-5
khademi khademi 328 modestus 323-5
papillosa 328 Harrison, C. J.O. 305-20
vichthofent 327 Hartshilia 324
laevis 328 herbivores 289
laiwuensis 328 Hitt, C. R. 289-94
n. subsp. aff. vichthofeni 327 Hinde collection 285
Sp. 321, 323, 327-8; pl. 11, figs, 1-2 Holloway, Miss M. L. 312, 319
Dorypygidae 327-8 Horsetown Group 301
Dovedale 277, 279-80 Hundwarella 328
Drepanura Zone 322-3 Hydrozoa 285
(Spinopanura) 322 Hypermecaspis 328
Dublin 282

Illaenidae 333
Illaenuridae 333

echinoderms 286
Illaenurus 333

Eleutherohomma 275-6
ELiiotT, G. F. 285-8 CANEUs 334
England 301, 303 Iniotoma 328
evetes Zone 323 Inouyella 335

Europe 301 ; petensis 335
lavi insects 291

Eymekops rvectangularis 335 in io ae
Ivania 328

falconi 328, 330-1

pisiformis 328-9
AeainEoe QOS Ivanoleesia 322, 324, 328-9, 330; Zone
Florida 311 oe4
Forest Marble 286 falconi 321, 328, 330
Fortey, R. A. 321-40 pisiformis 321-2, 328, 329-30; pl. 9,
France 285, 303, 305, 307, 309 figs 6, 8-10, 12
Fusella 275, 276-8, 279-84 Jell, Dr P.

convoluta 281-2

fusiformis 275-7, 278-82; pl. 1, figs 5-

Faringdon 303
Fermanagh 278-81
fish 291, 298

Jurassic 285-94

I2, 20-21 Kabarsero 295
llanoensis 278 Kalimale 295-6
vhomboidea 275, 277-8, 280*, 281*, 281- Kamwina 295-6, 298
2; pl. 1, figs 13-19 Kaolishania Zone 322
voemerianus 282 Kashmir 328
tornacensis 277 Keates, C. 319
tviangularis 282 Kenya 295-300, 305, 309, 311
trigonalis 281-2 Kingstonia 333
Koldinia 324
Germany 309 Koldiniella 322-3, 332-3, 338
Gervaisia 307-8, 308-9 convexa 333
croiseti 305, 306*, 307*, 308, 310 mitella 321-3, 329, 332, 333-4; pl. 10,
Gilbertson collection 278, 282 ’ figs 6-7, 9-10, 13-16

Gloucestershire 285 prolixa 333

INDEX 343

Kootenia 328 ostracods 295
Kulyumbe 323 OwEN, E. F. 301-4
laevigata Zone 323 Paeckelmanella 276, 282
Peakey, Dr L/S: Be 311, 317 Paeckelmanellidae 276
Leighton Buzzard 303 Papyriaspididae 335
Leistegiidae 335 Parakoldinia 333; pl. 10, figs 13-14
Lejopyge laevigata Zone 323 plana 334
Licharewiinae 276 ? Spio S21
Limekiln Pit 285 Paris Basin 285
Liopeishamia 333 Pedinocephalina — Toxotis (?) Zone 323

lubrica 333 Peichishania 335

spannensis 333 vectangularis 335
Lirrangopotamonautes see Potamonautes Peishania 333
Llanoaspis walcottt 335 lubrica 333

Pelecani 315

Macelloura 333 Pelecanidae 315
Ma’dan 322, 329 Pelecaniformes 315
mammal 291-2 Pelecanus 315-19
Mapan Formation 332 aethiopicus 315-19, 316*; pl. 7
Mapania 331 conspicillatus 318-19
Mapaniidae 335 crispus 316*, 318-19
Maryvillia 332 erythrovhynchus 318-19
Meneviella 324 gracilis 319
Mexico 301, 312 grandiceps 319
Mila Formation 321, 323-4, 331, 338 halieus 319

zones 322 intermedius 319
Mila Kuh 322 occidentalis 318-19
Mindyallan Stage 323, 328, 337 thagus 319
Miocene 295-300, 305-14, 319 odessanus 319
Mirifusella 277 onocrotalus 316*, 318-19
molluscs 286 proavus 319
Montana 323 vufescens 316*, 318
Morris, S. F. 295-300, 321 pelican 315-20
Mucrospiriferidae 275-82 Pérignat 308
Mural Limestone 301 Permian 282

Peronopsis 324, 326
Nailsworth 285 fallax ferox 326
Nairobi 296 aff. minor 321, 323, 326-7; pl. 12,
Namanoia 335 figs I-14
Namanoiidae 335 cf. minor 323, 326; pl. 12, fig. 15
New Mexico 278 integer 3206
New South Wales 278 Peru 319
Ngorora Formation 295-300 Phoeniconaias 305, 308-11
Nilssoniopteris vittata 291 floridanus 311-12
Nuneaton 323 gracilis 311
minor 305, 300*, 307*, 309, 309*, 311

Olduvai 315-20 novaehollandiae 311
Oligocene 305-14 stockt 312
oncolites 285 Phoeniconotius eyrensis 312
Ordosia 335 Phoenicoparrus 305, 308, 310

fimbricauda 335 andinus 305
ornithopod dinosaur 291 jamesit 305, 306*, 307*, 309*

Orthopotamonautes see Potamonautes Phoenicopteridae 310

344

Phoenicopterus 305, 308, 310

aethiopicus 305, 309*, 310*, 311-12;

pl. 6
copet 312
croiseti 305, 307-8
minutus 311
yvubey 305, 307*, 308-11
antiquorum 305
chilensis 305, 306*, 307*, 309*, 310*
yubery 305
Pickford, Dr M. 295
plant cuticles 289-92
Pleistocene 311-2, 315-20
Plethometopus convergens 334
obtusus 334
Plethopelioides lepidus 334
Pliocene 311-2, 319
Poshania 335
poshanensis 335
Poshauia 335
Potamidae 296-7
Potamon (Potamonautes) livrangensis 296
Potamonautes 296, 297-8
livrangensis 296
perlata 296
(Livvangopotamonautes) 295, 296, 297
livvangensis 296, 298
tugenensis 295, 296-7, 298; pl. 4
(Orthopotamonautes) 295, 297
depressus 298
johnstoni 298
Prochuangia Zone 322
Protohedinia 329
Pierophyllum 291
Pterospivifer 275-6, 282
Ptilophyllum 289-94
pectinoides 290*, 291; pl. 3
Ptychagnostus 327
atavus 326
ptychoparioids 321, 323
Puy-de-Déme 307
pygostyle 297

Qanli-Chapoghlu 322
Qazvin 321
Quadragnostidae 326-7
quasivespa Zone 323
Queensland 323

rabbits 289, 291
reptiles 291
Rhynchonella 301
densleonis 301-2
Rhynchonellidae 301-3

INDEX

Rhyssometopacea 335-7
Rhyssometopus princeps 337

Rift Valley 298

RissonE, A. 275-84

Roseberry Topping plant bed 289-92
Rusuton, A. W. A. 321-40

Rusinga Island 305, 309, 311

Salair 323
Sanghabad 321-2, 324, 338
Saukia Zone 322
Scandinavia 323
Schulze-Ammonia maceration 291
Sclerospongiae 286
Shahmirzad 322
Shasta County 301
sheep 291-2
Shiraz 321
Siberia 323
Solenites 291
Spinopanura see Drepanura
Spirifer 276
bicavinata 276, 282
convoluta 282
cuspidatus 281-2
distans 282; pl. 1, figs 1-4
fustformis 276, 278
mundula 278
novamexicana 278
vhomboidea 277
vhomboides 276
voemerianus 282
tornacensis 277
triangularis 282
Spiriferacea 275-82
Spiriferidae 275-6
Spiriferidina 276
Spirifinaella 276
sponges 286
stromatoporoids 285
Strophopleura 275, 282
antevosa 278
Strophopleurinae 275-6, 277-82
Syringothyridacea 276
Syringothyrididae 277
Syringothyris 281-2
cuspidatus 281-2

tabulates 285
Taitzma 335

imsueta 335
Taleghan Range 324, 338
Tang-i-Tehbud Bezuft 322
Tanout 298

INDEX 345

Tanzania 315, 317 Unispivifer 277-8
Tehran 321
Tengfengiidae 329 Victoria, Lake 309, 311
Tennessee 335 Viséan 279-82
Terebratulidae 301 Voiseyella 275-6, 278, 282
Thelphusa perlata 296 anterosa 278
Thomas, DrH. H. 289 novamexicana 278
Tiltup’s End 285-6
Tomes collection 285 WALKER, C. A. 305-20
Tournaisian 282 Wanwanoglobus convexus 334
Toxotis see Pedinocephalina Williamsonia 292
Treak Cliff 281-2 hildae 291
trilobites 321-40 leckenbyt 291

effaced 324, 332-3 Witchell collection 285
Tsinania 324, 334

caneus 334 Xanthoidea 296-7

?sp. 321, 323, 334; pl. 10, figs 11-12,

17-19 Yorkshire 289-94

Tsinaniidae 324, 333
Tugen Hills 295-6 Zagros Thrust 321

2 he

0

re oh ae ;

ea

by
i

pit Ea

Met Na

IO.

nn

A LIST OF SUPPLEMENTS
TO THE GEOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya.

Pp. 213; 30 Plates; 2 Text-figures. 1965. OUT OF PRINT.

. Ext-Naccar, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper

Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley,
Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. {11.

. Davey, R. J., Downie, C., SARJEANT, W. A. S. & WittiAMs, G.L. Studies on

Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 Plates; 64 Text-
figures. 1966. £8.20.

. APPENDIX. DAVEY, R. J., Downie, C., SARJEANT, W. A. S. & WILLIAMS, G, hy ®

Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 24.
1969. 95P.

. Etuiott, G. F. Permian to Palaeocene Calcareous Algae (Dasycaaaeenes ot

the Middle East. Pp. 111; 24 Plates; 16 Text-figures. 1968. OUT OF PRINT.

RHODES, ESB Austin. Rot. & DRUCE, E. C. British Avonian (Carboni-

ferous) Conodont faunas, and their value in local and continental correlation.
Pp. 313; 31 Plates; 92 Text-figures. 1969. £13.10.

. Cuttps, A. Upper Jurassic Rhynchonellid Brachiopods from Northwestern "

Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £5.25.

i GOODY Pi Gerais relationships of certain Upper Cretaceous Teleosts with

special reference to the Myctophoids. Pp. 255; 102 Text-figures. 19609.
£7.70.

. Owen, H. G. Middle Albian Stratigraphy in the Anglo-Paris Basin. Pp.

164; 3 Plates; 52 Text-figures. 1971. {7.20.

. SIDDIQUI, Q. A. Early Tertiary Ostracoda of the family Trichylebeaaeaae

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. £9.60.

Forey, P.L. A revision of the elopiform ‘fishes, fossil and Recent, Pp. 222;

92 Text-figures. 1973. 11.35. ng
Wit.iams, A. Ordovician Brachiopoda from the Shelve District, Shrapebien|
Pp. 163; 28 Plates; 11 Text-figures; 110 Tables. 1974. £12.80.

Printed in Great Britain by Ffohn Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 sNU

CRETACEOUS FAUNAS FROM...)
ZULULAND AND NATAL,
SOUTH AFRICA

THE AMMONITE FAMILY
PHYLLOCERATIDAE

W. J. KENNEDY

AND

H. C. KLINGER

+. .
‘
5
a

a

BULLETIN OF
e PTHE BRITISH MUSEUM (NATURAL HISTORY)
4 GEOLOGY Vol. 27 No. §
e LONDON : 1977

CRETACEOUS FAUNAS FROM ZULULAND" ss

AND NATAL, SOUTH AFRICA a
THE AMMONTTE FAMILY PHYLLOCERATIDAE

BY

WILLIAM JAMES KENNEDY
University of Oxford
AND

HERBERT CHRISTIAN KLINGER

South African Museum, Cape Town

Pp. 347-380 ; 15 Plates; 9 Text-figures

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
GEOLOGY Vol. 27, No: 5
LONDON: 1977

THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), ‘mstituted 1m 1949, 15
issued im five series corresponding to the Scientific
Departments of the Museum, and an Historical serves.

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

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

This paper is Vol. 27, No. 5 of the Geological
(Palaeontological) series. The abbreviated titles of
periodicals cited follow those of the World List of
Scientific Periodicals.

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

ISSN 0007-1471

© Trustees of the British Museum (Natural History), 1977

BRITISH MUSEUM (NATURAL HISTORY)

Issued 27 January 1977 | Price {6-10

CRERACEOUS: FAUNAS: FROM, ZUEULAND
AND NATAL, SOUTH AFRICA

THE AMMONITE FAMILY PHYLLOCERATIDAE

By W. |. KENNEDY & H.C. KEINGER

CONTENTS
Page
ABSTRACT : : : é : : - : ; ; 349
I. INTRODUCTION : : : ; : : 6 : ; 350
Location of Specimens . é ; : : : ; ; 350
Field Localities . : : : ‘ : : F : 351
Dimensions of Specimens. : : 5 5 : 5 351
II. SYSTEMATIC PALAEONTOLOGY . ; : : : : : 351
Genus Phylloceras : : ; : : : : : 351
Subgenus Phylloceras : : : - ; : 352
Phylloceras (Phyllocevas) serum . : 2 ‘ : 352
Subgenus Hypophylloceras . - : ; : 354
Group of Phylloceras (H. Neer iiinccnas| Weis : : : 354
Phylloceras (Hypophylloceras) subalpinum ellipticum ‘ 355
Phylloceras (Hypophylloceras) aphrodite . : : 3 358
Group of Phylloceras (Hypophylloceras) velledae . j : 359
Phylloceras (Hypophylloceras) velledae . : 3 : 360
Phylloceras (Hypophylloceras) velledae velledae_ . ; 360
Phyllocevas (Hypophylloceras) velledae inflatum . : 362
Phylloceras (Hypophylloceras) sp. aff. velledae . : : 363
Group of Phylloceras (Hypophylloceras) sevesitense . : 363
Phylloceras (Hypophylloceras) seresitense ; ; 363
Phylloceras (Hypophylloceras) seresitense eee : 364
Phyllocevas (Hypophylloceras) androiavense . : ; 3605
Phylloceras (Hypophylloceras) woodsi woodsi : : 366
Phylloceras (H. ei mikobohense . , : 368
Genus Partschiceras . : ; , ‘ ; 369
Partschiceras ree : : 5 : : : 370
Parschiceras sp. juv. ? cf. baborense é 2 ; : Bal
Genus Carinophylloceras : ‘ , é ‘ : ‘ 272
Carinophylloceras collignoni . ; . é ‘ : 373
III. ACKNOWLEDGEMENTS : : : : ; : : : 374
IV. REFERENCES . ; : ; 3 : : : : : 374
Wer -inpExe © ‘ ; ‘ : ‘ f : : : : 378
VI. Erratum ‘ ; ; ? ‘ ; é : ‘ : 380
ABSTRACT

Phylloceratid ammonites are present in rocks of Barremian, Albian, Cenomanian, Santonian,
Campanian and Maastrichtian ages in Zululand and Natal, South Africa. They form a sub-
stantial proportion of the normally coiled ammonite fauna during the Barremian and Albian,
but are rare in Aptian and post-Albian faunas. The following forms are described below.

350 CRETACEOUS FAUNAS

Phylloceras (Phylloceras) serum (Oppel), P. (Hypophylloceras) subalpinum (d’Orbigny)
ellipticum Kossmat, P. (Hypophylloceras) aphrodite Fallot & Termier, P. (Hypophylloceras)
velledae velledae (Michelin), P. (Hypophylloceras) velledae (Michelin) inflatum Collignon,
P. (Hypophylloceras) sp. aft. velledae (Michelin), P. (Hypophylloceras) seresitense seresitense
Pervinquiére, P. (Hypophylloceras) androiavense Besairie, P. (Hypophylloceras) woodsi woodsi
van Hoepen, P. (Hypophylloceras) mikobokense Collignon, Partschiceras umzambiense (van
Hoepen), Partschiceras sp. juv. cf. baborense (Coquand) and Carinophylloceras collignoni Klinger,
Wiedmann & Kennedy.

Eight of these species are new records from the area.

I. INTRODUCTION

THE Phylloceratidae are the root stock of the Jurassic and Cretaceous phyllo-
ceratids, appearing in the Lower Jurassic (Sinemurian) and ranging to the Upper
Cretaceous (Lower Maastrichtian). They have a world-wide distribution, but for
most of the Cretaceous they are typically rare in the boreal region of north-western
Europe and in the western interior of North America. In South Africa, species
are known from the Barremian to Lower Maastrichtian, and are relatively common
in the Barremian and Albian.
The following species are described below :

Phylloceras (Phylloceras) serum (Oppel)

. (Hypophylloceras) subalpinum (d’Orbigny) ellipticum Kossmat
. (Hypophylloceras) aphrodite Fallot & Termier

. (Hypophylloceras) velledae velledae (Michelin)

. (Hypophylloceras) velledae (Michelin) inflatwm Collignon

. (Hypophylloceras) sp. aff. velledae (Michelin)

. (Hypophylloceras) seresitense seresitense Pervinquiere

. (Hypophylloceras) androiavense Besairie

. (Hypophylloceras) woodst woodst van Hoepen

. (Hypophylloceras) mikobokense Collignon

Partschiceras umzambiense (van Hoepen)

Partschiceras sp. juv. cf. baborense (Coquand)
Carinophylloceras collignom Klinger, Wiedmann & Kennedy

ng) mesh Hel neh nel nch acl clans)

Location of Specimens

The following abbreviations are used to indicate the repositories of the material
studied :

BM(NH) _ British Museum (Natural History), London

EMP Ecole des Mines, Paris

MHNP Muséum d’Histoire Naturelle, Paris

SAM South African Museum, Cape Town

SAS South African Geological Survey, Pretoria
™ Transvaal Museum, Pretoria

UPE University of Pretoria, Pretoria

ZULULAND AND NATAL 351

Field Localities

Outline details of field localities referred to in this paper are given by Kennedy &
Klinger (1975) ; full descriptions of sections are deposited in the Palaeontology
Library of the British Museum (Natural History).

Dimensions of Specimens

All dimensions given below are in millimetres :

D = Diameter
Wb = Whorl breadth
Wh = Whorl height
U = Umbilicus

I ll

Figures in parentheses are dimensions as a percentage of the total diameter.

ik. SYSTEMATIC PALAEONTOLOGY

Class CEPHALOPODA Cuvier 1797
Subclass AMMONOIDEA Zittel 1884
Order PHYLLOCERATIDA Arkell 1950
Superfamily PHYLLOCERATACEAE Zittel 1884
Family PHYLLOCERATIDAE Zittel 1884
Subfamily PHYLLOCERATINAE Zittel 1884
Genus PHYLLOCERAS Suess 1865

TYPE SPECIES. Ammonites heterophyllus J. Sowerby 1820.

Discussion. An extensive and useful review of the scope and interpretation of
Phylloceras and its subgenera has been given by Wiedmann in a series of publications
(1962a, b, 1964; Wiedmann & Dieni 1968). He stresses the slight differences of
both species and genera within the phylloceratids, and regards many of the taxa
proposed by previous workers as of little value on either phylogenetic or strati-
graphic grounds. With minor reservations, his conclusions are accepted here, and
they clarify the relations between the bewildering host of taxa referred to the group.
Our following discussions thus rest heavily on his work, but are included here as a
review of work otherwise available only in German.

The genus Phylloceras is thus typified by the group of Phylloceras heterophyllum.
Rhacoceras Hyatt 1867 and Heterophylloceras Kovacs 1939 are both objective
synonyms, while the poorly defined Xeniophylloceras Buckman 1921, Hantkeniceras
Kovacs 1939 and Pseudophylloceras Beznosov 1958 do not bear separation on either
gross morphological features or sutural details. Furthermore, fasciculate ribbing
occurs in some Phylloceras s.s., suggesting that the Pseudophylloceratinae (Beznosov
1958) — genera such as Neophylloceras Shimizu 1934 and Epiphylloceras Collignon
1956 — do not bear separation from Phylloceras s.s. The following subgenera can
be recognized.

352 CRETACEOUS FAUNAS

Phylloceras Suess 1865, type species Ammonites heterophyllus J. Sowerby 1820,
Lower Jurassic to Lower Cretaceous.

Zetoceras Kovacs 1939, type species Ammonites zetes d’Orbigny 1850, Lower to
Middle and ? Upper Jurassic.

Geyeroceras Hyatt 1900, type species Ammonites cylindricus J. de C. Sowerby
1831, Lower Jurassic.

Hypophylloceras Salfeld 1924, type species Phylloceras onoense Stanton 1895,
Cretaceous.

Subgenus PHYLLOCERAS Suess 1865
Phylloceras (Phylloceras) serum (Oppel)
(Fig. 1; Pls 1-2; Biles, fics 12)

1865 Ammonites serus Oppel : 550.

1868 Phylloceras serum Oppel; Zittel: 66, pl. 7, figs 5, 6.

1885 Phylloceras semistriatum d’Orbigny ; Neumayr: 83, pl. 1, fig. 2 (non d’Orbigny).

1890 Phylloceras serum Oppel; Toucas: 574, pl. 13, fig. 2.

1891 Phylloceras cf. thetys d’Orbigny ; Sayn: 141, pl. 1, fig. 1.

1901 Phylloceras serum Oppel var. perlobata Sayn : 7, text-fig. 3, pl. 1, figs 6-8.

1901 Phylloceras thetys d’Orbigny ; Sayn: 36, pl. 1, fig. 3 only.

1907 Phylloceras serum Oppel; Pervinquiere: 12, pl. I, fig. 3.

1914 Phylloceras serum Oppel var. perlobata Sayn ; Zwierzycki : 33, pl. 4, figs 2, 3.

1949a Phylloceras spathi Collignon : 63, pl. 9, figs 1-1b.

1949a Phylloceras robionense Collignon : 64, 65, text-figs I-Ia.

1951 Phylloceras serum (Oppel) ; Arnould-Saget: 5, pl. 1, fig. 5.

1957 Hyporbulites betievi Busnardo; Busnardo & David: 84, pl. 1, figs 1-2; text-figs 4, 5.

1960 Euphylloceras serum Oppel; Drushchitza & Kudryavtseva : 251, pl. 1, fig. 8.

1960 Phylloceras serum Oppel; Collignon: pl. 134, figs 506, ? 507.

1962a Phylloceras serum Oppel; Collignon: 20, pl. 181, fig. 817; 33, pl. 186, fig. 853; 76,
pl. 214, fig. 916.

1962a Phylloceras spathi Collignon ; Collignon : 74, pl. 203, fig. 915; pl. 204, fig. 915a.

1966 Phylloceras cf. perlobatum Sayn ; Rawson : 455, pl. 72, figs 1-3.

1968 Ph. (Phylloceras) spathi Collignon ; Wiedmann & Dieni: 22.

1968 Ph. (Phylloceras) serum (Oppel) ; Wiedmann & Dieni: 20.

“wv

ww

on.

ww

LECTOTYPE. The specimen figured by von Zittel (1868) as his pl. 7, fig. 5 is
herein designated lectotype.

MATERIAL. Fourteen specimens, all from the Upper Barremian (Barremian I),
Loc. 170, Mlambongwenya Spruit, northern Zululand. BM(NH) C78652-61.
SAS LJE 132b, 54/19, 54/33, Z19.

DIMENSIONS. D Wb Wh Wb/Wh Ui
Zittel 1868 : 66 106-0 (28) (60) 0-46
Collignon 1960 :
pl. 134, fig. 566 96-0 35 (36) 57 (59) 0-61
Collignon 1962a :
pl. 204, fig. 916 1230 430 (35) 72:0 (59) 0-60
BM(NH) C78657 864 28-2 (33) 49°4 (57) O57 6-4 (7

)
BM(NH) C78660 120°5 44°5 (37) 70°8 (59) 0-62 8-6 (7)

ZULULAND AND NATAL 353

Fic. 1. External suture line of Phylloceras (Phylloceras) serum Oppel. SAS LJE 132b,
x 6.

Phylloceras spathi, from Collignon 19492 : 62.

Example I E126 46:0 (41) 63:0 (56) 0°73 6-0 (5)
Example 2 110-0 44:0 (40) 65:0 (59) 0-67 6-0 (5)
Example 3 113-0 40:0 (35) 66:0 (53) 0-60 6-0 (5)

Phylloceras robionense, from Collignon 1949a : 64.
Holotype T4t-o — 45°0 (32) 80 (57) 0°55 5°0 (4)

DESCRIPTION. The coiling is very involute with a tiny, deep umbilicus measuring
7%, of the diameter. The umbilicus is undercut, the umbilical wall rounded, and
the umbilical shoulder gently sloped, so that there is a shallow conical depression
surrounding a deep umbilical pit. The whorl section is a compressed oval, with
flattened subparallel lower flanks and upper flanks which converge towards broadly
rounded shoulders and a broad, rounded venter.

The shell surface is ornamented by fine, dense, flexuous striae. These arise at the
umbilical seam and are very fine, dense and crowded over the umbilical wall,
shoulder and lower flank, strengthening on the outer flank and venter. They pass
straight across the lower flank, flex slightly backwards some way below mid-flank,
to sweep gently forwards over the upper flank and shoulder. They pass across the
broad venter with a faint forwards projection. Internal moulds bear a fainter but
comparable ornament.

The suture line is complex, phylloid, the saddles having tetraphylloid terminations,
as shown in Fig. 1.

Discussion. Flattened sides, broad venter, flexuous ornament and tetraphylloid
saddles clearly place our specimens within the range of variation ascribed to
Phylloceras (Phylloceras) serum by Collignon (1962), Arnould-Saget (1951) and
Wiedmann & Dieni (1968), whom we follow in regarding Sayn’s variety perlobata

354 CRETACEOUS FAUNAS

and the ‘Hyporbulites’ betiert of Busnardo as synonyms. Phylloceras (Phylloceras)
spathi Collignon is very close to P. (P.) serum in ornament, and overlaps in relative
proportions, the large specimen figured by Collignon (1962a : pl. 203, fig. 915) having
a similar suture, which suggests that this species may also fall into synonymy.
Wiedmann & Dieni (1968 : 22), however, state that P. (P.) shathi lacks ornament on
internal moulds and differs from P. (P.) seywm mainly in whorl shape. Phylloceras
vobionense, from the Barremian of southern France (Basses-Alpes), may also be a
synonym.

P. (P.) serum has, on occasion, been confused with Phylloceras of the thetys group,
with which it is homoeomorphic; these forms have, however, diphylloid saddle
terminations as do forms such as Phylloceras sablyense Karakash.

OccURRENCE. Phylloceras (Phylloceras) serum has a long time range and a wide
geographic distribution, which may be summarized as follows :

Upper Jurassic (Tithonian) of central Europe (Stramberg Limestone), southern
France, north Africa (Tunisia) and Madagascar; Lower Cretaceous of southern
France (Valanginian and ? Barremian), Sardinia (Valanginian), Madagascar
(Hauterivian and Valanginian) and Zululand (Barremian).

Subgenus HYPOPHYLLOCERAS Salfeld 1924

TYPE SPECIES. Phylloceras onoense Stanton 1895.

DIAGNOsIS. Compressed, discoidal involute phylloceratids, inner whorls occasion-
ally constricted (‘Aphroditiceras’, ‘Euphylloceras’, ‘Goretophylloceras’). Ornament
of both shell and moulds consists of weak radial striae, strengthening on the venter ;
occasionally fasciculate in the umbilical region (‘Neophylloceras’, ‘Epiphylloceras’).
Suture with diphylloid (‘Goretophylloceras’), tetraphylloid (‘Hyporbulites’) or poly-
phylloid saddles (from Wiedmann 1964 : 173).

Discussion. Following Wiedmann, we regard Neophylloceras Shimizu 1934,
Paraphylloceras Shimizu 1935, Hyporbulites Breistroffer 1947, Goretophylloceras
Collignon 1949a, Euphylloceras Drushchitz 1953 and Aphroditiceras Mahmoud 1956
as synonyms of Hypophylloceras. The group thus represents a variable cluster of
phylloceratids linked by their apparent common origin in Phylloceras of the serum
group during the early Cretaceous ; detailed study shows that there is every inter-
mediate between the supposedly diagnostic morphologies of the various genera
placed in synonymy. Furthermore, the intimate gradational relationship between
P. (P.) serum and some Hypfophylloceras species suggests that even subgeneric
separation is more of a stratigraphic convenience than of phylogenetic value.

OccURRENCE. Hypophylloceras has a world-wide distribution, although rare in
the boreal region and the western interior of North America. Species first appear
in the Valanginian, and range to the Lower Maastrichtian.

1. Group of P. (Hypophylloceras) thetys (d’Orbigny)

This species group is characterized by di- or triphylloid saddles throughout
ontogeny. It is thus separable from the velledae group where saddles are at first

ZULULAND AND NATAL 355

diphylloid, becoming tetraphylloid when adult, and the sevesitense group where
saddles are tetraphylloid throughout ontogeny. The following are referred to the

group :

1G

Hypophylloceras thetys (d’Orbigny) (1841: 174; pl. 53, figs 7-9), Valanginian
to Barremian.

. Hypophylloceras thetys (d’Orbigny) diegot Boule, Lemoine & Thévenin (1906 : ro ;

pl. 1, figs 5, 7; text-fig. 4), Barremian to Cenomanian.

. Hypophylloceras thetys (d’Orbigny) majoricense Wiedmann (1964:178; pl. 15,

fig. 1; pl. 18, fig. 3; text-fig. 37), Aptian to Albian.

. Hypophylloceras sablyense Karakash (1907: 39; pl. 3, figs 12a-b; pl. 14,

figs 8a—b; pl. 24, fig. 4), Barremian.

. Hypophylloceras fortune: (Honnorat-Bastide 1892) (Wiedmann 1964: 192),

Aptian.

. Hypophylloceras aptiense aptiense Sayn (1920:195; pl. 1, fig. 4; text-fig. 2),

Aptian.

. Hypophylloceras aptiense Sayn applanatum Wiedmann (1964:190; pl. 15,

fig. 5; text-fig. 43), Aptian.

. Hypophylloceras morett Mahmoud (1955:76; pl. 5, figs 2-4; text-fig, 44),

Albian.

. Hypophylloceras aphrodite Fallot & Termier (1923 : 25; pl. 2, figs 5-7; text-fig.

5), Albian.

. Hypophylloceras subalpinum subalpinum d’Orbigny (1841: 283; pl. 83, figs

1-3), Albian.

. Hypophylloceras subalpinum dW Orbigny ellipticum Kossmat (1895: 11; pl. I,

fig. 2; pl. 6, fig. 1), ? Aptian, Albian to Cenomanian.

. Hypophylloceras wmprovisum (Stoliczka) (1865: 113; pl. 58, figs 4-4b), Ceno-

manian.

. Hypophylloceras subseresitense Wiedmann (1964: 202; pl. 14, fig. 9; pl. 15,

fig. 6; text-fig. 47), Aptian.

. Hypophylloceras sardoum Wiedmann & Dieni (1968: 24; pl. 1, figs 1, 2; pl. 2,

fig. 2; text-fig. 3), Albian.

Phylloceras (Hypophylloceras) subalpinum (d’Orbigny) ellipticum Kossmat

(Figs 2-3; Pl. 4, figs 1-4; Pl. 5, figs 2-3; Pl. 12, fig. 2)

1865 Ammonites subalpinus d’Orbigny ; Stoliczka: 114; pl. 58, figs 3a—c (non d’Orbigny).
?1877. Ammonites velledae Michelin ; Simonovich, Batsevich & Sorokin: pl. 3, fig. 3 (non
Michelin).
1895 Phylloceras ellipticum Kossmat:107; pl. 15, figs 2a-b; pl. 20, figs 1a-b.

non 1897 Phylloceras ellipticum Kossmat ; Parona & Bonarelli: 78; pl. to, fig. 7 [= Phyllo-

cevas aphrodite Fallot & Termier].

1907a Phylloceras cf. ellipticum Kossmat ; Crick: 169; pl. Io, figs 12-12a.

1907 Phylloceras ellipticum ? Kossmat; Pervinquiére: 51; pl. 3, figs 1, 2a-b; text-
fig. 4.

1910 _ Phylloceras ellipticum Kossmat ; Pervinquiere : 1o.

1928 Phylloceras cf. ellipticum Kossmat ; Collignon: 147; pl. 1, figs 10o-Ioa.

356 CRETACEOUS FAUNAS
U,
<a a erL lee L
E
Fic. 2. External suture line of Phylloceras (Hypophylloceras) subalpinum ellipticum
Kossmat. SAS Ar1i16, x6.
1943 Phylloceras japonicum Matsumoto : 674 (nom. nud.).
1953 Phylloceras japonicum Matsumoto: 147; text-fig. 10 (nom. nud.).
non 1956 Phylloceras cf. ellipticum Kossmat; Collignon:15; pl. 1, figs I, Ia, Ib
[= Phylloceras improvisum (Stoliczka) ].
1959 Partschiceras ? japonicum (Matsumoto) ; Matsumoto: 52; pl. 12, figs Ia—c, 2a-b,
3a-c; pl. 28, figs 2a-d; pl. 29, figs 1a-b; text-fig. 2.
1964 Ph. (H.) subalpinum ellipticum Kossmat; Wiedmann:197; pl. 13, figs 5a-b;
pl. 14, fig. 8; pl. 15, figs 3a-c.
1968 Ph. (Hypophylloceras) subalpinum ellipticum Kossmat; Wiedmann & Dieni: 23 ;
pl. 1, fig. 4; pl. 2, figs 1a-b.
HoLotyPE. Kossmat’s original specimen from the Utatur Group of southern

India (1895: pl. 15, figs 2a—b; pl. 20, fig. 1).

MATERIAL. We have fourteen specimens: BM(NH) C18139 (figured by Crick
1907a) probably from the Cenomanian of the Skoenberg region, BM(NH) C78662
from Bed 6, Loc. 51, BM(NH) C78664 from Loc. 54, SAS 2477 and Z38 from Loc. 57,
all from the Mzinene River (Albian V); BM(NH) C78663 from Bed 4, Loc. 64,
BM(NH) C78665, SAS 945, 66/P1 from Loc. 66, Munywana Creek (Albian V),
BM(NH) C78666, SAS A4g9, 10822 and Ag45 from Loc. 62, the Skoenberg, Mzinene
River (Cenomanian II-III) and SAS Ar116 from the Albian of the Mzinene (precise
horizon unknown). BM(NH) C78686 from Bed 4, Loc. 64, on the Munywana Creek,
may also belong here.

DIMENSIONS. D Wb Wh Wb/Wh U
Holotype (from

Kossmat 1895) 710 28°5 (40) 41:0 (58) 0-70 50 (7)
BM(NH) 78662 45:0 «178 (40) 253 (56) o-71 3°3 (7)
BM(NH) C78669 585 20:0 (34) 35:2 (60) 057 3°6 (6)
SAS 10822 565 17 (Go) ar 55) 0°55 5°0 (9)
SAS 2447 59°5 20). (BAvor, 465 1(59) 0°57 4:0 (7)
SAS 66/Pr 56:0 L7 5st) 2r) (55) 0°56 4°5 (8)

ZULULAND AND NATAL 357

DESCRIPTION. The coiling is very involute, with a tiny, funnel-shaped umbilicus
(6-9% of diameter). The whorl section is a compressed oval, the whorl breadth
to height ratio varying from 0-55 to 0-71, the greatest breadth being typically at
or just below mid-flank. The umbilical shoulder slopes gently outwards, producing
a funnel-shaped umbilical depression, the flanks being gently inflated, converging
to a narrow rounded venter. The surface of the shell is ornamented by low, flat,
radial folds which arise at the umbilical seam as mere striae, but strengthen across
the lower flank, flexing faintly forwards as they do so, to flex faintly back at mid-
flank and pass almost straight across the upper flank and venter. The internal
mould is smooth, or bears only faint, low folds.

The suture is moderately subdivided with triphylloid saddles (Figs 2-3).

Discussion. Ornament and whorl section clearly place our specimens close to
P. (Hypophylloceras) subalpinum ellipticum ; specimens figured as Pl. 5, figs 3a—c
and Pl. 6, figs 2a—b have the identical weak fold-like ribs and elliptical whorl section
clearly shown in Kossmat’s figures and in material illustrated recently by Wiedmann
(1964). Other specimens, more compressed than previously-described material
referred to this species (whorl breadth to height ratio as low as 0:55) but otherwise
identical, we also place within the range of swbalpinum ellipticum. None of our
specimens show the constrictions described by Kossmat, nor the fine striations
developed on well-preserved material (e.g. Wiedmann 1964: pl. 15, fig. 3c). The
dimensions of ellipticum overlap those of P. (H.) subalpinum subalpinum, and
involution and sculpture are identical. These similarities led Stoliczka to refer
his Indian material to P. (H.) subalpinum s.s. Kossmat, however, separated the
Indian material on the basis of the elliptical whorl section with the greatest breadth
at or about mid-flank, and the presence of constrictions and fine striations in
ellipticum, as opposed to the trigonal whorl section in subalpinum. As Wiedmann
(1964 : 198) has noted, no body chambers of subalpinum have been described, so
that the constrictions and fine striations cannot be used to separate these forms.
Whorl section remains as one distinguishing character, while the suture lines of the

U,
0 see ee

)

2

Fic. 3. External suture line of Phylloceras (Hypophylloceras) subalpinum ellipticum
Kossmat. SAS Z447, x 3.

358 CRETACEOUS FAUNAS

present material show that virtually all the saddles are subtriphylloid, thus
substantiating Wiedmann’s (1964) suggestion that the greatest difference between
the two subspecies is in the configuration of the saddles, especially L/U,.

‘Partschiceras’ japonicum Matsumoto, from the Cenomanian of Japan and Alaska,
is a clear synonym of P. subalpinum ellipticum, whilst the forms compared with this
subspecies by Pervinquiére (1907, 1910), Crick (1907a) and Collignon (1928) clearly
belong to it. The specimen of Phylloceras cf. ellipticum described by Collignon in
1956 from the Turonian of Madagascar appears to be a P. (Hypophylloceras)
improvisum Stoliczka; the Phylloceras elpticum of Parona & Bonarelli (1897)
from the Albian of Escragnolles should be referred to P. (Hypophylloceras) aphrodite.

P. (H.) subalpinum ellipticum is a distinctive form. Whorl section alone readily
separates it from P. (H.) subalpinum subalpinum, as noted above, whilst forms such
as P. (H.) fortunes and P. (H.) sardoum have distinctly trigonal whorl sections.
P. (H.) improvisum is a form with rounded whorls and a broad venter, and P. (H.)
aptiense aptiense and applanatum and P. (H.) aphrodite all bear constrictions
throughout ontogeny. P. (Hypophylloceras) thetys thetys, thetys diegot and thetys
majoricense, P. (H.) moreti and P. (H.) subseresitense are as readily distinguished,
being ornamented by fine striae rather than the low folds of P. (H.) subalpinum
ellipticum, and there are other obvious differences (Wiedmann 1964; Wiedmann &
Dieni 1968).

OccURRENCE. Phylloceras (Hypophylloceras) subalpinum ellipticum is known
from the Lower Albian of the Balearics, the Upper Albian of Sardinia, Tunisia,
southern Russia and South Africa (Zululand), and the Cenomanian of Algeria,
southern India, Madagascar, South Africa (Zululand), Alaska and Japan.

Phylloceras (Hypophylloceras) aphrodite Fallot & Termier
(Pl3, fig. 3)

1897 Phylloceras ellipticum Kossmat ; Parona & Bonarelli: 78; pl. 10, fig. 7.

1923 Phyllocevas aphrodite Fallot & Termier: 25; pl. 2, figs 5-7; text-fig. 5.

1947 Phylloceras escragnollensis Breistroffer : 55.

1955 Salfeldiella (A phroditiceras) aphrodite var. venus Mahmoud : 75; pl. 5, fig. 1; text-fig. 43.

1957 LPhylloceras aphrodite Fallot & Termier; Almela & Revilla: 19; pl. 3, fig. 5.

1957 Phylloceras (Salfeldiella) guettardi Raspail; Almela & Revilla: 16; pl. 2, fig. 6.

1957 Phylloceras aff. grothti Fallot; Almela & Revilla: 16; pl. 2, fig. 7.

1962a Hypophylloceras aphrodite (Fallot & Termier) ; Wiedmann: 252; pl. 16, fig. 5; text-
fig. 5.

1962b Hypophylloceras aphrodite (Fallot & Termier) ; Wiedmann: 141.

1964 Ph. (Hypophylloceras) aphrodite Fallot & Termier ; Wiedmann: 184; pl. 18, figs 4-6;
pl. 20, fig. 1; text-fig. 4o.

1968 Ph. (Hypophylloceras) aphrodite Fallot & Termier ; Wiedmann & Dieni: 25.

NEOTYPE. Designated by Wiedmann, 1962a: 252; refigured by him in 1964:
pl. 18, figs 4a-c. From the Upper Albian of Son Vida-Son Suredeta, Mallorca
(Balearics).

MATERIAL. Two specimens. BM(NH)C78667 from Bed 6, Loc. 51 and
SAS PJR/z from Loc. 57, both along the Mzinene River, Zululand (Albian V).

ZULULAND AND NATAL 359

DIMENSIONS. D Wb Wh Wb/Wh U
SAS PJR/1 34:0 14°7 (43) 17:0 (50) 0°87 5°0 (7)
BM(NH) C78667 37°0 16-9 (46) 20:0 (54) 0-84 5:0 (13)

DESCRIPTION. The coiling is very involute, with a small, deep, conical umbilicus
(7-13% of diameter). The whorl section is slightly compressed, suboval, with the
greatest breadth at mid-flank. The umbilical wall is subvertical, sloping outwards ;
the umbilical shoulder is abruptly rounded, the flanks convex and markedly inflated
and the venter broadly and evenly rounded. The surface of the test bears fine
flexuous striae, and irregular low folds which arise at the umbilical seam, sweep
forwards across the inner flank, slightly backwards across mid-flank and forwards
again across shoulder and venter. There are four to five deep, narrow, flexuous
constrictions per whorl, paralleling the growth striae, which are weak at the
umbilical shoulder but deep across the flank and venter. The internal mould is
virtually smooth save for constrictions.

The suture includes a rather narrow diphyllic E/L saddle, separated by a trifid
lobe from a smaller but also diphyllic L/U,.

Discussion. The presence of constrictions throughout ontogeny serves to
distinguish this species from most other members of the thetys group. P. (Hypo-
phylloceras) aptiense aptiense is more compressed (whorl breadth to height ratio is
typically 0-7) and has broader, straighter constrictions ; P. (H.) aptiense applanatum
has an equally compressed whorl, flat, subparallel flanks and a somewhat flattened
venter (Wiedmann 1964 : text-fig. 43a).

OccURRENCE. Upper Albian of the Balearics, Sardinia, southern France
(Escragnolles), Sinai (Moghara Massif) and South Africa (Zululand).

2. Group of P. (Hypophylloceras) velledae (Michelin)

This species group is transitional between the ¢hetys and seresitense groups, having
diphyllic saddles during early ontogeny and tetraphyllic saddles when adult. The
following are referred to the group:

1. Hypophylloceras velledae velledae (Michelin) (1834 : pl. 35), Albian to Cenomanian.

2. Hypophylloceras velledae (Michelin) morelianum (d’Orbigny) (1840:176; pl. 54,
figs 1-3), Aptian to Albian.

3. Hypophylloceras velledae (Michelin) aschiltaense Breistroffer (1947 : 55), Aptian
to Albian.

4. Hypophylloceras velledae (Michelin) imflatum Collignon (1950: 35; pl. 4, figs
16-16b), Albian.

5. Hypophylloceras ponticult (Rousseau) (1842 : 783), Aptian.

6. Hypophylloceras onoense Stanton 1895 [= Ammonites ramosus Gabb 1864 : 65 ;
pl. 11, fig. 12; non Meek], Aptian.

7. Hypophylloceras cypris cypris Fallot & Termier (1923 : 23; pl. 3, figs Ia-c, 2c),
Aptian to Albian.

8. Hypophylloceras cypris Fallot & Termier cytherae Wiedmann (1964 : 218 ; pl. 12,
figs 2a—c, 7a—b), Aptian to Albian.

360 CRETACEOUS FAUNAS

9. Hypophylloceras pervinquierer Collignon (1928: 147, text-fig. 2; pl. XV, figs 9,
ga), Lower Cenomanian.

Phylloceras (Hypophylloceras) velledae (Michelin)

This is the most commonly cited Albian phylloceratid, and the most frequently
misinterpreted. Wiedmann (1964 : 205-207) has now clarified much of the con-
fusion surrounding the species, and has designated a neotype for the restricted form.
The species can be subdivided as follows :

1. Whorl section compressed, oval, maximum whorl breadth at umbilical wall,
sides flat — P. (H.) velledae morelianum (d’Orbigny), Aptian ? to Albian.

2. Whorl section compressed, oval, sides moderately inflated, greatest breadth on
inner third of flanks — P. (H.) velledae velledae (Michelin), Albian to Cenomanian.

3. Whorl section compressed, elliptical, flanks inflated, greatest breadth at mid-
flank — P. (H.) velledae aschiltaense Breistroffer, Aptian to Albian.

4. Whorl section subcircular, maximum whorl breadth on inner third of strongly
rounded flanks — P. (H.) velledae inflatum Collignon, Albian.

Phylloceras (Hypophylloceras) velledae velledae (Michelin)
(Pl. 4, fig. 5; Pl.5,f¢.4 > PL.6, fig: 1; Pl. 7.mes 2-3. Pls 0) aera

1834 Ammonites velledae Michelin : pl. 35.

1841 Ammonites velledae Michelin; d’Orbigny : 280; pl. 82, figs 1-4.
1907b Phylloceras sp. Crick : 236.
1921a Phylloceras velledae (Michelin
1925 Phylloceras velledae (Michelin) ; Spath : 181-182.

1928 Phylloceras velledae (Michelin) ; Collignon:6; pl. 1, figs I-TIa.
1930 Phylloceras velledae (Michelin) ; Besairie: 212; pl. 17, fig. 2.
(
(

);.Spath > 273.
ys;
3
);

1936 Phylloceras velledae (Michelin) ; Venzo: 66; pl. 5, figs 4a-b.

1947 Phylloceras velledae (Michelin) ; Breistroffer : 55.

1949a Phyllocervas velledae (Michelin), var. ; Collignon: 44; pl. 7, figs I-1a; text-figs 4, 5.

1960 Euphylloceras velledae Michelin; Drushchitza & Kudryavtseva : 252; pl. 2, figs 5a-b.

1962a Hypophylloceras velledae velledae (Michelin) ; Wiedmann: 250; pl. 16, fig. 4.

1962b Hypophylloceras velledae velledae (Michelin) ; Wiedmann: 142.

1963 Phylloceras velledae (Michelin) ; Collignon: 4; pl. 241, figs 1036.

1964 Phylloceras velledae (Michelin) ; Collignon: 4; pl. 318, fig. 1347.

1964 Phylloceras (Hypophylloceras) velledae velledae (Michelin) ; Wiedmann: 209; pl. I1,
figs Ia-c; pl. 13, figs 4a—b; pl. 21, figs 4a-—b; text-fig. 49. (With full synonymy.)

1968 Ph. (Hypophylloceras) velledae velledae (Michelin); Wiedmann & Dieni:25; pl. 1,
figs 5a-b.

NEOTYPE. Designated by Wiedmann, 1964: 211. MHNP 1895-17 (d’Orbigny
Collection) from the Albian of Epothémont, near Bar-sur-Aube, Aube, France.

MATERIAL. Numerous specimens. SASLJE86 from Loc. 163, Mfongosi
Creek (Aptian III) ; BM(NH) C78685 from Loc. 36 on the Mzinene River (Albian
III); BM(NH) C78668 from Bed 1 at Loc. 51, on the Mzinene (Albian IV) ;
BM(NH) C78677, C78670 and C78683 from Bed 6, C78684, C78682 from Beds 12-13

ZULULAND AND NATAL 361

and BM(NH) C78669 (loose) also from Loc. 51 (Albian V); BM(NH) C78678,
? C78679, SAS Z1503, Z1491, Z1492a-b, from Loc. 54 on the Mzinene ; BM(NH)
C78671-6 from Loc. 56 on the Mzinene ; SAS Ar1847, from Loc. 57 on the Mzinene ;
aly Albian V. SAS A2053, PJR/2, Z1195, 2564, Z1248, Zr247 from Loc. 64;
SAS 26/3214, 26/3227, 26/3223, BM(NH) C78680-1 and C78689 from Loc. 65, and
SAS Ar189 from Loc. 66, all on the Munywana (Albian V). SAS Z17(5), Z17(6)
and Ziga—b come from Haughton’s (1936: 291; fig. 3) localities 17 and 1g in
northern Zululand.

The specimen referred to by Spath (1921a : 273) from the ‘Middle Branch of the
Manuan Creek’, SAM 4992, also belongs to this species, as does Crick’s (1907b)
Phylloceras sp., BM(NH) C18264.

DIMENSIONS. D Wb Wh Wb/Wh U
Neotype 165-0 61-0 (37) 98-0 (59) 0-62 —
Collignon 1949b:45 1180 44:0 (37) 69:0 (59) 0°63 5 (4)
Collignon 1963 : 4 II4°0 43°0 (38) 68-0 60) 0-63 Sui (3)
SAS 3214 G97 5) 87037) 57°2 (57) 0°65 6-0 (6)
SAS Z19 117-2 400 (35) 700 (60) 057 5°5 (5)
SAS Z17 (5) 1375 °9 4777-135) 83:2 (60) 0°57 6°8 (5)
SAS Z1492a E62 Or 995 5"0 (34) 97°2 (60) 0°57 6°5 (4)
SAS Z1496 169-0 60:2 (36) 103°2 (61) 0-60 9°5 (6)
BM(NH) C78668 208-0 777°5 (37) 121-0 (58) 0°63 II°5 (6)

DeEscrRIPTION. The coiling is involute, with a tiny umbilicus, 3-7% of the
diameter, with an outwards sloping wall. The whorl section is compressed, the
whorl breadth to height ratio varying from 0-55 to 0-63, with the greatest breadth
well below mid-flank. The sides are gently rounded, and converge from the lower
flanks towards a quite broad, rounded venter. Ornament consists of fine flexuous
striae, hair-like around the umbilicus, but coarsening and flattening on the upper
flank and venter where they are approximately as wide as the interspaces ; there
are about two hundred per whorl. They sweep gently forwards across the inner
flank, backwards across mid-flank, and then forwards across the upper flank to pass
across the venter with a shallow forward projection. There are no constrictions
nor do the striae bunch. Internal moulds bear traces of the same ornament as the
shell surface.

The suture of large specimens includes an asymmetrically diphyllic saddle E/L,
separated by a trifid lateral lobe from an L/U, saddle, with five folioles (PI. 6, fig. 3a).

Discussion. Wiedmann gives a complete synonymy for this form, and demon-
strates very clearly how few of the specimens referred to Phylloceras velledae actually
belong to the restricted species. Our collections contain, however, a large number
of specimens ranging from juveniles to individuals which are still wholly septate at
diameters of over 200 mm and match the neotype in proportions, style of ornament
and suture. There is some variation in relative proportions and in strength of
ornament, but all the material belongs to a single form. Our material appears to
be the largest described collection of P. (H.) velledae velledae, and an ontogenetic
series and a number of adult variants are figured here.

362 CRETACEOUS FAUNAS

Differences between P. (Hypophylloceras) velledae velledae, morelianum, aschiltaense
and inflatum have already been given. The other Albian members of this group,
P. (H.) cypris cypris and P. (H.) cypris cytherae, both show distinctly fasciculate
ornament ; the former has a subtrigonal whorl section, and both have a more open
umbilicus.

P. (Hypophylloceras) seresitense, the other species of Hypophylloceras represented
by large individuals in our collections, has a much wider umbilicus, and has much
finer, straighter striae at similar diameters.

OcCURRENCE. Lower Albian of the Balearics ; Upper Albian of France ; Middle
Aptian and Upper Albian of South Africa ; Upper Albian of Mozambique ; Lower
Albian to Cenomanian of Madagascar.

Phylloceras (Hypophylloceras) velledae (Michelin) inflatum Collignon
(Pl 13) feo)

1950 Phylloceras velledae Michelin var. inflata Collignon : 35; pl. 4, figs 16-16b.
1962a Phylloceras velledae inflatum Collignon ; Wiedmann : 142.

1963 Phylloceras velledae Michelin var. infiata Collignon: 6; pl. 242, fig. 1040.
1964 Phylloceras velledae inflatum Collignon ; Wiedmann : 207.

LECTOTYPE. The specimen figured by Collignon (1950) as his pl. 4, figs 16-16b,
from the Lower Albian of Komihevitra, Madagascar, is herein designated lectotype.

MATERIAL. One specimen only, SAS 1742, from the Middle or Upper Albian of
the Mzinene River, Zululand.

DIMENSIONS. D Wb Wh Wb/Wh U
Lectotype 98-0 49:0 (50) 58-0 (59) 0°84 20 ' (2)
Collignon 1963 : 6 104:0 52°0 (50) 64:0 (62) 0-81 4:0 (4)
SAS 1742 81-0 40:0 (50) 45°3 (56) 0-88 4:0? (5?)

DEscRIPTION. The coiling is very involute, with a tiny, occluded umbilicus.
An outward-sloping umbilical shoulder produces a conical circumbilical depression.
The whorl section is subrounded, the whorl breadth to height ratio being 0-88.
The whorl sides are inflated and the venter broadly rounded ; the greatest breadth
is close to the umbilical shoulder. Ornament consists of fine, dense, flexuous
striae, weak on the lower flank, but strong on the upper flanks and venter. The
striae arise at the umbilical seam, pass straight across the lower flank, flex gently
backwards at mid-flank, and gently forwards across shoulders and venter. The
internal mould bears weaker but otherwise similar ornament.

Discussion. The occluded umbilicus, whorl section and ornament clearly place
this specimen in P. (Hypophylloceras) velledae inflatum ; it is identical with the
Malagasy material figured by Collignon, save for slight differences in relative pro-
portions. This form is readily distinguished from other subspecies of velledae, none
of which approaches its degree of inflation. Form of umbilicus and ornament, as
well as inflation, clearly differentiate it from other members of the velledae group.

ZULULAND AND NATAL 363

OCCURRENCE. Lower Albian of Madagascar; Middle or Upper Albian of South
Africa (Zululand).

Phylloceras (Hypophylloceras) sp. aff. velledae (Michelin)
(Pl. 5, fig. x; Pl. 7, fig. 1)
cf. 1972 Hypophylloceras aff. californicum (Anderson) ; McLearn: 22; pl. 1, figs 1a—b, 2a-b.

MATERIAL. Three specimens: BM(NH) C78696, from Loc. 51, collected loose
but of Upper Albian age ; BM(NH) C78695 from Loc. 56 ; SAS A2239 from Loc. 57 ;
all Mzinene River, Zululand (Albian V).

DIMENSIONS. D Wb Wh Wb/Wh U
BM(NH) C78695 81-8 31°0 (38) 46°5 (57) 0-66 5-0 (6)
BM(NH) C78696 107°5 —- 43°5 (40) 64-2 (59) 0°67 7:2 (6)
SAS A2239 147°4 “3 89°5 (60) rig g'I (6)

DEscrRiPTION. Three specimens in our collection appear to belong to the velledae
group, but differ from P. (H.) velledae velledae in having a broader whorl section and
from velledae velledae and velledae inflatum in their coarser ornament, effaced on the
inner flank and not obviously flexuous. McLearn has recently figured (1972 : pl. 1,
figs I-2, especially 2) similar specimens from the Lower Albian of the Queen
Charlotte Islands, British Columbia, referring to them as Hypophylloceras aff.
californicum (Anderson). The type of this species is, however, a finely ornamented
form, known only as a fragment from the Lower Albian of California (Anderson
1938 : 143; pl. 12, fig. 7) and which is too poorly described and figured for adequate
comparisons.

OccURRENCE. Upper Albian of Zululand. Lower and Upper Albian of the
Queen Charlotte Islands, British Columbia.

3. Group of P. (Hypophylloceras) seresitense Pervinquiére

This group, characterized by possession of tetraphylloid saddles even in the early
stages, includes seven Lower Cretaceous forms, most of which are listed by Collignon
(1956: 29 et seq.). Included here are the Neophylloceras, Epiphylloceras and
Hyporbulites of previous workers.

Phylloceras (Hypophylloceras) seresitense Pervinquiére

The species can be divided as follows (Wiedmann 1964 : 221).

1. Whorl section moderately compressed, whorl breadth to height ratio up to 0-65,
umbilicus narrow—P. (H.) seresitense seresitense Pervinquiére, Aptian to
Cenomanian.

2. Whorl section very compressed, whorl breadth to height ratio 0-50-0°57,
umbilicus narrow — P. (H.) seresitense boule Collignon, Albian to Cenomanian.

3. Whorl section high, umbilicus moderately open-P. (H.) seresitense tanit
Pervinquiére, Albian to Cenomanian.

29

364 CRETACEOUS FAUNAS

Phylloceras (Hypophylloceras) seresitense seresitense Pervinquiére
(Pl. 4, fig. 65° Pl 6, fe a Pe 7 ties 2 = ero)

1865 Ammonites velledae Michelin ; Stoliczka: 116; pl. 59, fig. 1.
1895 Phylloceras velledae Michelin; Kossmat: 12; pl. 1, fig. 3.
1907 Phylloceras velledae var. sevesitense Pervinquiere : 52.
1907a Phylloceras velledae Michelin ; Crick: 166; pl. Io, figs 10-11.
1910 Phylloceras velledae var. sevesitensis Pervinquiére ; Pervinquiere: 9, text-fig. 2 ;
pl. 1, figs 1-3.
non 1923 Phylloceras sevesitense Pervinquiere; Spath:18; pl. 1, fig. 2; pl. 2, fig. 1
[= P. servesitense tanit Pervinquiere].
1928 Phylloceras ex aff. vramosi Meek; Collignon:1; pl. 1, figs 2-4.
1936 Phylloceras sevesitense Pervinquiére ; Collignon: 190; pl. 21, figs 1-2.
1940 Phylloceras sevesitense Pervinquiére ; Fabre: 211; pl. 5, fig. 1.
1943 Phylloceras velledae Michelin ; Matsumoto : 676.
1947 Hyporbulites seresitensis Pervinquiére ; Breistroffer : 82.
1950 Phylloceras (Hyporbulties) seresitense Pervinquiere ; Collignon : 66.
1956 Hyporbulites seresitensis var. vraynaudiensis Collignon: 16; pl. 4, fig. 1.
1959 Neophylloceras seresitense (Pervinquiére) ; Matsumoto: 55, text-fig. 3; pl. 12,
figs 4-5.
1962a Hypophylloceras seresitense sevesitense (Pervinquiere) ; Wiedmann: 142, text-fig.
8; pl. 8, figs 1-2.
1962b Hypophylloceras seresitense sevesitense (Pervinquiere) ; Wiedmann: 249; pl. 16,
He. L.
1963 Phylloceras (Hyporbulites) sevesitensis Pervinquiere ; Collignon: 4; pl. 241, fig.
1038; pl. 142, fig. 1041.
1964 Ph. (Hypophylloceras) seresitense seresitense Pervinquiére ; Wiedmann : 221; pl. 15,
figs 4a-b; pl. 21, figs 1a-b; text-fig. 52.
1968 Phylloceras (Hypophylloceras) seresitense sevesitense Pervinquiere; Wiedmann &
Dieni : 26.

ow OU

Lectotype. The original of Pervinquiére 1910: pl. 1, fig. 1, designated by
Wiedmann (1964).
MATERIAL. Four specimens: BM(NH) C78690, C18137-8 from the Skoenberg

(? Cenomanian III), and SAS Ar4or from Loc. 36 on the Mzinene River (Albian
III), all from Zululand.

DIMENSIONS. D Wb Wh Wb/Wh U
BM(NH) Cr18138 28:3 10°6 (37) 16-7 (59) 0°64 I°5 (5)
SAS Ar4or 190-0 69°3 (36) 109°2 (57) 0°63 10°5 (5)

DESCRIPTION. Our specimens add nothing to knowledge of the early develop-
ment of this species, which is well known. We have, however, a near adult specimen
(SAS Ar4ot, Pl. 9) still wholly septate at 190 mm, which appears to be the largest
known individual referred to it. The coiling is very involute, the umbilicus small,
conical and deep (5% of diameter) with a rounded undercut wall and an outwards
sloping shoulder which merges with the flanks. The whorl section is compressed,
the greatest breadth being one-third of the way up the flanks. The sides are gently
inflated, converging to a broad rounded venter.

Ornament consists of dense, fine striae, flat-topped, and as wide as the interspaces.
There are between 400 and 450 of these on the outer whorl. They arise at the

ZULULAND AND NATAL 365

umbilical seam and are very faint, but strengthen across the flanks. They flex
gently forwards on the inner flank, backwards over the outer flank and forwards
across the shoulders to pass straight across the venter. The major saddles of the
suture (E/L and L/U,) are tetraphylloid.

Discussion. Our smaller specimens, originally figured by Crick as Phylloceras
velledae Stoliczka (? Michelin), clearly belong to P. (H.) seresitense seresitense, as
indicated by Wiedmann (1964). Our adult specimen, still wholly septate at a
diameter of 190 mm, is referred to P. (Hypophylloceras) seresitense on the basis of
the style and strength of ornament, relative proportions and such of the suture line
as is visible. The whorl breadth to height ratio clearly excludes reference to
P. (H—.) seresitense boule, but separation from P. (H.) seresitense tanit cannot be
definitely confirmed at this size because the largest described tanit has a diameter
of only 62mm. It seems, however, that the umbilicus of tamit is usually 10% or
more of the diameter.

OCCURRENCE. P. (Hypophylloceras) seresitense seresitense ranges from the Upper
Aptian (Balearics) to the Cenomanian. The geographic distribution includes
southern France, northern Spain, north Africa, the Balearics, southern India,
Madagascar, South Africa and Alaska.

Phylloceras (Hypophylloceras) androiavense Besairie

(Pls 4;fige'7)

1936 Phylloceras androiavensis Besairie: 165; pl. 16, figs 9-11.
1963 Phylloceras (Hyporbulites) androiavense Besairie ; Collignon: 4; pl. 241, fig. 1037.

LECTOTYPE. Besairie’s figured specimen (1936: pl. 16, figs g-11), from the
upper part of the Middle Albian of Androiavy, Madagascar, is herein designated
lectotype.

MATERIAL. One specimen only, BM(NH) C78667, from Loc. 54, Mzinene River,
Zululand (Albian V).

DESCRIPTION. Our specimen is fragmentary, with an original diameter of
approximately 40mm. The coiling is very involute, with a tiny umbilicus (approxi-
mately 5°% of the diameter). The whorl section is very compressed (whorl breadth
to height ratio about 0-46), the greatest breadth being at the umbilical shoulder.
The whorls are rapidly expanding, the whorl height being approximately 60% of
the total diameter. The umbilical shoulder is quite abruptly rounded, the sides
flat and subparallel with a narrow rounded venter. Ornament consists of fine,
flexuous striae grouped into bunches; about ten low, fold-like ribs arise at the
umbilical shoulder, pass forwards across the inner flank, and flex backwards at mid-
flank where they give rise to numerous striae which sweep gently forwards across
the shoulder and venter, together with similar intercalated striae. The inner
flanks between folds bear fine irregular striae. The sutures are not exposed.

Discussion. Our specimen is identical in style of ornament, whorl section and
(as far as can be determined) relative proportions to P. (Hypophylloceras) androta-
vense. Apart from this species, fasciculate ornament is present in the following

29*

366 CRETACEOUS FAUNAS

mid-Cretaceous forms. P.(Hypophylloceras) cypris cypris Fallot & Termier (1923 : 22;
pl. 3, figs ra—c, 2c) is an Albian form, differing, however, in having a subtrigonal
whorl section, a whorl breadth to height ratio of 0-66, and a larger umbilicus 12%
of the diameter. P. (Hypophylloceras) cypris cytherae Wiedmann (1964: 218 ;
pl. 12, figs 2a—b, 7a-b; pl. 15, fig. 8a—b; text-fig. 12) is an Upper Aptian to Albian
form with a compressed, oval whorl section, rounded flanks, a whorl breadth to
height ratio of 0-66 and a larger umbilicus 12% of the diameter. Hypophylloceras
algeriense Wiedmann (1962:144; pl. 8, fig. 3a-b; text-fig. 9), an Albian to
Cenomanian species with a whorl breadth to height ratio of 0-65, has rounded,
convergent flanks and an arched venter.

OccURRENCE. High Middle Albian of Madagascar; Upper Albian of South
Africa (Zululand).

Phylloceras (Hypophylloceras) woodsi woodsi van Hoepen
(Figs 4-6; Pl. 13, figs 3-5)

1906 Phylloceras sp., Woods : 331; pl. 41, fig. 4.

1921 Phylloceras woodsi van Hoepen: 3; pl. 2, figs 1-6; text-fig. 1.

1921b Phylloceras neva (Forbes) ; Spath : 4o.

1922 Phylloceras woodsi van Hoepen; Spath: 117.

1929 Phylloceras woodsit van Hoepen; Barrabé: 174; pl. 9, fig. 17.

1930 Phylloceras woodsi van Hoepen; Besairie : 567 (223); pl. 21, figs 2-2a.

1931 Phylloceras woodsi van Hoepen ; Collignon : ro.

1956 Hyporbulites woodsi van Hoepen ; Collignon : 19.

1956 Hyporbulites hoepent Collignon: 19; pl. 1, figs 4, 4b, 5, 5b, 6-6b; text-fig. 4.
1966 Phylloceras (Hyporbulites) hoepeni Collignon ; Collignon:1; pl. 455, fig. 1406.

HorotyPe. TM 537, figured by van Hoepen (1921: pl. 2, figs 3-4).

“wae

Fic. 4. External suture line of Phylloceras (Hypophylloceras) woodsi woodsit van Hoepen.
Holotype, TM 537, x 12°5.

ZULULAND AND NATAL 367

Fic. 5. Partial external suture of Phyllo-
cevas (Hypophylloceras) woodst woodsi van
Hloepen. , TM 520, x'6:

MATERIAL. Eight specimens: the holotype TM 537; four paratypes, TM 5209,
530, 533, 534; a specimen in the Durban Museum (unregistered) ; a specimen in
the Griesbach Collection, figured by Woods (1906) ; and BM(NH) C78826. All are
from the Umzamba Formation of Loc. 1, Umzamba, Pondoland, and are of Santonian
or Campanian age.

DEscriPTION. The coiling is moderately involute (umbilicus approximately
7-10% of diameter). The whorl section is a compressed oval with virtually parallel
flanks, converging only slightly towards the gently rounded venter. Maximum
whorl width is at the middle to inner third of the flanks. The umbilicus is pit-like,
with nearly vertical walls, thus forming a pronounced, angular umbilical shoulder.

Ornament consists of numerous, fine single ribs which are most pronounced on
the outer third of the flanks. Ribbing on the inner third is very faint, but the ribs
appear to originate at the umbilical edge, flex slightly backward on the inner third
of the flanks and gently forward at mid-flank, and then pass in a prorsiradiate
direction across the ventrolateral shoulders.

The suture is typically phylloid on the inner whorls, but tends to become more
dendritic at the largest diameters studied. E/L is diphylloid, separated from the
triphylloid L/U, by a trifid lateral lobe.

Discussion. Collignon (1956: 19), when erecting P. (H.) hoepent, stated that
the species was very close to P. (H.) woods; as far as relative proportions are
concerned, the two species overlap to a great extent. P. (H.) hoepens has a smaller
umbilicus than the present species, 3-6% of diameter as against 7-10% in P. (#.)
woodst, but Collignon realized that this criterion is difficult and rather impracticable
to apply. He stated, in addition, that the suture lines were different ; in P. (#.)
hoepeni the suture line is more zigzag-like and the saddles have smaller stems and
are more incised. Unfortunately, only E/L and part of the lateral lobe of P. (#.)
hoepent have been figured by Collignon (reproduced here as Fig. 6), whilst the
suture line given by van Hoepen for P. (H.) woods is rather schematic.

368 CRETACEOUS FAUNAS

oe Fic. 6. Partial external suture line of
Ye, 3 | ‘Hyporbulites’ hoepeni Collignon (= P.
|
|
|

(Hypophylloceras) woodsi woodsi van Hoepen
herein), after Collignon (1956 : text-fig. 4),

X 4°5.

Virtually the same part of the suture line as figured by Collignon is visible on one
of the paratypes of P. (H.) woodst, TM 529 (Fig. 5); it is quite clear that, apart
from the folioles in the saddles of P. (H.) hoepent being more phylloid than in
P. (H.) woodsi, the two suture lines are identical. Because of the comparable
relative proportions, similar suture lines and (probably) the same stratigraphic
occurrence, P. (H.) hoepeni is here regarded as a junior synonym of P. (H.) woodsv.
P. (H.) hoepent infundibuliformis Collignon is also referred to P. (H.) woodsi as a
subspecies. It is more inflated than the nominal subspecies and stands in the same
relationship to P. (H.) woodsi that P. (H.) subalpinum ellipticum does to P. (H.)
subalpinum subalpinum.

Less readily resolved is the relationship of P. (H.) woods: to ‘Neophylloceras’
marshall Shimizu [= Phylloceras nera Marshall (non Forbes) 1926: 124; pl. Io,
fig. 4; pl. 26, figs 1-2], ‘N.’ hetonaiense Matsumoto (1943 : 675, text-figs Ias, bs) of
which ‘N.’ lambertense Usher (1952: 50; pl. I, figs 1-3) is a synonym (Jones 1963 :
23), and ‘N.’ neva Forbes (1846: 106; pl. 8, fig. 7). All have ornament well within
a range of intraspecific variation which we would find acceptable in phylloceratids.
In the suture line of N. hetonaiense, however, the elements of the suspensive lobes
form a concave arc ; in N. nera they lie on a straight line, and in the present species
the suspensive lobe is slightly retracted (Fig. 4). Given better collections, it may be
possible to show that all these species are junior subjective synonyms of P. (Hypo-
phylloceras) nera.

OCCURRENCE. Santonian or Campanian of Zululand; Santonian Texanites
olivett Zone, Madagascar.

Phylloceras (Hypophylloceras) mikobokense (Collignon)
(Ph, fig 2)

1938 Phylloceras aff. surya (Forbes) ; Collignon: 41; pl. 7, fig. 1.
1956 Epiphylloceras mikobokense Collignon : 24; pl. 2, figs 3-3a; pl. 4, figs 5-5b.

HoLotyPeE. MNHP 1567, figured by Collignon (1956: pl. 4, figs 5-5b), from the
Lower Maastrichtian of Lajona, Madagascar.

MATERIAL. Two specimens, SAS H163c/7 and 163/9, from Loc. 20, Mfolozi
River, Zululand (Maastrichtian I).

ZULULAND AND NATAL 369

DIMENSIONS. (From Collignon 1956 : 24)

Wb Wh Wob/Wh U
EMP 1567 86:0 25°0 (29) 42-0 (49) 59 =
EMP 1564 73°0 23°0 (32) 42:0 (58) 55 II-0 (15)

DESCRIPTION. Our specimens are fragments of discoidal shells with a moderately
high expansion rate and a compressed whorl section (whorl breadth to height ratio
is 0°51), the greatest breadth being close to the umbilical shoulder. The umbilicus
is small, about 10% of the inferred diameter, quite deep, with a subvertical, rounded
wall and an abruptly rounded shoulder. The flanks are rounded, converging to a
narrowly rounded venter.

Ornament consists of fasciculate ribs and striae. Our specimens bear strong
fold-like ribs which arise at the umbilical seam, pass straight across the lower flank,
thicken and flex backwards across the outer flank, and die out across the ventro-
lateral shoulder, being represented on the venter merely by low, broad swellings.
Superimposed on these are fine, dense striae, arising at the umbilical seam, branching
on the flanks and following a course parallel to the ribs.

The suture is deeply incised, having lost the typical phylloid terminations
(rier, fis. rb).

Discussion. The very coarse, fold-like ribs and fine striae, and relative pro-
portions, indicate that our specimens belong to Collignon’s species, comparing
closely with the large paratype fragment figured by him (1956: pl. 2, figs 3-3a).
Ornament readily separates this species from most Upper Cretaceous Hypophylloceras,
whilst those forms with comparable fasciculate ornament (‘Epiphylloceras’ of
authors) differ as follows. Hypophylloceras surya (Forbes) (1846: 106; pl. 7,
fig. 10) has much coarser ribbing rather than striae; Hypophylloceras cottreaus
Collignon (1956: 25; pl. 4, figs 6a—b) lacks strong fold-like ribs, having instead
fasciculate striae arising from umbilical bullae. Efiphylloceras radiatum Marshall
(1926: 135; pl. 19, fig. 7; pl. 26, figs 3-4; see also Henderson 1970: 6, text-fig.
26) has a smaller umbilicus and less compressed whorls, the whorl breadth to height
ratio being 0-72.

OccURRENCE. Lower Maastrichtian of Madagascar and South Africa (Zululand).

Genus PARTSCHICERAS Fucini 1920

TYPE SPECIES. Ammonites partschi Stur 1851 (non Klipstein 1843), by the
subsequent designation of Spath, 1927.

Discussion. This genus is reviewed at length by Wiedmann (1962b: 257 ;
1964: 229). Our material is too poor to allow profitable discussion, and we have
therefore followed him in regarding Phyllopachyceras Spath 1925, Procliviceras
Fucini 1920, Macrophylloceras Spath 1927, Hoplophylloceras Spath 1927, Partchi-
phylloceras Roman 1938 and ? Calaiceras Kovacs 1939, as synonyms.

OccURRENCE. Sinemurian to Maastrichtian ; in the Cretaceous species have a
world-wide distribution save for the boreal region of north-west Europe and the
western interior of North America.

370 CRETACEOUS FAUNAS

Partschiceras umzambiense (van Hoepen)
(Fig. 7; Pl. 13) fig. 2)

1920 Phylloceras umzambiense van Hoepen: 142; pl. 24, figs 1-3.

1921b Phylloceras umzambiense van Hoepen ; Spath: 50.

1922 Phylloceras umzambiense van Hoepen ; Spath: 117.

1956 Phyllopachyceras umzambiense (van Hoepen) ; Collignon : 27 ; pl. 2, figs 2-2b; text-fig. 8.
1966 Phyllopachyceras umzambiense (van Hoepen) ; Collignon: pl. 455, fig. 1850.

HototyPeE. TM 524, figured by van Hoepen (1920: pl. 24, figs 1-3), from the
Umzamba Formation, Loc. 1, Pondoland, and of Santonian-Campanian age.

MATERIAL. The holotype only.

DIMENSIONS. D Wb Wh Wb/Wh U
TM 524 44:0 17°0 (39) 27°0 (61) 0°63 2:0 (5)
EMP 1587

(Collignon 1956) 22:0 100 (45) 13°0 (59) 0-76 I-0 (5)
Collignon 1966 :

pl. 455, fig. 1850 21 100 (49) 12. (57) 0°83 1:0 (9)

DescriPTIion. The holotype is an incomplete, septate internal mould, badly
abraded on one side. Coiling is involute with a small umbilicus, 5% of the diameter.
The umbilicus is funnel-shaped and not distinctly demarcated from the flanks.
Whorl section is ovoid with rounded flanks and a broad venter. Maximum breadth
is at the inner third of the whorl.

Ornament consists of thin, sinusoidal ribs, separated by somewhat wider inter-
spaces. The ribs flex forwards on the inner half of the flanks and then backwards on
the outer half to traverse the venter radially. Approximately eight pairs of fascicu-
late ribs arise at the umbilicus as low, radial swellings which are confined to the
umbilicus, disappearing on the flanks. Apart from the normal ribs, intercalatories
or bifurcating ribs originate at about mid-flank.

The main saddles (E/L and L/U,) are tetraphyllic. L is larger than U, (Fig. 7).

Discussion. This is a relatively rare species and has only been recorded on
three occasions to date (van Hoepen 1921; Collignon 1956, 1966). The main
distinguishing feature is the ornamentation of fasciculate ribbing near the umbilicus
and intercalatory bifurcating ribbing on the vental part of the flanks.

Partschiceras besairiet besairier Collignon and P. besairiet tsianalokensis (Collignon
I931:10; pl. 1, figs 2-2b, 3-3a, 4-4b; pl. 8, fig. 1) from the Santonian of
Madagascar bear a superficial resemblance as far as inflation of the whorls is con-
cerned, but has different ornamentation; P. besairier besairier has strong ribs on
the ventral area and P. besairiei tstanalokensis lacks ribbing altogether. Of other
Upper Cretaceous species, Parschiceras forbesianum (d’Orbigny 1850 : 213) (of which
according to Henderson (1970: 7) P. ezoense Yokoyama (1890: 170; pl. 19, figs
2a—c), P. minimum Marshall (1926 : 137; pl. 19, fig. 8; pl. 26, figs 5-6), P. bistriatum
Marshall (1926: 138; pl. 19, fig. 5; pl. 27, figs 1-2), P. inflatum Shimizu (1935 :
178), P. marshalli Collignon (1937 : 26) and P. zelandicum Collignon (1956 : 31) are
synonyms), P. whiteavesi Kossmat (1897:124-125 [= Phylloceras forbesianum

ZULULAND AND NATAL 371

4

Fic. 7. External suture line of Partschiceras umzambiense (van Hoepen). Holotype,
TM 524, x6.

Kossmat (non d’Orbigny) 1895: 109; pl. 15, figs Ia-c; = Ammonites rowyanus
Stoliczka (non d’Orbigny) 1865:117; pl. 69, figs 5-7]) and P. occlusum Tate
(1865 : 36; pl. 3, figs ra—b) all lack the fine, dense, wiry ribbing of P. wmzambiense
and are virtually smooth when preserved as internal moulds. ‘Epiphylloceras’
species, which may bear similar fasciculate ribbing, all differ from P. wmzambiense
in their compressed whorl section.

OCCURRENCE. Santonian of Madagascar (Texanites oliveti Zone) ; Santonian or
Lower Campanian of Pondoland.

Partschiceras sp. juv. ? cf. baborense (Coquand)
(Pl. 6, figs 2-3)

cf. 1880 Ammonites baborensis Coquand : 26.
cf. 1964 Partschiceras baborense (Coquand) ; Wiedmann: 243; pl. 14, figs 2, 4, 5; pl. 16,
figs 1-2; pl. 21, figs 5, 6; text-fig. 59. (With synonymy.)
MATERIAL. Three specimens, BM(NH) C78692-4, from Loc. 170, Mlambong-
wenya Spruit, northern Zululand (Barremian I).

DIMENSIONS. D Wb Wh U
BM(NH) C78693 18-0 — (—) IOI (56) —

DESCRIPTION. The shell is involute, with an occluded umbilicus. The whorls
are slightly compressed, the greatest breadth being close to mid-flank. Although
the umbilicus is occluded, the umbilical wall slopes outwards to produce a shallow
conical circumbilical pit before merging with the broadly rounded, subparallel
flanks. The ventrolateral shoulder and venter are broadly and evenly rounded.
The shell surface is quite without ornament, although this may be a result of partial
decortication.

The suture is too poorly exposed and corroded for detailed description.

Discussion. These specimens are too poor for satisfactory determination.
The nature of the umbilicus, lack of ornament and relative proportions recall,

372 CRETACEOUS FAUNAS

however, Partschiceras baborense, especially pyritic juveniles figured by Wiedmann
(1964) and Collignon (1937: pl. 1, figs 4-4b to 6-6b).

OccURRENCE. P. baborense ranges from Barremian to Aptian, and occurs widely
in southern France, northern Spain, the Balearics, Sardinia, central Europe, the
Caucasus, Bulgaria, the Crimea, north Africa and Madagascar.

Genus CARINOPHYLLOCERAS Klinger, Wiedmann & Kennedy 1975

TYPE SPECIES. Carinophylloceras collignont Klinger, Wiedmann & Kennedy
1975: 658; pls 76-77; text-figs 1-3.

Diacnosis. Phylloceratid ammonites with fastigate to distinctly keeled venters.
Whorl section ovoid, higher than wide, with maximum width at umbilical margin ;
narrowly umbilicated. Ornament typically phylloceratid, consisting of striae.
Suture phylloid, with lituid I, trifid L, saddles E/L assymmetrically diphyllic,
L/U, asymmetrically tetraphyllic. Saddles in U; asymmetrically diphyllic (Figs
8-0).

Discussion. Carinophylloceras is unique amongst Cretaceous phylloceratids by
virtue of the presence of a keel. Keeled phylloceratids occur in the Jurassic,
e.g. Harpophylloceras Spath 1927 and Menegheniceras Hyatt 1900. There are,
however, no Cretaceous taxa referable to these genera and any affinity of Carino-
phylloceras with these forms may be ruled out.

The suture line of Carinophylloceras with an asymmetrical diphyllic saddle E/L
and asymmetrical tetraphyllic L/U,, the ornamentation, degree of evolution and to
a lesser extent whorl section point to affinities with the Albian/Cenomanian Phyllo-
ceras (Hypophylloceras) velledae (sensu Wiedmann 1964), and to the Albian/Aptian
P. (H.) cypris cypris Fallot & Termier (Wiedmann 1964: fig. 50; pl. 13, fig. 3 etc.).
Apart from the keel, the whorl section is somewhat intermediate between P. (H.)
velledae velledae and P. (H.) velledae morelianum. ‘The presence of a keel, however,
clearly separates Carinophylloceras colliignomi from the group of P. (H.) velledae.
Carinophylloceras is a homeomorph of Damesites, and was originally believed to be
a new species of that genus. The similarities are remarkable. Not only the whorl
section and the presence and shape of the keel, but also the degree of shell involution,

Sle so: al

Fic. 8. External suture line of Carinophyllocevas collignoni Klinger, Wiedmann &
Kennedy. “UP B404 91-3:

ZULULAND AND NATAL 373

the course of the ornamentation and even the external suture line show such
similarities that these genera can scarcely be distinguished on their external
morphology. Examination of the suture line, however, especially the internal part,
reveals that it is typically phylloid with a lituid internal lobe (Fig. 9), a characteristic
of all phylloceratinids (Wiedmann 1972:115). The internal lobe of Damesites as
figured by Matsumoto (1954 : fig. 11) is intensively frilled and of desmoceratid type ;
indeed vo known desmoceratids possess a lituid internal lobe.

OcCURRENCE. Lower Albian (Albian III) of Zululand.

Carinophylloceras collignoni Klinger, Wiedmann & Kennedy
(Figs 8-9; Pl. 14; Pl. 15, figs 1-2)
1975 Damesites ? sp.nov. Kennedy & Klinger : 276.
1975 Carinophylloceras collignoni Klinger, Wiedmann & Kennedy: 658; pl. 76, figs 1a-b;
pl. 77, figs 1-3; text-figs 1-3.

HoitotyPe. SAS A1577, from the Mzinene Formation, stream cliff section along
the Mzinene River 1200 m NE of the farm Amatis, north of Hluhluwe, Zululand,
South Africa, 27°58’03” S, 32°18’34” E. Loc. 35 of Kennedy & Klinger (1975).

MATERIAL. In addition to the holotype we have 39 paratypes: SAS UMS/2,
SAS A1133 and BM(NH) C78639, C78644, C78647-8, C78767, C78769, C78770 from
Loc. 35, and BM(NH) C78640-3, C78645-6, C78651 and C78768 from Loc. 36,
beth on the Mzmene River. SAS Ho3D/1, SAS Ho3/1, SAS Ho3/2, SAS 93/3,
SAS H93/5, from Loc. 142, Nxala Estate, southern part of Mkuze Game Reserve,
Zululand. SAS EM oz, SASEMg2z, SASEM77 from the Msunduzi Pan at
a7 25 5, 321240 E>; UPEB33 from the same area, at 26°57’1I0° 5,
32°12'45’ E. SAS EM 245, b, c, SAS EM 93, SAS EM 244, SAS EM 114, from the
Ndumu region, northern Zululand at 26°55'55"S, 32°12'55” E. SAS LJE 134A,
UPE B463, UPE B464, UPE B4rr and BM(NH) C78649-50, from Loc. 174, and
BM(NH) C78766 and C78771, from Loc. 171, both Mlambongwenya Spruit, northern
Zululand. UPE B23, from Aloe Flats Estate, northern Zululand, at 26°59’50” S,
321150’ E.

All specimens are from the Mzinene Formation of late early Albian age (Albian
III).

DIMENSIONS. D Wb Wh Wb/Wh U

Holotype SAS A1577 149 60:5 (41) 88-5 (59) 0:68 8-5 (6)

Paratypes SASEM244 123°5 44:5 (36) 68-0 (55) 0-65 8-0 (6:5)
SAS H98/z G93 HS2 DAT) A454 58) 0°73 6-0 (8)
SAS Hg3/3 108 = 40°5 (37) 63°'5 2 0°64 = 7°5 (7)

SAS UMS/z E325 AZ-01(30\n> G75 (58
DESCRIPTION. Coiling is moderately involute with a narrow funnel-shaped
umbilicus (6-8 of diameter). Whorl section is subtrigonal with a fastigate to
distinctly keeled venter. Maximum width is at the umbilical edge. In juvenile
stages the venter is fastigate, but in the adult a distinct keel is developed. The

374 CRETACEOUS FAUNAS

Fic. 9. Partial internal suture line of
Carinophylloceras collignoni Klinger, Wied-
mann & Kennedy, showing overlapping
lituid internal lobe. SAS EM 114, x 8:3.

keel is of the floored type, and, depending on the mode of preservation, may be
either present or absent on internal moulds.

Ornament consists of pronounced striae which arise at the umbilical wall, are
bent forwards at first, then sweeping gently backwards near the middle of the flanks,
being finally strongly projected on the outer part of the flanks. They are bundled
at their origin, and much stronger on the outer part of the whorls and venter,
producing a chevron-like ventrolateral and ventral ornament. On internal moulds
the ornamentation is still present, though very much subdued.

Suture line as for genus. Auxiliary saddles in U; are triphyllic.

OccURRENCE. Lower Albian (Albian III) of Zululand.

III. ACKNOWLEDGEMENTS

We are grateful to Dr H. W. Ball, Dr M. K. Howarth and Mr D. Phillips of the
Department of Palaeontology, British Museum (Natural History), General
M. Collignon (Moirans), Mr C. W. Wright (London), Professor Dr J. Wiedmann
(Tubingen), Professor D. L. Visser (Pretoria), Dr C. K. Brain and Mrs E. Voigt
(Transvaal Museum, Pretoria), Mr P. J. Rossouw (Pretoria), Mr M. R. Cooper
(South African Museum, Cape Town) and the technical staff of the Department of
Geology, Oxford, Geological Survey of South Africa, Pretoria, Institut fiir Geologie
und Palaontologie, Tubingen, and Transvaal Museum, Pretoria, for their help and
assistance in many ways. Thanks are also due to the Director, Geological Survey,
Pretoria, for permission for one of us (H. C. K.) to publish these results.

IV. REFERENCES

ALMELA, A. & DE LA REviLA, J. 1957. Fédsiles piritosos de Cretaceo de la Sierra de Ricote
(Murcia). Boln Inst. geol. min. Esp., Madrid, 68 : 45-83, pls 1-11.

ANDERSON, F. M. 1938. Lower Cretaceous deposits in California and Oregon. Spec. Pap.
geol. Soc. Am., New York, 16, 244 pp., 84 pls.

ARNOULD-SAGET,S. 1951. Lesammonites pyriteuses du Tithonique supérieur et du Berriasien
de Tunisie central. Annls Mines Géol., Tunis, 10: 1-132, pls I-11.

ZULULAND AND NATAL 3)//2)

BaRRABE, L. 1929. Contribution a l'étude stratigraphique et pétrographique du pays de
Sakalave, Madagascar. Mém. Soc. géol. Fr., Paris, n.s. 12 : 1-270, pls I-Io.

BESAIRIE, H. 1930. JRecherches géologiques 4 Madagascar. Contribution a l'étude des

ressources minérales. Bull. Soc. Hist. nat. Toulouse, 60 : 345-616 (1-272), pls 1-27.

1936. Recherches géologiques a Madagascar. Premiére suite. La géologie du nord-
ouest. Mém. Acad. malgache, Tananarive, 21 : 9-259, pls 1-24.

Beznosov, N. V. 1958. Jurassic ammonites of the northern Caucasus and Crimea. Phyllo-
ceratina and Lytoceratina. 118 pp, 34 pls. Gostoptekhizdat, Leningrad. [In Russian.]

Bout, M., Lemoine, P. & THEVENIN, A. 1906-07. Paléontologie de Madagascar. III —
Céphalopodes crétacés des environs de Diego-Suarez. Annis Paléont., Paris, 1 : 173-192,
pls 14-20 (1906) ; 2: 1-56, pls 1-8 (1907).

BREISTROFFER, M. 1947. Sur les zones d’ammonites dans |’Albien de France et d’Angleterre.
Trav. Lab. Géol. Univ. Grenoble, 26 : 1-88.

Busnarpbo, R. & Davin, L. 1957. Contribution a l’étude des faunes d’Ammonoides de
Medjez Sfa (Est Constantinois). Publs Serv. Carte géol. Algér., Algiers, n.s. 13 : 67-123,
pls 1-3.

CoLtLticNon, M. 1928-29. Les Céphalopodes du Cénomanien pyriteux de Diego-Suarez.
Annls Paléont., Paris, 17: 139-160 (1-24), pls 15-19 (1-5) (1928); 18: 1-56 (25-79),
pls 1-2 (6-7) (1929).

1931. Faunes senoniennes du nord et de l’ouest de Madagacar. Annis géol. Serv. Mines
Madagascar, Tananarive, 1 : 1-66, pls 1-9.

1936. Fossiles du Turonien supérieur d’Antanitiloky. Jn BrsairigE, H. Recherches
géologiques a Madagascar. Premiere suite. La géologie du Nord-Ouest. Mém. Acad.
malgache, Tananarive, 21 : 200-202, pl. 21.

1937. Les ammonites pyriteuses de l’Aptien d’Antanatanamirafy. Annis Paléont.,
Paris, 26 : 107-132, pls 16-18.

1938. Ammonites Campaniennes et Maastrichtiennes de l’ouest et du sud de Madagascar.
Annls géol. Serv. Mines Madagascar, Tananarive, 9 : 55-118, pls 1-9.

1949a. Faune néocomienne des couches a Criocevas de Belohasifaka (cercle de Sitampiky).
Annls géol. Serv. Mines Madagascar, Tananarive, 15 : 53-83, pls 8-13.

1949b. La faune albiene d’Ambarimaninga. Annis géol. Serv. Mines Madagascar,
Tananarive, 16: 1-128, pls 1-22.

1950. L’Albien de Mokaraha. Annis géol. Serv. Mines Madagascar, Tananarive,
17 : 57-85, pls 10-14.

1956. Ammonites néocrétacées du Menabe (Madagascar). IV.-—Les Phylloceratidae.
V.—LesGaudryceratidae. VI.—LesTetragonitidae. Avwnmnls géol. Serv. Mines Madagascar,
Tananarive, 23 : 1-106, pls I-11.

—— 1962. Allas des fossiles cavactéristiques de Madagascar (Ammonites). III (Berriasien,

Valanginien, Hauterivien, Barremien) : 96 pp, pls 176-214. Service Géologique, Tananarive.

1963. Atlas des fossiles cavactéristiques de Madagascar (Ammonites). X (Albien) :
xv+184 pp, pls 241-317. Service Géologique, Tananarive.

1964. Atlas des fossiles caractéristiques de Madagascar (Ammonites). XI (Cénomanien) :
x1+152 pp, pls 318-375. Service Géologique, Tananarive.

1966. Ailas des fossiles cavactéristiques de Madagascar (Ammonites). XIV (Santonien) :
x+134 pp, pls 455-513. Service Géologique, Tananarive.

Coguanp, H. 1880. Etudes supplémentaires sur la paléontologie algérienne. Bull. Acad.
Hippone, Bone, 15: 1-449.

Crick, G. C. 1907a. The Cephalopoda from the deposit at the North End of False Bay,

Zululand. Rep. geol. Surv. Natal Zululand, Pietermaritzburg, 3 : 163-234, pls 10-15.

1907b. The Cephalopoda from the tributaries of the Manuan Creek, Zululand. Rep.

geol. Surv. Natal Zululand, Pietermaritzburg, 3 : 235-249.

DrusHcuHiTtza, V. V. & KupRryAvtTsEvA, M. P. 1960. Atlas of Lower Cretaceous fauna of the
northern Caucasus and Crimea. 396 pp, 149 pls. Gostoptekhizdat, Moscow. [In Russian.]

376 CRETACEOUS FAUNAS

FABRE, S. 1940. Le crétacé supérieur de la Basse-Provence occidentale, 1. Cénomanien et
Turonien. Annis Fac. Sci. Marseille, Sér. 11, 14: 355 pp, Io pls.

Fattot, P. & TreRMIER, H. 1923. Ammonites nouvelles des [les Baléares. Tvab. Mus.
nac. Cienc. nat. Madr., ser. Geol. 32 : 1-83, pls 1-6.

ForBES, E. 1846. Report on the Cretaceous fossil invertebrates from southern India, collected
by Mr. Kaye and Mr. Cunliffe. Tvans. geol. Soc. Lond., (2) 7: 97-174, pls 7-19.

GaBB, W. M. & MEEk, F. B. 1864. Palaeontology of California, 1:xx+243 pp, 32 pls.
Geol. Surv. Calif., San Francisco.

HENDERSON, R. A. 1970. Ammonoidea from the Mata Series (Santonian-Maastrichtian) of
New Zealand. Spec. Pap. Palaeont., London, 6: 81 pp, 15 pls.

Jones, D. L. 1963. Upper Cretaceous (Campanian and Maastrichtian) ammonites from
southern Alaska. Prof. Pap. U.S. geol. Surv., Washington, 432 : 1-53, pls 1-41.

KarAKASH, N. I. 1907. Le Crétacé de la Crimeé et sa faune. Tvudy imp. S.-Peterb. Obshch.
Estest., 32 : 1-482, pls 1-28.

KENNEDY, W. J. & KLINGER, H. C. 1975. Cretaceous faunas from Zululand and Natal,
South Africa. Introduction, Stratigraphy. Bull. Br. Mus. nat. Hist. (Geol.), London,
25 (4) : 263-315, plate.

KLINGER, H. C., WIEDMANN, J. & KENNEDY, W. J. 1975. A new carinate phylloceratid
ammonite from the early Albian (Cretaceous) of Zululand, South Africa. Palaeontology,
London, 18 (3) : 657-664, pls 76-77.

KossmaT, F. 1895-98. Untersuchungen iiber die sudindische Kreideformation. Bevtr.
Paléont. Geol. Ost-Ung., Vienna and Leipzig, 9 (1895) : 97-203 (1-107), pls 15-25 (I-11) ;
11 (1897) : 1-46 (108-153), pls 1-8 (12-19); 12 (1898) : 89-152 (154-217), pls 14-19
(20-25).

McLEarn, D. 1972. Ammonoids of the Lower Cretaceous sandstone member of the Haida
Formation, Skidegate Inlet, Queen Charlotte Islands, western British Columbia. Bull.
geol. Surv. Can., Ottawa, 188 : 1-78, 45 pls.

Maumoun, I. G. EL-DIN 1955. Etudes paléontologiques sur la faune crétacique du massif du
Moghara (Sinai, Egypte). Publis Inst. Désert Egypte, Heliopolis, 8 : 1-192, 19 pls.

MARSHALL, P. 1926. The upper Cretaceous ammonites of New Zealand. Tyvans. Proc. N.Z.
Inst., Wellington, 56 : 129-210, pls 19-47.

Matsumoto, T. 1943. A short note on the Japanese Cretaceous Phylloceratidae. Proc.

imp. Acad. Japan, Tokyo, 18 (for 1942) : 674-676.

1953. The ontogeny of Metaplacenticeras subtilistviatum (Jimbo). Jap. J. Geol. Geogr.,
Tokyo, 23 : 139-150, pl. 13.

1954. Appendix. Selected Cretaceous leading ammonites in Hokkaido and Saghalien.
In Matsumoto, T. (Ed.) The Cretaceous System in the Japanese Islands : 243-313, pls
17-36. Jap. Soc. Promotion Sci., Tokyo.

1959. Cretaceous ammonites from the Upper Chitina Valley, Alaska. Mem. Fac. Sct.
Kyushu Univ. (D), 8: 49-90, pls 12-29.

MICHELIN, H. 1834. Ammonites velledae. Magasin Zool., Paris, 3: CIV, pl. 35.

NeEumAyR, M. 1885. Die geographische Verbreitung der Juraformation. Denkschr. Akad.
Wiss. Wien, 50: 57-142, pl. I.

OppEL, A. 1865. Die Tithonische Etage. Z. dt. geol. Ges., Berlin, 17 : 535-558.

ORBIGNY, A. D’. 1840-42. Paléontologie francaise. Tervains Crétacés I. Céphalopodes.

662 pp, 148 pls. Paris.

1850. Pvodvome de Paléontologie stratigraphique universelle des animaux mollusques et
vayonnés, 2. 427 pp. Paris.

Parona, C. F. & BoNARELLI, G. 1897. Fossili albani d’Escragnolles, del Nizzardo e della
Liguria occidentale. Palaeontogr. ital., Pisa, 2: 53-112, pls 10-14.

PERVINQUIERE, L. 1907. Carte géologique de la Tunisie. Etudes de paléontologie tunisienne,

1. Céphalopodes des terrains secondaires. 438 pp, 27 pls. Paris.

1910. Sur quelques ammonites du Crétacé algérien. Mém. Soc. géol. Fr. Paléont.,
Paris, 42 : 1-86, 7 pls.

ZULULAND AND NATAL 377

Rawson, P. F. 1966. A phylloceratid ammonite from the Speeton Clay (Lower Cretaceous)
of Yorkshire. Palaeontology, London, 9 : 455-457, pl. 72.

RoussEAu, L. 1842. Description des principaux fossiles de la Crimée. Jn DEMIDOFF, A. DE
1840-42. Voyage dans la Russie méridionale et la Crimée par la Hongrie, la Valachie et la
Moldavie (&c.) 2: 781-823, Atlas pls 1-12. Paris.

Sayn, G. 1891. Description des ammonites nouvelles ou peu connues du Crétacé inférieur

du sud-est de la France. Mem. Soc. géol. Fr., Paris, 17 : 679-688, pl. 17.

1901-07. Les ammonites pyriteuses des marnes valanginiennes du sud-est de la France.
Mém. Soc. géol. Fr., Paris, 23 : 1-66, pls 1-6.

1920. Les Phylloceras gargasiens du sud-est de la France (espéces nouvelles ou peu
connues). Jv KiL1an, W. Contribution a l’étude des céphalopodes paléocrétacés du
sud-est de la France. Mém. Serv. Carte géol. dét. France, Paris : 191-203, pl. I.

SHIMIZU, S. 1935. The Upper Cretaceous cephalopods of Japan, Part I. J. Shanghai Sci.
Inst., sect. 2, 1 : 159-226.

SIMONOVICH, S., BaTSEvicH, L. & Soroxin, A. 1877. Description of the geology of the
Kutais region etc. Mater. Geol. Kavk., Tiflis, 1876 : 1-112, pls 1-8.

SpatH, L. F. 1921a. On Cretaceous Cephalopoda from Zululand. Amn. S. Afr. Mus.,

Cape Town, 12: 217-321, pls 19-26.

192Ib. On Upper Cretaceous Ammonoidea from Pondoland. Aun. Durban Mus.,
3 : 39-57, pls 6, 7.

1922. On the Senonian ammonite fauna of Pondoland. Tvans. R. Soc. S. Afr., Cape
Town, 10: 113-147, pls 5-9.

1923-43. A monograph of the Ammonoidea of the Gault. Palaeontogr. Soc. (Monogr.),
London, 787 pp, 72 pls.

1925. On Upper Albian Ammonoidea from Portuguese East Africa, with an appendix
on Upper Cretaceous ammonites from Maputoland. Ann. Tvansv. Mus., Pretoria, 11:
179-200, pls 28-37.

1927. Revision of the Jurassic Cephalopod fauna of Kachh (Cutch). Mem. geol. Surv.
India Palaeont. indica, Calcutta, n.s. 11: 1-71, pls 1-7.

STOLICZKA, F. 1863-66. The fossil Cephalopoda of the Cretaceous rocks of southern India.
Mem. geol. Surv. India Palaeont. indica, Calcutta (3) 1: 41-56, pls 26-31 (1863); 2-5:
57-106, pls 32-54 (1864) ; 6-9: 107-154, pls 55-80 (1865) ; 10-13: 155-216, pls 81-94
(1866).

TaTE, R. 1865. Correlation of the Cretaceous Formations of the north-east of Ireland.
Q. Jl geol. Soc. Lond., 21: 15-44, pls 3-6.

Toucas, A. 1890. Etude de la faune des couches Tithonique de l’Ardéche. Bull. Soc. géol.
Fr., Paris, 18 : 560-630, pls 13-18.

UsHER, J. L. 1952. Ammonite faunas of the Upper Cretaceous of Vancouver Island, British
Columbia. Bull. geol. Surv. Can., Ottawa, 21 : 1-182, pls I-30.

vAN HoeEpEN, E. C. N. 1920. Description of some Cretaceous ammonites from Pondoland.
Ann. Transv. Mus., Pretoria, 7: 142-147, pls 24-26.

1921. Cretaceous Cephalopoda from Pondoland. Ibid., 8: 1-48, pls I-11.

VENZO, S. 1936. Cefalopodi del Cretacea medio-superiore dello Zululand. Palaeontogr.
ital., Pisa, 36 : 59-133, pls 5-12.

WIEDMANN, J. 1962a. Die systematische Stellung von Hypophylloceras Salfeld. Neues Jb.

Geol. Paldont. Abh., Stuttgart, 115 : 243-256, pl. 16.

1962b. Ammoniten aus der Vascogotischen Kreide (Nordspanien). 1. Phylloceratina,
Lytoceratina. Palaeontographica, Stuttgart, A 118 : 119-237, pls 8-14.

1964. Unterkreide-Ammoniten von Mallorca, 2. Lieferung: Phylloceratina. <Abh.
math.-naturw. Kl. Akad. Wiss. Mainz, Jahr. 1963 (4) : 157-264, pls 11-21.

1972. Neue Vorstellungen iiber Stammesgeschichte und System der Kreideammoniten.

Proc. int. palaeont. Un., int. geol. Congr. XXIII Sess., Czechoslovakia 1968 : 93-120,

ipl. Warsaw.

378 CRETACEOUS FAUNAS

WIEDMANN, J. & DiEn1,I. 1968. Die Kreide Sardiniens und ihre Cephalopoden. Palaeontogr.
ital., Pisa, 64: 1-171, 18 pls.

Woops, H. 1906. The Cretaceous fauna of Pondoland. Ann. S. Afr. Mus., Cape Town,
4: 275-350, pls 33-44.

Yokoyama, M. 1890. Versteinerungen aus der japanischen Kreide. Palaeontographica,
Stuttgart, 36 : 159-202, pls 18-25.

ZITTEL, K. A. 1868. Die Cephalopoden der Stramberger Schichten. Jn OppEt, A. Palaeont.
Mitth. Mus. k. Bayer. St. 2 (1): 1-118, pls 1-24. Stuttgart.

ZWIERZYCKI, J. 1914. Die Cephalopodenfauna der Tendaguru-Schichten in Deutsch-
Ostafrika. Avch. Biontol., Berlin, 3 (4) : 7-96, pls I-Io.

V. INDEX

The page numbers of the principal references are printed in bold type. An asterisk (*) denotes
a figure.

acknowledgements 374 Heterophylloceras 351

Ammonites baborensis 371 Hoplophylloceras 369
cylindricus 352 Hypophylloceras 369 &c.; see Phylloceras
partschi 369 aphrodite 358
vamMosUus 359 cottreaut 369
yvouyanus 371 sevesitense 304
SEVYUS 352 surya 369
subalpinus 355 Hyporbulites 354, 363
velledae 355, 300, 364 androtavense 305
zetes 352 betiert 352, 354

A phroditiceras 354; see Salfeldiella hoepeni 366-8, 368*

sevesitensis 364

Calaiceras 369 venaudiensis 3604

Carinophylloceras 372-3 woodst 366
collignont 350, 372, 372*, 373-4, 374* ; pls
ales location of specimens 350

Damesites 372-3

? Sp. Nov. 373
dimensions of specimens 351

Macrophylloceras 369
Menegheniceras 372

Epiphylloceras 351, 354, 303, 369, 371 Neophylloceras 351, 354, 363
mikobokense 368 hetonatense 368
vadiatum 369 lambertense 368
erratum 380 marshalli 368
Euphylloceras 354 neva 308
SEYUM 352 sevesitense 364

velledae 360
Paraphylloceras 354

field localities 351 Partschiceras 369
baborense 371-2
Geyeroceras see Phylloceras sp. juv. cf. baborense 350, 371-2; pl. 6,
Goretophylloceras 354 figs 2-3
besairier besairiei 370
Hantheniceras 351 - tsianalokensis 370

Harpophylloceras 372 bistriatum 370

INDEX

ezoense 370
forbestanum 370-1
inflatum 370
? japonicum 356, 358
marshallt 370
minimum 370
occlusum 370
umzambiense 350, 370-1, 371*;
fig. 2
whiteavest 370
zelandicum 370
Partschiphylloceras 369
Phyllocevas 351 ; see also Hyporbulites
androiavensis 365
aphrodite 355, 358
ellipticum 355-6, 358
escragnollensis 358
aff. grotht 358
heterophyllum 351
improvisum 350
japonicum 356
neva 306, 308; pl. 13, fig. 3
onoense 352, 354
cf. perlobatum 352
ex aff. vamosi 364
vobionense 352-4
sablyense 354
semistviatum 352
sevesitense tanit 364-5
Sevum 352, 354
perlobata 352-3
spatht 352-3
aff. surya 368
thetys 352, 354, 359
umzambiense 370
velledae 360-1, 364-5
inflatum 362
sevesitense 304
woodsi 366
sp. 360-1, 366
(Geyeroceras) 352
cylindricus 352
(Hypophylloceras) 352, 354-69
algeriense 306
androtavense 350, 365-6; pl. 4, fig. 7
aphrodite 350, 355, 358-9 ; pl. 3, fig. 3
aptiense applanatum 355, 358-9
aptiense 355, 358-9
aff. californicum 363
cypris cypris 359, 362, 366, 372
cythevae 359, 302, 366
fortunet 355, 358
hoepeni 368
infundibuliformis 368

pl. 13,

379

improvisum 355-6, 358
mikobokense 350, 368-9; pl. 12, fig. 1
morett 355, 358
neva 368
onoense 352, 359
pervinquieret 360
ponticult 359
sablyense 355
savdoum 355, 358
sevesitense 354,
pl. 4, fig. 6
boulet 363, 365
sevesitense 350, 363, 364-5;
fig. 4; pl. 7, fig. 4; pl.9
tanit 363, 365
subalpinum ellipticum 350, 354, 355-8,
35077 357, 309, pl4, “figs 1-4;
pl. 5, figs 2-3, > pl. 12, fig. 2
subalpinum 354, 357-8, 368
subseresitensé 355, 358
thetys diegot 355, 358
MAJOVICENSE 355, 358
thetys 354-5, 358-9
velledae 354, 359, 360, 362, 372; pl. 6,
fig. I
aschiltaense 359-60, 362
inflatum 350, 359-60, 362-3; pl. 13,
fig. 7
movelianum 359-60, 362, 372
velledae 350, 359, 360-2, 363, 372;
pl 4fig- 5, pl 5, fe: 4; pl 7,
figs 2-3; pls 8, Io, 11
sp. aff. velledae 350, 363; pl. 5, fig. 1;
pl.. 7, fig. 1
woodst 367-8
infundibuliformis 368
woodst 350, 366-8, 366*, 367*, 368* ;
pl. 13, figs 3-5
(Phylloceras) 351, 352-4
heterophyllum 352
sevum 350, 352-4, 353*; pls 1-2; pl. 3,
figs 1-2
spatht 352, 354
(Salfeldiella) guettardi 358
(Zetoceras) 352
ZeteS 352

359, 362, 363, 365;

ply G

Phyllocerataceae 351-74
Phylloceratidae 351-74
Phylloceratinae 351-74
Phyllopachyceras 369

umzambiense 370

Procliviceras 369
Pseudophylloceras 351

380 INDEX
Rhacoceras 351 Texanites oliveti Zone 368, 371

Xeniophylloceras 351
Salfeldiella (A phroditiceras) see Phylloceras
aphrodite venus 358 Zetocevas see Phylloceras

VI. ERRATUM

The following error has been noted in the first paper of this series (Bull. By. Mus. nat. Hist.
(Geol.), 25 (4) : 263-315) :
For ‘Besaire’ vead ‘Besairie’ throughout.

W. J. KENNEDY H. C. KLINGER

Dept of Geology & Mineralogy SouTH AFRICAN MuSEUM
UNIVERSITY OF OXFORD P.O. Box 61

Parks Roap CaPrE Town 8000

OxFORD OXI 3PR REPUBLIC OF SOUTH AFRICA
ENGLAND

Accepted for publication 12 February 1976

PARE x

Phylloceras (Phylloceras) serum (Oppel)
BM(NH) C78658, Makatini Formation, Barremian I, Loc. 170, Mlambongwenya Spruit,
northern Zululand. x1. (p. 352; see also Pl. 2 and PI. 3, figs 1-2.)

PATE vy

Bull. By. Mus. nat. Hist. (Geol.) 27, 5

30

PLATE 2

Phylloceras (Phylloceras) serum (Oppel)

Makatini Formation, Barremian I, Loc. 170, Mlambongwenya Spruit, northern Zululand.
(p. 352 ; see also Pl. 1 and Pl. 3, figs 1-2.)

Fic. 1a, b. BM(NH) C78660, x1.

Fic. 2a-g. BM(NH) C78659. a-d, x1. e-g, x2.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5

PLATE 2

PICA)

Phylloceras (Phylloceras) serum (Oppel)
Makatini Formation, Barremian I, Loc. 170, Mlambongwenya Spruit, northern Zululand.
<Is (ps 3523) scervalso: Piss =2-)
Fic. ta-c. BM(NH) C78657.
Fic. 2a-b. BM(NH) C78661.

Phylloceras (Hypophylloceras) aphrodite Fallot & Termier
Fic. 3. BM(NH) C78691, Mzinene Formation, Albian V, Loc. 51 (Bed 1), Mzinene River,
Zululand. x1. (p. 358.)

Bull. By. Mus. nat. Hist. (Geol.) 27, 5

PLATE 4

Phylloceras (Hypophylloceras) subalpinum ellipticum Kossmat
(p: 355. See also Pl5, figs 2—3 and: Pita hier see All 51

Fic. 1. BM(NH) C78665, Mzinene Formation, Albian V, Loc. 66, Munywana Creek
Zululand.

Fic. 2. BM(NH) C78664, Mzinene Formation, Albian V, Loc. 54, Munywana Creek, Zululand.

Fic. 3. BM(NH) C18139, figured Crick (1907a: pl. Io, figs 12, 12a), Mzinene Formation,
Skoenberg region on the Lower Mzinene River, Zululand, probably Cenomanian III.

Fic. 4. SAS A116, Mzinene Formation, Albian V, Loc. 67, Munywana Creek, Zululand.

Phylloceras (Hypophylloceras) velledae velledae (Michelin)
Fic. 5. -BM(NH) C78683, Mzinene Formation, Albian V, Loc. 51 (Bed 6), Mzinene River,
Zululand. -xt. (p. 360; seealsorwPl. 5, fig. 4; Pl. 6) fig: a Pl 7 ies 2-3 andericgersuo sic)

Phylloceras (Hypophylloceras) seresitense Pervinquiere
Fic. 6. BM(NH) C18138, figured Crick (1907a: pl. 10, figs 10-10a), Mzinene Formation,
Cenomanian, Skoenberg region, Mzinene River, Zululand. x1. (p. 364; see also Pl. 6,
iia LO TEAL a, sake, Al ehiauGleledls O),)

Phylloceras (Hypophylloceras) androiavense Besairie
’ Fic. 7. BM(NH) C78667, Mzinene Formation, Albian V, Loc. 54, Mzinene River, Zululand.
<a (ps 365¢)

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5

Pe Aa

Phylloceras (Hypophylloceras) sp. aff. velledae (Michelin)
Fic. 1. BM(NH) C78696, Mzinene Formation, Albian V, Loc. 51, Mzinene River, Zululand

(found loose)= Gin (pesOsie jseerdlsomelts7eedica sty)

Phylloceras (Hypophylloceras) subalpinum ellipticum Kossmat
(p. 355 ; see also PIA ies 1—4 andeiher2 tice, 2, sell <an-25 2
Fic. 2. BM(NH) C78663, Mzinene Formation, Albian V, Loc. 64 (Bed 4), Munywana Creek,
Zululand.
Fic. 3. BM(NH) C78062, Mzinene Formation, Albian V, Loc. 51 (Bed 6), Mzinene River,
Zululand.
Phylloceras (Hypophylloceras) velledae velledae (Michelin)

Fic. 4. BM(NH) C78672, Mzinene Formation, Albian V, Loc. 56, Mzinene River, Zululand.
xr. (p. 300; seealso Pl 4 fig. 530 RE oF ie. 1 ela etles 2 sya giaissS) LO, ary)

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5

PLATE 6

Phylloceras (Hypophylloceras) velledae (Michelin)
Fig. 1. SAS Ziroa, from Haughton’s (1936: 201, fig. 3) Moc. 19, <1. “(ps 300m meypical
form see also Pl. 4, fig.5; Pl. 5, fig. 4; Pl. 7, figs 2-3 and Pls 8, Io, 11.)

Partschiceras sp. juv. ? cf. baborense (Coquand)
Fies 2-3. BM(NH) C78693, Makatini Formation, Barremian I, Loc. 170, Mlambongwenya
Spruit, northern Zululand: j<1493a—3b >< 25 (p37)

Phylloceras (Hypophylloceras) seresitense seresitense Pervinquiere
Fic. 4. BM(NH) C18137, figured Crick (1907a: pl. Io, figs 11-11a), Mzinene Formation,
Cenomanian, Skoenberg Region, Mzinene River, Zululand. x1. (p. 364; see also Pl. 4, fig. 6;
PIE eatige 4s anid Ios)

Bull. By. Mus. nat. Hist. (Geol.) 27, 5 PLATE 6

PLATE 7

Phylloceras (Hypophylloceras) sp. aff. velledae (Michelin)

Fic. 1. BM(NH) C78695, Mzinene Formation, Albian V, Loc. 56, Mzinene River, Zululand.

x15 (pi sosF sec alsomel 5h sean)

Phylloceras (Hypophylloceras) velledae velledae (Michelin)

360; see also Pl. 4, fig. 5; Pl 5, fig. 4; Plo fig. 1 and Piss) ro; 11.) | bornener

(P:
Fic. 2. BM(NH) C78673, Mzinene Formation, Albian V, Loc. 56, Mzinene River, Zululand.
Fic. 3. BM(NH) C78684, Mzinene Formation, Albian V, Loc. 51 (Bed 12 or 13), Mzinene

River, Zululand.

Phylloceras (Hypophylloceras) seresitense seresitense Pervinquiere
BM(NH) C78690, Mzinene Formation, ? Cenomanian III, The Skoenberg, Mzinene

FIG. 4.
(p. 3645 “see also PL. 4; igo 7 Pino hie aaa: IIs o%)

River, Zululand. x1.

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5

IPIL ANTI, 3

Phylloceras (Hypophylloceras) velledae velledae Michelin
BM(NH) C78668, Mzinene Formation, Albian IV, Loc. 51 (Bed 1), Mzinene River, Zululand.
xo-9; actual diameter 208mm. (p. 3605 See also Pl 4) fig. 5; Pl5 fig. 43) Pe omien:
Pl. 7, figs 2-3 and Pls 10-11.)

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5 PLATE 8

PV AE

Phylloceras (Hypophylloceras) seresitense seresitense Pervinquiere
SAS Argo1, Mzinene Formation, Albian III, Loc. 36, Mzinene River, Zululand.
xo-74; actual diameter 190 mm. (p. 304; see also Pl. 4; fie. 6; (EINO) ne, 4 and estiee 4.)

Bull. By. Mus. nat. Hist. (Geoll\i2785

LeU IAD ID, 6)

PLATE to

Phylloceras (Hypophylloceras) velledae velledae (Michelin)
SAS Z1492a, Mzinene Formation, Albian V, Loc. 54, Mzinene River, Zululand.
x 0:86; actual diameter 162mm. (p. 360; see also Pl. 4, fig. 5; Pl. 5, fig. 4;
Pl. 6, fig. 1; Pl. 7, figs 2-3 and Pls 8, 11.)

Bull. By. Mus. nat. Hist. (Geol.) 27, 5 PLATE to

IRIE ANID, iti

Phylloceras (Hypophylloceras) velledae velledae (Michelin)
SAS Z1496, Mzinene Formation, Albian V, Mzinene River, Zululand.
x 0-85; actual diameter 169mm. (p. 360; see also Pl. 4, fig. 5; Pl. 5, fig. 4;
Plo fee. Pleo. nes 2-3" anid ISS som)

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5 IPI AN WIS, ie i

PLATE 12

Phylloceras (Hypophylloceras) mikobokense (Collignon)
FIG. “I, SAS H163c/7, St Lucia Formation, Maastnchtian 1, oc.) 20, Mielozimikaven,
/golkoleuaels, Sat, (70s BOS)

Phylloceras (Aypophylloceras) subalpinum ellipticum Kossmat
Fic. 2. BM(NH) C78666, Mzinene Formation, Cenomanian II or III, The Skoenberg,
Mzinene River, Zululand. x1. (p. 355; see also Pl. 4, figs 1-4 and Pl. 5, figs 2-3.)

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5 IPILAN IID,

PLATE 13

Phylloceras (Hypophylloceras) velledae inflatum Collignon
Fig. 1. SAS Z1742, Mzinene Formation, Albian IV or V, Mzinene River, Zululand. x1.
(pass62.)
Partschiceras umzambiense (van Hoepen)
Fic. 2. Holotype, TM 524, Umzamba Formation, Santonian or early Campanian, Loc. I,

mouth of Umzamba River, southern Natal (Pondoland). x1. (p. 370.)

Phylloceras (Hypophylloceras) woodsi woodsi van Hoepen
Umzamba Formation, Santonian or early Campanian, Loc. I,
mouth of Umzamba River, southern Natal (Pondoland). (p. 366.)
Fic. 3. Durban Museum specimen, mentioned by Spath (1921: 50) as Phyllocevas nera
(Forbes). x 1:
Ric. 4. Paratype, 2M 534, xz.
5. BLOolotype mis 3755 5<2:

PLATE 133

Il) Bay Bi

x€O

nat. Hist. (¢

Mus.

Bull. Br.

PLATS a4

Carinophylloceras collignoni Klinger, Wiedmann & Kennedy
Paratype, BM(NH) C78640, Mzinene Formation, Albian III, Loc. 36, Mzinene River,
Zululand. x1. (p. 373; see also Pl. 15.)

LAID. ae

Bull. By. Mus. nat. Hist. (Geol.) 27, 5

IGNITED) 1

Carinophylloceras collignoni Klinger, Wiedmann & Kennedy
Albian Ill, Mzimenetniver, Zululand: Both <1 (ps373) see alsomelapess)
Fig. 1. Paratype, BM(NH) C78644, Loc. 35 (Bed 2).
Fic. 2. Paratype, BM(NH) C78768, Loc. 36.

PLATE 15

Bull. Br. Mus. nat. Hist. (Geol.) 27, 5

Fieig iy

¥
=a Aw @.4

Ties

2s
%

Io.

cE:

A LIST OF SUPPLEMENTS
TO THE: GEOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

. Cox, L. R. Jurassic Bivalvia and Gastropoda from Tanganyika and Kenya.

Pp. 213; 30 Plates; 2 Text-figures. 1965. OUT OF PRINT.

. Ex-Nacear, Z. R. Stratigraphy and Planktonic Foraminifera of the Upper

Cretaceous—Lower Tertiary Succession in the Esna-Idfu Region, Nile Valley,
Egypt, U.A.R. Pp. 291; 23 Plates; 18 Text-figures. 1966. 11.

. DAVEY, R. J., DOWNIE, C., SARJEANT, W. A. S. & WiLLiAmMs, G. L. Studies on

Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 248; 28 Plates; 64 Text-
figures. 1966. £8.20.

. APPENDIX. DAVEY, R. J., DowntizE, C., SARJEANT, W. A. S. & WILLIAMS, G. L.

Appendix to Studies on Mesozoic and Cainozoic Dinoflagellate Cysts. Pp. 24.
1909. 95P.

. E..iott, G. F. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of

the Middle East. Pp. 111; 24 Plates; 16 Text-figures. 1968. OUT OF PRINT.

. Ruopes, F. H. T., Austin, R. L. & Drucs, E. C. British Avonian (Carboni-

ferous) Conodont faunas, and their value in local and continental correlation.
Pp. 313; 31 Plates; 92 Text-figures. 1969. £13.10.

. Cups, A. Upper Jurassic Rhynchonellid Brachiopods from Northwestern

Europe. Pp. 119; 12 Plates; 40 Text-figures. 1969. £5.25.

. Goopy, P. C. The relationships of certain Upper Cretaceous Teleosts with

special reference to the Myctophoids. Pp. 255; 102 Text-figures. 1969.
£7.70.

. OwEN, H.G. Middle Albian Stratigraphy in the Anglo-Paris Basin. Pp. 164;

3 Plates; 52 Text-figures. 1971. £7.20.

. SIppIguI, Q. A. Early Tertiary Ostracoda of the family Trachyleberididae

from West Pakistan. Pp. 98; 42 Plates; 7 Text-figures. 1971. {9.60.
Forey, P. L. A revision of the elopiform fishes, fossil and Recent. Pp. 222;
g2 Text-figures. 1973. {I1.35.

WitLiAMs, A. Ordovician Brachiopoda from the Shelve District, Shropshire.
Pp. 163; 28 Plates; 11 Text-figures; 110 Tables. 1974. £12.80.

Printedin Great Britain by John Wright and Sons Lid. at The Stonebridge Press, Bristol BS4 5NU

ne

Sem y RS OTS eee tp teot uur haven nc ae gunviesdnae eat eainittea re SNE SRS te a mr teens ayes ea NS INE BRE He CEE RS INR Le TOE

TEP LOR WP I EOWA RETO RENEE UN RCs MBE E ES Hath tas FILM,

eT)
Dy

et
a
bes

tas

“Y

cat

~

ban
hare

% “ 2 x a 7 wy = : a

SN Ns EE

——— - es

ms
ey

Pook Toren Paks

ate
Fe AA
Sjore xe
i *
aaa A RR
NSS

Pa ate