133845.pdf

40 JOHN C. W. COPE

the Amadeus Basin, Australia, by Pojeta &

Gilbert-Tomlinson (1977). The same genus

occurs in the Llanvirn Series of mid-Wales

(Cope 1999) showing that it was able to migrate

for large distances around the Gondwanan

margins to Avalonia, apparently quite indepen-

dently of water temperature. Eritropis may also

have reached Bolivia as Arca gracilis Hoek, 1912

(in Steinmann & Hoek 1912, pp. 248-249, p1. 8,

fig. 17) appears remarkably similar in external

morphology to Eritropis, but regrettably the

dentition is unknown. The only other interpre-

tation of this latter fossil is that it is a glyptarcoid

bivalve related to either Glyptarca or Hemipri-

onodonta. Despite the wide latitudinal tolerance

of Eritropis, the other eritropid genus

described by Pojeta & Gilbert-Tomlinson

(1977), Inaequidens, appears restricted to its

type area.

Trigonioida

The Trigonioida are a distinctive autolamelli-

branch group that can be distinguished from the

early Ordovician; they may be separated from

the heteroconchs by their type of dentition and

a short parivincular ligament with strong

nymphae and deep ligament grooves (Johnston

1996; Johnston & Zhang 1998); this type of

ligament is unlike that of the actinodonts, a

group of heteroconchs to which the trigonioids

have often been related (e.g. Pojeta 1978; Babin

1993; Waller 1998). Its earliest representative

appears to be Noradonta from the Early Arenig

of the Montagne Noire (Babin 1982a); this genus

was first described from the Nora Formation

(earliest Mid-Ordovician) of the Georgina

Basin, Australia, by Pojeta & Gilbert-Tomlinson

(1977) and this is again an example of a genus

that was able to migrate from the very high

latitudes of the Montagne Noire to the low

latitude of Australia. The dentition of

Noradonta has affinities with that of the cardio-

larioids, suggesting derivation from them (Cope

2000). Noradonta is a genus which forms a

suitable ancestor for Tromelinodonta (Babin

1982b) from the Late Arenig of Armorica. In

turn, Tromelinodonta, by loss of posterior teeth,

could produce the typical form of Lyrodesma

that has a radiating subumbonal sheaf of crenu-

late teeth. The earliest Lyrodesma recorded

hitherto is from the murchisoni Zone of the

Llanvirn Series of mid-Wales (Cope 1999).

Slightly younger ones are known from Spain

(Gutierrez-Marco & Babin 1999) and there are

many species known from the Late Ordo-

vician; these include forms from high-latitude

Gondwana such as the Armorican L. secure de

Tromelin & Lebesconte, 1876. or the Lyrodesma

described from Bolivia by Sanchez & Suarez-

Soruco (1996). Undescribed Avalonian forms

occur in both the Caradoc and Ashgill of Wales

and NW England (collections of the British

Geological Survey). Low-latitude Laurentian

forms include such species as L. majus (Ulrich,

1879) or the Baltican lyrodesmatid figured by

Neuman et al. (1997). Unlike contemporary

bivalves which favoured silty muds, lyrodes-

matids seem characteristic of the unstable

habitats represented by high-energy sands

(Cope 1999) and it is noteworthy that this is

similar to the habitat of Mesozoic trigoniids

(Stanley 1977). Post-Ordovician lyrodesmatids

were first reported from the Llandovery Series

of the Laurentian Silurian by Harrison &

Harrison (1975), but have subsequently been

found to occur as late as the Middle Devonian

(Pojeta & Zhang 1986). The ventrally flaring

crenulate teeth and dorsal musculature with

pedal retractors at the hinge-plate extremities

provide good evidence of a phylogenetic link

between Lyrodesma and the trigonioids. Upper

Palaeozoic trigonioids lack the dental crenula-

tion but have the same musculature as both the

lyrodesmatids and Mesozoic trigonioids (Newell

& Boyd 1975) and the juvenile dentition of

Lyrodesma was found to be very similar to the

schizodid dentition, suggesting paedomorphic

retention of this feature in the Upper Palaeozoic

schizodids (Harrison & Harrison 1975). Within

the Late Ordovician the lyrodesmatids gave

rise to the genus Pseudarca de Tromelin &

Lebesconte, 1875. For over a century this genus

was of doubtful affinities and was included in the

Nuculanoidea by Cox et al. (1969), although

Babin (1966) had suggested it was probably a

lyrodesmatid. Later, Babin (1987) described the

rediscovered type material, and, noting its

peculiar dentition with a diverging sheath of

short teeth, concluded that it was a lyrodes-

matid. Tunnicliff (1987) described material from

the Caradoc rocks of North Wales that included

a species of Pseudarca, previously unknown in

Britain, and independently came to the con-

clusion that its affinities were lyrodesmatid.

Brachilyrodesma Pojeta & Gilbert-Tomlinson,

1977 is a Mid- to Late Ordovician lyrodesmatid

from the Toko Group of the Georgina Basin,

Australia, unknown from elsewhere.

Heteroconchia

The Heteroconchia Hertwig, 1895 is a combi-

nation of the subclasses Palaeoheterodonta

Newell, 1965 and Heterodonta Neumayr, 1884.

but as used here excludes the Trigonioida which