BRYOZOANS FROM CHATHAM ISLAND 
lato, related genera, and aspidostomatids. A perusal of a range of 
literature covering Late Cretaceous and Tertiary onychocellids and 
aspidostomatids (e.g., Brydone 1930; Voigt 1930, 1949, 1957, 1962, 
1967, 1968, 1975, 1979, 1981, 1983, 1985b, 1987, 1989; Berthelsen, 
1962) shows that a very wide range of colonial, zooidal, opesial, 
larval-brooding, and avicularian morphologies have been attributed 
to Onychocella alone. These include encrusting, erect uni- and 
bilamellar, and erect cylindrical and quadrilateral colony forms, 
broad subcircular to transversely narrow opesiae, nonexistent to 
vestigial to hyperstomial ovicells, and large and small, symmetrical 
to asymmetrical avicularia. Theoretically, heterochrony during 
zooidal ontogeny could account for all or most of this variation, but 
it is likely that, within the huge and bewildering array of species of 
Onychocella sensu lato, fine-grained statistical analysis would iden- 
tify a number of clades, some of which would correspond to the 
variety of generic concepts within the family. 
The type species of Onychocella is encrusting, the opesiae are 
relatively large and somewhat bell-shaped in outline, the ‘ovicells’ 
are endozooidal, and the avicularium is asymmetrical, with the 
rostral tip reflexed against the distolateral corner of an adjacent 
autozooid and the mandible unimembranous (i.e., on one side of the 
rachis only, the other half of the membrane being suppressed on the 
reflexed side of the rostrum). The nominal type species isOnychocella 
marioni Jullien, 1882, Recent, distributed from the Cape Verde 
Islands to France and the Mediterranean. Unquestionably this is a 
junior synonym of O. angulosa (Reuss, 1848), first described from 
the Austrian Tertiary (see Harmer 1926, and compare scanning 
electron micrographs of Recent (Hayward 1974) and fossil (Schmid 
1989) material). Schmid’s (1989) illustrations show more clearly 
than those of other authors that the avicularia have distinct, but 
weakly-developed mandibular pivots at the widest part of the 
avicularium. There is little information on the arrangement for larval 
brooding in O. angulosa. If it is the same as in Floridina levinseni 
Canu & Bassler (type species of Velumella Canu & Bassler, 1917), 
then there is no ovicell as such, just a cap-like overhang of the 
proximal end of the next zooid in the series (see Levinsen 1909: pl. 
24, fig.10). In the sum of their characters, Semieschara d’ Orbigny 
(type species S. flabellata d’Orbigny, 1852) and Rhagasostoma 
Koschinsky (type species R. hexagonum Koschinsky, 1885) accord 
with those of Onychocella as conservatively based on O. angulosa 
and these genera may be considered as synonyms. 
How stable are the above morphological characters? The form of 
the avicularium and mandible appear visually distinctive, but Hast- 
ings (1930) has shown remarkable variation in the form of the 
mandible, at least, in O. alula Hastings, confirmed by Cook (1985). 
In this species most avicularia are skeletally symmetrical and the 
majority of the mandibles straight but, surprisingly, untmembranous. 
Occasionally the rachis may be set obliquely with a slight develop- 
ment of a membrane opposite the larger membrane. In older parts of 
colonies the thin and fragile membrane is normally lacking alto- 
gether. The evidence from this species, then, is that symmetrical 
avicularia need not imply bimembranous mandibles (as inSmittipora 
Jullien, 1882 and Rectonychocella Canu & Bassler, 1917 (a prob- 
able synonym of Smittipora)). Skeletally, the distal deflection of the 
avicularian rostrum is a variable character also. In several Creta- 
ceous species attributed to Onychocella the avicularium may be 
symmetrical with no distal curvature, or have varying degrees of 
curvature, within the same colony (see for example, Voigt 1989). 
Mandibular pivots are lacking in many species attributed to 
Onychocella — it may be that the occurrence and shape of these is a 
useful character, along with the shape of the avicularian opesiae, but 
little study has been made. Frequently, published photographs and 
micrographs are not of adequate quality or magnification to give 
25 
detailed information on some avicularian characters. 
The presence of hyperstomial ovicells has been taken to be a 
‘good’ character. Thus their occurrence in Recent ‘Rectonychocella’ 
disjuncta Canu & Bassler, 1930 led Hayward (1974) to state that a 
new genus is required for this species. The presence of hyperstomial 
ovicells, and their consistent occurrence in erect cylindrical colonies 
in several species has allowed for the discrimination of Latereschara 
d’Orbigny, 1851 (Brydone 1930; Voigt 1959, 1967), which appears 
to be a ‘good’ genus. Cheethamia Shaw, 1967 also has hyperstomial 
ovicells but is encrusting. Scanning electron microscopy of the type 
species of both of these genera is needed to determine if they are 
congeneric. Semiescharellina @ Orbigny, 1852 (type species. mumia 
d’Orbigny, 1852) may also be a synonym. Voigt (1989: 54) used 
Cheethamia subgenerically for a species of Onychocella. 
The present species is excluded from Cheethamia in having 
transversely narrow opesiae with distinct opesiular indentations, 
well-developed articular ridges in the avicularium (lacking in 
Cheethamia), and a separate ovicell opening (not cucullate as in 
Cheethamia). Large subvicarious avicularia are not characteristic of 
Aspidostoma so the present species cannot be included in that genus 
— in the type species, A. giganteum, they are proportionately much 
smaller and interzooidal; they have similar articular ridges but a 
separated avicularian opesia (Hayward 1995). 
?Onychocella sp. Figs 60-63 
MATERIAL. IGNS BZ 210, a single eroded colony from Pukekio, 
Chatham Island. 
DESCRIPTION. Colony encrusting. Autozooids arranged quin- 
cuncially, mostly longer than wide; length = 0.44—0.62 mm, width = 
0.20—-0.22 mm, with the lateral margins and periopesial rim thick- 
ened and raised. Frontal cryptocyst of an even level below the rim, 
the opesia-orifice high-arched in regular autozooids, generally as 
wide as long (0.13—0.19 mm), proximal rim with rounded opesiular 
indentations at the corners. Avicularium subvicarious, occurring 
sporadically at the bifurcation of zooid rows but not every new zooid 
row starts with an avicularium; 0.48—0.50 mm long, the broad 
lingulate rostrum either rounded distally or subacute, with an exten- 
sive palate; an indentation occurring on each side approximately 
half-way along the avicularium; a single, elongate-oval medial 
foramen. Ovicells hyperstomial, the narrow transverse opening 
separated from the zooidal opesiae by the distal rim of the zooid and 
apparently unable to be closed by the zooidal operculum in life. 
REMARKS. As with the preceding species the generic and familial 
attributions of this species are uncertain. In both species the clear 
separation of the ovicellular opening from the opesia-orifice is not 
compatible with Onychocella sensu stricto. 
Genus INCERTAE SEDIS 
Onychocellid sp. not figured 
MATERIAL. IGNS BZ 211, asingle colony fragment from Pukekio, 
Chatham Island. 
DESCRIPTION. Colony encrusting? Autozooids large, exceedingly 
thick-walled. Measured from orifice to orifice at the colony surface 
the zooids are 0.73—0.92 mm long, but internally one zooid is only 
0.52 mm long between orifices whose length at the surface is 0.73 
mm. This appears to be a consequence of the convex colony surface 
in the sole fragment whereby, as the zooidal cryptocysts thickened 
frontally in life, the peristomial orifices became more separated. 
