48 CARBONICOLA, ANTHRACOMYA, AND NAIADITES. 
attachment of the posterior end of the external hgament, as is the case in the 
recent Anodonta. 
The umbones of this species are seldom eroded. The only specimens in which 
I have seen this feature are from Codnor Park. One specimen from the Collection 
of Mr. Neilson, of Glasgow (see woodcut), appears to have been bored with a 
circular hole; and a similar hole is shown in one of Mr. Salter’s figures (op. supra 
cit.). To what these holes are due it is impossible to guess, for Gasteropoda 
have never been found in the beds with Carbonicola (see also p. 74). 

The pit for the reception of the cardinal tooth of the opposite valve is very 
large in the specimen figured, PI. II, fig. 1a, much larger than is necessary. The 
same feature may be observed in many of the valves on Pls. V and VI, and more 
particularly in the transverse section on Pl. XI, fig. 2. This raises a question 
as to the possibility of there having been present an internal ligament or cartilage 
lodged in this hollow, and the fact that striz can be made out would also point to 
this conclusion ; but this cartilage could not have been always present, as there 
is no room for it in many of the hinges figured on Pls. V and VI differing from 
that which I have considered to be the normal type. It is possible that this 
hollow lodged a process of the external ligament, which had here its anterior and 
strongest attachment. Whilst, however, on the one hand there are forms showing 
a shghter degree of development in the posterior end, there are others (Pl. II, 
figs. 1, 2), from Limerigg, Slamanuan, in the collection of Mr. Smith, of Kilwin- 
ning, which show a more elongated shape, with much smaller comparative dorso- 
ventral diameter ; but these have the characteristic constriction at the junction 
of the middle and posterior thirds of the valves, and the beaked posterior end. 
This species is the largest form of Carbonicola, and is very much larger than 
other members of the genus. In England the large form has a very limited 
distribution, both horizontally and locally. It is much more common in Scotland ; 
though, there, it appears to keep to one horizon. In the North Midlands of 
Kngland there is a dwarf form, but generally obtained crushed. 
The shape is very characteristic; the comparatively great dorso-ventral depth 
and the tumidity in the anterior half of the shell are of diagnostic importance. 
Salter considered the form to be identical with Unio rugosus (Pachyodon, Brown). 
