CARBONIFEROUS FISH-FAUNA, 
portion. The width of the exserted portion gradually increases 
to its maximum opposite the hinder part of the inserted base, and 
then the spine gradually tapers as it curves into the long and 
slender extremity, which is observable through a thin film of 
intractable matrix in the fossil represented in Pl. III, fig. 1. 
There is no evidence of the shortening of the spine by wear 
during life, such as occurs commonly in Gyracanthus. ‘The 
vertical compression of the spine is well seen in a fragment of the 
type-specimen, which indicates that just at the hinder end of the 
inserted base the maximum vertical diameter equals nearly one- 
third of the width, while the dorsal and ventral faces are about 
equally flattened. These two faces are also similar in their 
ornamentation, which consists of oblique ridges, all surmounted 
with a regular row of smooth, rounded tubercles (Pl. IV, fig. 1d), 
and separated by shallow, smooth grooves, which are wider than 
the ridges themselves. All the ridges are complete from edge to 
edge of the spine, and thirteen to fifteen of them are cut by a 
transverse section at the hinder end of the base. Their direction 
tends to become more and more longitudinal as they are traced 
from the base towards the apex of the spine, and they may have 
been absent on the terminal portion. ‘The anterior margin of the 
spine (PI. IV, fig. 10) is compressed to a sharp edge, along which 
the several ridges of the dorsal and ventral faces meet exactly in 
pais. The posterior margin (PI. IV, fig. le) is a narrow smooth 
area impressed by a deep longitudinal groove, which is repre- 
sented by a ridge of sandstone in the fossil-casts (Pl. II, figs. 2a, 
2e). It is evident that this groove was not quite median or 
bilaterally symmetrical. Its borders are not provided with any 
denticles. The calcified tissue of the spine is rather coarse, 
longitudinal vascular canals being visible to the naked eye, and 
imparting to longitudinal sections (PI. III, fig. 14) a fibrous 
appearance. ‘The internal cavity is large, and a remnant of it 
seems to extend to the apex of the spine. 
Anterior Free Pectoral Spine.—Vhere is no clear evidence of 
fin-supports or a pectoral arch at the base of the pectoral fin- 
spines; but there are two pairs of wide, hollow, triangular spines, 
which may have projected freely from the ventral aspect of the 
[6] 
