2 Proceedings of the Royal Physical Soerety. 
vertebre on cach side, that a deceptive conclusion — is 
suggested as to their being continuous in the middle line 
over that aspect of the axis, the dorsal one, which is 
attached to the matrix (as in Fig, 2). 
1 must confess that I find it, for the present, very difficult 
to put an interpretation upon this curious paired structure, 
and | must own also that it is possible that in the eyes of 
some if may appear to indicate a rudiment or remnant of a 
shoulder-girdle—an idea which would be somewhat fatal to the 
Marsipobranch theory of this organism, But if this piece on 
cach side appertains to a pectoral arch, it can only, from. its 
position, represent either a supra-clavicular or a post-temporal 
element, which would indicate a condition somewhat incon- 
sistent with the otherwise very primitive structure of 
Paleospondylus. It seems to me quite possible that these 
two pieces may have been someway concerned in the 
support of the branchial apparatus, 
The vertebral column is rarely well preserved ; in fact, in 
the majority of specimens, it appears merely as a bituminous 
streak, with only faint indications of transverse segmentation, 
It is well shown in the specimen represented in Fig. 1, and in 
Mie, 8 T have represented, on. a more magnified scale, the 
vertebral centra immediately behind the head in another 
example, These most anterior centra are rather shorter than 
broad, but farther back their length equals their breadth, 
Mxternally they are marked by several longitudinal grooves, 
of which one is median and ventral. Internally they are 
undoubtedly hollow, and are genuine ring-vertebre, the 
hollow tubular interior being often opened into and shown 
by erosion, as is the case in the last vertebra, that on the left, 
represented in Fig, 3. The dorsal aspect of the vertebrae 
immediately behind the head is never shown, but so soon as 
the column gets twisted round on its side“it may be seen 
that each centrum is surmounted by a somewhat square- 
shaped newral piece or arch. About the posterior two-thirds 
of the column these neural pieces become elongated into 
slender and very oblique spines, producing a fin-like eleva- 
tion, which again falls away towards the point of the tail. 
A similar development of heemal spines occurs at the same 
