Description of Paleeospondylus Gunni. 89 
It must first be noted that all the specimens seem to lie 
on their backs in the stone, the ventral aspect of the head 
only being exposed: where the vertebral column is well 
preserved, this can be verified by observing the relation of 
the neural arches to the centra, as shortly behind the head 
the column is ordinarily twisted so as to lie on its side 
(Woodcut, p. 90). These facts indicate that the head not 
only was flattened, but that the upper or dorsal aspect had 
irregularities causing it to adhere more closely to the matrix. 
Secondly, we must bear in mind that almost every specimen 
which has occurred has the skull in a more or less abraded 
condition. The specimens are not conspicuous except upon 
weathered surfaces, and then it will be found that the 
weathering has eaten more or less through the surface of 
the head, so as to produce an appearance naturally subject 
to some variation in different individuals. The substance 
of which the remains consists is black, and of the appear- 
ance of jet; I have not examined it microscopically} but 
under a simple lens it presents no trace of structure. 
Fig. 2, Pl. 1, represents the most uninjured head which I 
have ever obtained, and which is here, as usual, looked at 
from the ventral surface. It is oblong, divided by a con- 
striction on each side into anterior and posterior parts, each 
of a somewhat hexagonal shape. The anterior part (Woodcut, 
t.p.) has a narrow anterior margin, and a broader posterior one 
where it joins the hinder division; on each side there are two 
lateral margins, joining each other in an obtuse external angle. 
The surface of this division of the cranium is gently convex, 
with a median longitudinal depression or furrow, the edges of 
which are somewhat prominent. The surface on each side is 
not smooth, but shows several pitted markings radiating some- 
what from each external angle. Each antero-lateral angle is 
produced into an anteriorly directed pointed feeler-like process, 
reminding us indeed of the prepalatine process of Myzxine, 
but an examination of other specimens (see Figs. 1 and 4) 
shows that there were at least. two other similar but shorter 
1 The fish remains, in general, which occur in the Caithness flags, are 
quite useless for microscopic work, owing to the tissue being saturated with 
opaque black bituminous matter. 
