THE OLD RED SANDSTONE. - 49 
shoulders, one of the legs also wanting, and the arms spread 
to the full. The figure of the Coccosteus I would compare 
to a boy’s kite. (See Plate III., fig. 1.) There is a rounded 
head, a triangular body, a long tail attached to the apex of 
the triangle, and arms thin and rounded where they attach to 
the body, and spreading out towards their termination like 
the ancient one-sided shovel which we see sculptured on old 
tombstones, or the rudder of an ancient galley.* The 
manner in which the plates are arranged on the head is 
peculiarly beautiful; but I am afraid I cannot adequately 
describe them. A ring of plates, like the ring-stones of an 
arch, runs along what may be called the hoop of the kite ; 
the form of the keystone-plate is perfect ; the shapes of the 
others are elegantly varied, as if for ornament; and what 
would be otherwise the opening of the arch, is filled up with 
one large plate, of an outline singularly elegant. A single 
plate, still larger than any of the others, covers the greater 
part of the creature’s triangular body, to the shape of which 
it nearly conforms. It rises saddle-wise towards the centre :. 
on the ridge there is a longitudinal groove ending in a _perfo- 
ration, a little over the apex, (Plate III., fig. 2 ;) two small lat- 
eral plates on either side fill up the base of the angle ; and 
the long tail, with its numerous vertebral joints, terminates the 
figure. 
Does the reader possess a copy of Lyell’s lately published 
elementary work, edition 1838? If so, let him first turn up 
the description of the Upper Silurian rocks, from Murchison, 
which occurs in page 459, and mark the form of the trilobite 
Asaphus caudatus, a fossil of the Wenlock formation. (See 
* I have since ascertained that these seeming arms or paddles were 
simply plates of a peculiar form. (See Plate IX.) 
