CLASSIFICATION OF DEVONIAN FISHES 427 
exhibited no ornamentation. Their inferior faces were also smooth, 
and formed by a very delicate layer of bone. The mass of the 
scale was formed by an osseous and spongy substance, adorned 
with fine rays, which radiated from the centre of the scale. These 
rays were intersected by concentric and circular lines, so that a 
scale, whose smooth layer is worn away, presents a number of small, 
elongated cells, disposed in circular series, almost like the seats of an 
amphitheatre. 
In the “ Additions et Corrections,” (1. c. p. 140), Agassiz adds a 
description of the scales of Glyptolepis elegans, which supplies 
an important correction to that just given. Referring to Tab. 21a 
he says, “ The figure 2 a represents a scale of its natural size, and 
“fig. 2 the same magnified. The folds of the surface which con- 
“stitute the ornantents of the visible portion of the scale are more 
“distant than in the foregoing species.” Thus it is admitted that 
Glyptolepis has not smooth, but sculptured scales, as, indeed, the 
name of the genus implies. 
Agassiz enumerates a caudal, two dorsal, and two anal fins, but 
states that the existence of pectorals is doubtful, and that, in any 
case, they must have been small and inconsiderable. The ventral 
fins, on the other hand, are said to possess a singular structure, 
“which is also to be found in Megalichthys.’” A series of plates 
extends as a pointed band along the belly, and, becoming free at its 
posterior extremity, carries numerous rays on both sides, and thus. 
forms a ventral fin, which, from the manner in which its rays are 
disposed, is very like an eel’s tail. (Tab. 21, fig. 2.) I have ex- 
amined the specimen here referred to, which forms a part of Sir 
Philip Egerton’s collection, and, with Professor Pander, I feel satis- 
fied that the fin in question is the very long, acutely lobate, pectoral, 
bent back in such a manner, that the proximal half of its posterior 
edge is covered by the lower margin of the abdomen of the fish. 
Professor Agassiz goes on to say that the two dorsals are opposed 
to the two anals, and are situated so far back that the caudal 
directly follows them. They are so close together that the last ray 
of the first touches the first ray of the second. The second dorsal 
and anal are higher than the first, and the caudal is large, hetero- 
cercal, and triangular, appearing to be almost vertically truncated ; 
its superior division -bears numerous little fulcra. 
Hugh Miller (‘Old Red Sandstone,” 1841,) made some important 
improvements upon Agassiz’ description and definition of Glyptolepis. 
He pointed out with great justice (and figured a specimen demon- 
strating the fact), that there is only one anal, the second, or posterior 
