Fossil Fishes of the Upper Old Red Sandstone. 38] 
Cosmacanthus Malcolmsoni, Ag. 
[Plate X., Figs. 2 and 3.] 
Agassiz, Poiss, Foss. v. g. r., p. 120, pl. 33, fig. 8; Traquair, Ext. 
Verteb. Moray Firth Area, p. 263, pl. vi., figs. 6-9. 
In my “Extinct Vertebrata,’ I showed not only that 
Agassiz was right in referring the original fragment of 
Cosmacanthus to the category of Selachian spines, but that 
it was not bilaterally symmetrical, and consequently, like 
Gyracanthus, was pretty certainly a lateral appendage. 
I am now able to repeat those statements, and to complete 
the description of Cosmacanthus Malcolmsoni, from a beautiful 
and nearly entire example contained in the Brickenden 
Collection in the British Museum (P. 8298). The spine 
(Pl. X., Figs. 2 and 3) measures 34 inches in length, and 
has the point nearly complete; in its form it is compressed, 
but one side (Fig. 3) is more convex than the other (Fig. 2). 
The anterior margin is rounded; on the posterior aspect is 
seen the sulcus, extending to about 14 inch from the point. 
The covered or inserted part of the lateral surface, behind, is 
broader on the flattened than on the rounded side of the 
spine, and is on both sides sharply defined from the sculptured 
surface; and as the sculptured surface is cut off at the basal 
end, it is clear that the inserted extremity has been lost. .The 
exposed surface is covered with stellate tubercles, larger and 
more crowded on the front of the spine, smaller and not so 
closely placed on the sides ; though there is some tendency 
to a linear arrangement of these tubercles on the sides, there 
is no longitudinal interval along the anterior margin, as in 
the other two specimens which are known. A posterior 
marginal row of large tubercles, sixteen in number, occupies 
the place of the marginal denticles of Gyracanthus, being 
placed along the hinder edge of the flat surface of the 
spine. 
The present specimen, nearly entire as it is, in this way 
completely confirms the evidence of the fragment which I 
figured in my previous memoir, namely, that Cosmacanthus 
