4 Dr. R. H. Traquair—Homosteus and Coccosteus compared. 
into the possession of the Museum of Science and Art, Edinburgh, 
I propose to make this, the most perfect specimen of Homusieus 
which has ever been found, the text of the following remarks on 
the genus. 
Along with Homosteus it may, however, be as well to re-examine 
the structure of Coccosteus itself as a basis of comparison. 
The reading of the cranial buckler of Coccosteus is much com- 
plicated by the fact that certain superficial grooves belonging to the 
lateral-line system are very conspicuous and apt to be mistaken for 
sutures, while the true sutures are visible with difficulty, and can 
only be made out in exceptionally well-preserved examples. They 
seem, indeed, to have almost entirely escaped the observation of 
Agassiz, as the lines on the head indicated on his restoration of 
Coccosteus (“Old Red,” tab. 6, fig. 4) belong almost without 
exception to the lateral-line system. The figures given by Hugh 
Miller (Quart. Journ. Geol. Soc. 1859, p. 129, and “ Footprints,” 
Ist ed. fig. 11), in which he attempts to reduce the plates of the 
cephalic shield of Coecosteus to the same plan as that of the bones 
of the top of the head in the Cod, are much better, inasmuch as 
many of the true sutures are given ; but it is also too plain that he 
also looked upon the superficial grooves as indicating the real 
boundaries of the plates which he considered as the homologues of 
the cranial roof-bones in osseous fishes. Pander’s interpretation, 
aithough his figures give both sets of lines on the upper surface 
with considerable though not perfect accuracy, is correct only as 
regards the hinder part of the buckler, his reading of the anterior 
half being hopelessly wrong, and consequently his elaborate com- 
parison with the arrangement in Asterolepis falls to the ground. 
By far the most correct restoration of the cranial shield of Coccosteus 
is that of Huxley,’ in which he omits the superficial grooves alto- 
gether, and in which the only faults I can find are of omission, viz. 
the non-recognition of the median suture between the two central 
plates which he letters as “frontal,” and of the pair of premaxillary 
bones on each side of that median bone in front, to which he applies 
the name “ premaxilla.”’ 
In Pl. I. Fig. 2 I have given a sketch of the head of Coccosteus 
decipiens, Ag., the superficial grooves being given in dotted, the 
sutures in continuous lines, and as regards the names I have applied 
to the bony plates, I have thought it best to use as few as possible 
which might lead the reader to infer that I considered them the 
morphological equivalents of the cranial bones of ordinary fishes. 
Posteriorly we have the trapezoidal median occipital plate (m. o.), 
flanked on each side by the triangular external occipital (e.0.). In 
front of these are the two central plates (c.), external to which and 
forming the antero-external margin of the buckler are three plates, 
marginal (m.), post-orbital (pt. o.), and pre-orbital (p. o.), the two 
latter forming the upper margin of the orbit. The two pre-orbitals 
come together in the middle line posteriorly only for a very short 
distance, in front of which they are separated by a narrow oval 
1 Dec. Geol. Survey, x. p. 30. 
