6oo 



NATURAL SCIENCE. 



there were supporting elements of the nature of gill-arches. A pair 

 of >shaped impressions {(JU.) between these markings are exactly 

 such as might result from contact with ridges upon the auditory 

 capsules, due to a great development of the upper semicircular 

 canals, as in sharks. In short, it seems most probable that the shield 

 in question covered a truly fish-like head ; and if it were not for the 

 invariable absence of limbs and jaws, and the forward position of the 

 breathing apparatus in Pteraspis and its allies, these animals might 

 be placed, without hesitation, in the class of fishes. The possibility 

 that limbs and jaws were present, but not calcified enough to be 

 preserved, must, however, be borne in mind ; while the negative 

 evidence on this subject, and the want of information as to the nature 

 of the tail, are factors necessitating caution in determining affinities 

 at present. 



The next group of Ostracodernis (3,6, 12) is typified by the genus 

 Cephalaspis (Figs. 5, 10), and is remarkable for the manner in 

 which its members mimic in outward shape the contemporaneous 

 aquatic arachnids {Euryptcrus, &c.) The anterior dorsal shield (Fig. 5) 

 which was doubtless confined to the head and branchial region, 

 consists of a single piece, rounded or tapering in front, abruptly trun- 



FiG 9.— Restoration of rt.n7s/ir5. (From Brit. Mus. Catal. Foss. Fishes.) 



cated behind, and with the rounded margin reflexed inwards below to 

 form a flattened and ornamented rim. Like that of the Pteraspidians, 

 the substance of the shield consists of three superposed layers ; but 

 the inner layer is bony, the thick middle zone is soHd though traversed 

 by a coarse network of blood-vessels, and the outer surface is covered 

 with a dense tuberculated tooth-like tissue. The eyes (Fig. 5, orb.) are 

 placed close together in the middle of the shield ; and there is a large 

 " postorbital valley " {p.o.v.) behind them, really a vacuity in the 

 shield filled with a separate plate of pecuHar structure (see No. 14, p. 4). 

 There are no apertures for nostrils, which must thus have been 

 situated on the lower side of the head close to the mouth, and of this 

 nothing is as yet known. At the back of the shield, on each side, 

 occurs a small flap stiff"ened with little plates (Fig. 10) evidently to 

 be regarded as the operculum ; and immediately behind the shield 

 the ordinary scaling of the body begins, without any special develop- 

 ment for the support of limbs. Paired fins are, in fact, absent ; and 

 the Cephalaspidians possess only a system of median fins repre- 

 sented by a dorsal and a caudal, not stiffened by fin-rays but by little 

 elongated scales. The great deep quadrangular body-scales form a 

 series of interlocking rings, no doubt corresponding to the series of 

 muscle-plates of the trunk. 



