FOSSIL PISHES. 31 



maximum thickness of about one inch. It projects beyond the curved 

 border from six to eight inches in a flattened neck. 



As has been before stated, the dorsal shield of D. Hertzeri is, as yet, 

 imperfectly known. Two incomplete specimens, which I have, indicate 

 that it was of nearly the same general size and form as that of D. TerreUi; 

 but the neck-like projection is relatively much shorter, as though cut off, 

 obliquely, from above. 



The Sujpra^soapulas (or Post Temporals) of Dinichthys TerreUi are 

 somewhat imperfectly represented in Figs. 1, la, 2, 2a, of Plate 34, Vol. I. 

 They are flattened, triangular, or trapezoidal bones, about one foot in 

 length by eight inches in width at the broadest part. They are thickest 

 in the middle, where the exposed portion is comparatively small, and thin 

 off on either side, where they are overlapped by other plates. ISTear the 

 center of the thickest border a strong, flattened condyle is obliquely set, 

 which fits into a deep cavity in the os articulare capitis. 



The exposed portion of the Supra-scapulas, like the cranial surface, is 

 marked by simple, linear furrows, which form some large pattern of orna- 

 mentation, as yet not fully made out. 



The Plastron or ventral shield of Dinichthys Hertzeri is represented 

 nearly complete, and of the natural size on Chart VI. The inside of the 

 bones composing it is there shown. In the smaller diagram on the same 

 chart, these bones are represented in position and seen from the outside. 

 This shield is composed of five flat bones ; two pairs and an elongated cen- 

 tral one, which is interposed between them. The central plate is some- 

 what overlapped by the lateral ones, and when all are in their normal posi- 

 tions, the shield is twenty inches wide, and about three feet in length. 

 The exposed surfaces of these bones are granulated, like those of the head, 

 and the hinder pair are marked with the peculiar linear furrows seen on 

 the bones of the cranium and the Supra-scapulas. The homologies of 

 the bones of the ventral shield, and their correspondence with those of the 

 ventral shield of Coccosteus have been discussed in the general descrip- 

 tion. The exterior margins of the anterior pair of plates (pre-ventro late- 

 rals, of Owen) show contact with other plates, and it is almost certain that 

 the sides of the body between the dorsal shield and Supra-scapulas, and 

 the ventral shield, were defended by plates of some kind, but none have 

 yet been found that can certainly be referred to this position. Among 

 the specimens collected at Delaware by Mr. Hertzer, and appertaining to 

 D. Hertzeri, is one imperfect triangular plate, nearly three feet in length, 

 and one foot wide at its broadest end. This, I have thought, might have 

 occupied the side of the body, as there seemed no otlier place for it, but its 

 location is yet only conjectural. lS.o such plate has been found entire in 

 connection with the remains of D. TerreUi, but a large number of frag- 

 3 



