188 



of closely-appressed osseous rays of different widths, each of which is com- 

 posed of narrow, oblique segments. The sutures of these segments are of 

 different character in ahiiost every ray : being in the marginal ray en chevron ; 

 in others, step-like ; and, in others, dovetailed (see p. 200, and Plate XLIV, 

 fig. 4). I suppose this compound spine to belong to one or other of the 

 borders of the caudal fin. 



The vertebrae in all the species certainly assignable to this group are, 

 where known, deeply two-grooved on each side, besides the pits for the inser- 

 tion of neurapophyses and pleurapophyses, except in the cervical region, 

 where the lateral grooves are wanting. There are no diapophyses. The 

 caudal vertebrae are rather numerous, but not so much so as in Amia, nor are 

 they so much recurved as in that genus. 



Affinities of the Saurodontidce. — More perfect specimens received since 

 the description of the cranial structure on p. 183 was printed, render it 

 almost certain that the median bone of the superior cranial walls is a supra- 

 occipital; that the parietals are produced upward into an angle (epiotic, p. 

 183) on each side; and that the epiotics (opisthotic, p. 183) form the postero- 

 lateral angles of the skull. If this interpretation be true, there is no opis- 

 thotic bone. It is quite possible to interpret the superior cranial structure 

 of the Siluroids in the same way, while the arrangement is very different 

 from that seen in Salmonidce, Cijprinidce, Characinidce, and Esocidce, where 

 the opisthotic is present, and where the supraoccipital does not present any. 

 such anterior prolongation. The structure of the scapular arch, if we except 

 the position of the scapula, has much in common with that of the Siluroids; 

 while the two basilars and double articulation of the pectoral spine are 

 striking points of resemblance to the same group. As characters of more 

 typical Physostoini, we have the maxillary arcade of the mouth, the form and 

 position of tlie ventral fins, and the apparent absence of dorsal spines. The 

 form of the bones of the mouth, the presence of symplectic, subopercular, 

 and postfrontal bones, the basioccipital muscular tube, and the unmodified 

 anterior vertebrae, distinguish the Saurodontidce widely IVom the Nema- 

 tognathi, and ally them to the Isospondyli; and with the latter I have for the 

 present allowed them to remain. It must not be forgotten, however, that 

 the long supraoccipital and osseous vientral rays arc physoclystous charac- 

 ters, and that the form and position of the femora arc much nearer to those 

 of i?(?/o«e than to those of any jjliysostomous fish known tome. The three 

 persistent vertcbr.ie of the caudal fin rec&W Snlmo, while (he dorsals nearly 

 resemble those of the Clupeidce. The peculiar malleolus of the palatine bone 

 is closely imitated by the Ponwlobus iJS&udoharengus (alewife) ; but in that 



