2 Mr. A. S. Woodward on iico 



Diphmystus longicostatus, Cope. (PI. I. fig. 1.) 



The type specimen of this species, now in the collection of 

 Professor Cope, Philadelphia, was discovered by Mr. Joseph 

 Mawson, F.G.S., in the Cretaceous of Itacaranha, near 

 Bahia, Brazil. Other fragments were met with at the same 

 time, and some presented to the British Museum * ; but it 

 was not until quite lately that our national collection became 

 enriched by another complete specimen of the fish, discovered 

 by the same indefatigable explorer of the Brazilian Cretaceous. 

 This fossil, which seems to be imperfect in the anterior 

 dorsal region, is shown of the natural size in PI. I. fig. 1, 

 and partly confirms, partly extends, Cope's original description 

 of the species. 



The general form and proportions of the fish are seen to 

 be exactly as already described, while the numbers of the 

 dorsal and anal fin-rays have been correctly given. Now 

 for the first time, however, the pelvic fins are indicated, of 

 very small size, arising opposite the origin of the dorsal, and 

 the pectoral fins are not placed quite so high upon the flank 

 as in the type specimen. The skeleton is obscured in places 

 by a thin film of matrix ; but it is clear that there were about 

 twenty-four abdominal vertebrae, as stated by Cope, while the 

 number of caudals is twelve (not ten). The long ribs extend 

 to the ventral border, and the hindermost five or six pairs are 

 attached to rapidly elongating processes, tlie ribs shortening 

 in proportion as the latter increase. The neural arches are 

 all fused with the vertebral centra; but in the anterior caudals 

 the line of union between each neural arch and its spine can 

 be distinguished. There are also traces of intermuscular 

 bones in the dorsal region. The dorsal and anal fins are 

 imperfect, but the slender caudal lobes are especially well 

 shown. The dorsal ridge-scales are also obscured or destroyed 

 and the contour of the back apparently distorted, but the 

 characteristic ventral ridge-scales appear to be more satisfac- 

 torily displayed than previously. In the anterior half of the 

 abdomen these scales are of normal size ; but immediately 

 behind the pelvic fins they begin to increase, and the three 

 hindermost scales are relatively gigantic. Each exhibits a 

 long posteriorly directed point, and is apparently smooth. 



The new example of I), longicostatus thus described was 

 obtained by Mr. Mawson from the beach between Itacaranha 

 and Plataforma ; and to complete our knowledge of the 

 species it now only remains to discover the dentition and the 

 squamation of the flank. 



* A. S. Woodward, "Notes on some Vertebrate Fossils from the 

 Province of Bahia, Brazil, collected by Joseph Mawson, Esq., F.G.S.," 

 Ann. & Mag. Nat. Hist. [(3J vol. ii. (1888j p. 132. 



