﻿Yol. 64.] FROM THE CRETACEOUS FOEMA.TION OF BAHIA. 361 



of the mandible. The opercular apparatus is narrow, and there is 

 space for an expanded preoperculum, of which part can be seen. 



The vertebraB are about 40 in number, half referable to the 

 caudal region. The centra are much-constricted cylinders, scarcely 

 longer than deep, and slightly strengthened by a few longitudinal 

 ridges. The ribs are long and stout, each impressed with a 

 longitudinal groove. The separate neural spines immediately 

 behind the head are expanded into narrow laminae. A few inter- 

 muscular bones are seen in the abdominal region. 



In all the fins except the caudal the rays are slender, with a long 

 undivided basal portion. In each pectoral fin eight rays can be 

 counted. The pelvic fins are opposed to the dorsal, and in the 

 type-specimen they appear to be not much smaller than the 

 pectorals. The dorsal fin in the second specimen is observed to 

 comprise twelve rays. Its foremost support is expanded into a 

 large wing, and the following six supports are also more or less 

 winged. The anal fin in the same specimen arises at least as near 

 to the pelvic fins as to the caudal. It is quite small and short, 

 with only nine supports ; extending along the long interval behind 

 it there is, however, a spaced series of seven thick, shining scales, 

 which are of uncertain shape but have the appearance of ridge- 

 scales accidentally displaced upwards. Although not connected 

 with any fin-supports or fin-rays, they suggest the thickened ven- 

 tral ridge-scales associated with the anal finlets in Scomhroclwpea. 

 The caudal fin is forked, and its relatively stout rays are articulated 

 nearly to the base. 



Scales are not preserved in the type, but traces of thin cycloid 

 scales are seen on the tail of the second specimen. One very large 

 ventral ridge-scale (PL XLIII, fig. 4, r) is also observable in the 

 front part of the abdominal region in the latter specimen, its hinder 

 prominence (turned forwards in the fossil) being sharply pointed. 



The fish thus described is obviously a Clupeoid, not differing 

 much from Glupea itself; and its most striking peculiarity is the 

 spaced row of small thickened scales between the anal and the 

 caudal fins. These scales, as already mentioned, suggest the anal 

 finlets of the European Cretaceous Scomhrodupea ; and, if they do 

 not actually mark the points of attachment of such finlets, they 

 seem to represent the little thickened scales which accompany 

 them in the genus just mentioned. I therefore name this new 

 Brazilian fish Scomhroclupea scut at a. It is distinguished 

 from the two definitely-known species ^ not only by its anal ridge- 

 scales, but also by the number of the rays in its fins. 



There is fragmentary evidence of at least one other ganoid fish 

 in Prof. Ennes de Souza's collection, but the specimens described 

 in the foregoing pages are all that admit of satisfactory determina- 

 tion. These are enough to prove that the shales at Ilheos belong 

 to the same Lower Cretaceous Series as those in the neighbourhood 

 of Bahia itself ; and they indicate the same mingling of typically 



1 ' Catal. Foss. Fishes Brit. Mus.' pt. iv (1901) pp. 135-38. 



