662 MR. A. SMITH WOODWARD OX THE SO-CALLED [Nov. 20, 



clavicle (s.cl.) and the arched clavicle (cl.) expanded at its lower 

 end. Posteriorly may also be seen the large scaim.la (sc), pierced 

 by an oval foramen. The pelvic, arch (tig. 6, plv.) is of the 

 ordinarv form, the two halves separate, each expanded at the 

 origin of the fin and tapering forwards. It is situated in a 

 remarkably advanced position, for the figured specimen seems to 

 show that the anterior fins represented in the type specimen 

 are truly the pelvic pair ; while each of these fins is very well 

 developed and comprises not less than eight rays l . The dorsal fin 

 (fig. 2, d.f.) occupies about the middle of the back and seems to 

 comprise two or three spinous rays anteriorly. 



The anal Jin is small and remote, consisting of robust rays 

 which are unjointed for a considerable length at their base. 



The scales are thick, cycloid, and very deeply overlapping : the 

 external markings are merely those produced by the concentric 

 lines of growth, but there are a few traces of radiating lines and 

 crimping in the covered portion. The peculiar longitudinal 

 elevations, described by Agassiz as marking the course of the 

 " lateral line," are apparently confined to the scales of the caudal 

 region. 



4. ACROGNATHUS BOOPS. 



Acrognatlws hoops, L. Agassiz, Poiss. Foss. vol. v. pt. i. p. 14, 

 pt. ii. p. 108, pi. lx.a. figs. 1-4 (1834-44); A. S. Woodward, Proc. 

 Geol. Assoc, vol. x. p. 323 (1888). 



No new evidence is forthcoming as to this supposed deep-sea 

 Salmonoid. It is represented in the British Museum only by the 

 type specimen and a fragment. 



Conchfsion. 



In determining the systematic position of these fishes from the 

 English Chalk it is, of course, impossible to refer to the most 

 distinctive external feature by which Salmonoids can be separated 

 from Clupeoids. The nature of the matrix would not admit of 

 the preservation of an adipose dorsal fin, even if it were originally 

 well-developed. Three osteological characters of Osmeroides and 

 Aulolepis, however, now made known for the first time, combine 

 to suggest comparisons only in one direction, namely, with the 

 modern genera Elo^is, Meyalops, and their extinct allies. These 

 characters are : — (i.) the union of the parietal bones mesially to the 

 exclusion of the supraoccipital from the cranial roof ; (ii.) the 

 arched maxilla overlapped above by two large supramaxillary 

 bones ; and (iii.) the presence of a large gular plate. It is true 

 that although in the typical Salmonidse the supraoccipital separates 

 the parietals on the cranial roof, there are rare instances (e.y., 

 ThymaUus) in which the parietals are in contact throughout their 



1 Since this paper was read the writer lias observed a specimen in the 

 Woodwardian Museum, Cambridge (Forbes-Young collection), in which the 

 pelvic fins are beautifully preserved in the advanced situation here described. 



