Dr. E. H. Traquair — Fossil Fish from Borough Lee. 251 



Hemictenodus, Jaekel. 



In a recent paper 1 Dr. Otto Jaekel has proposed to institute the 

 genus Hemictenodus for C. obliquus of Atthey and a species from the 

 German Trias. This new genus he characterizes by the small 

 number of ridges on the teeth, as compared with those of such a form 

 as Ctenodus cristatus, and considers it to be intermediate between 

 Ctenodus and Ceratodus, in fact an illustration of the evolution of 

 the one genus from the other. Into the latter aspect of the question 

 I cannot follow Dr. Jaekel, as 1 consider the gap between Ctenodus 

 and Ceratodus, in spite of the opinions of Prof. Anton Fritsch, to 

 be far too wide to be bridged over simply by a comparison of shapes 

 of teeth. The consideration as to whether Ctenodus obliquus may 

 not be adopted as the type of a new genus is, however, quite another 

 matter. 



In briefly describing portions of cranial shields of Ctenodus from 

 the Northumbrian Coal-field, Messrs. Hancock and Atthey 2 drew 

 attention to the fact that in the species obliquus the median occipital 

 plate was concave in front, while in tuberculatus (= cristatus) it 

 "projects, and has a wedge-shaped process in the centre." And an 

 undoubted skull of Ct. cristatus, belonging to Mr. Ward, F.G.S., of 

 Longton, and figured in outline by Prof. A. Fritsch, shows that in 

 point of fact the median occipital plate was pointed in front, and 

 related to the neighbouring plates as in Dipterus, the anterior pair of 

 the three pairs of plates, which border it ou each side meeting in 

 the middle line in front of it. A different arrangement is, however, 

 shown in the skulls figured by Mr. T. P. Barkas 3 and Prof. Miall, 4 

 in which the median occipital has the anterior concavity as in the 

 specimens attributed to obliquus by Hancock and Atthey, and to 

 this concavity articulates a second median plate of considerable size, 

 which entirely separates the anterior pair of lateral bones from each 

 other. That the skulls showing this arrangement belong to a group 

 typified by C. obliquus there can be no doubt, any more than that 

 the other pattern is characteristic of cristatus and such allied forms 

 as interruptus. The question comes to be, — Is this distinction generic ? 

 I think so myself, though, as in other cases, it may not seem so to 

 other minds differently constituted. 



Hemictenodus quinquecostatus, Traquair. 



Ctenodus obliquus, var. quinquecostatus, Traquair, Geol. Mag. Dec. II. Vol. X. 



December, 1883. 



Owing to the general resemblance which the teeth of this species 

 bear to those of C. obliquus I originally described it as only a variety 

 of that form, but the constancy of the smaller number of ridges 

 induces me to elevate it to the rank of a species. In H. obliquus 

 the usual number is six to eight ; here the regular number is five, 

 with only rarely a rudimentary sixth one. 



The top of the skull shows the same general arrangement of 



1 Sitzungsb. der Gesellsch. naturforschender Freunde in Berlin, 1890. 



2 Ann. and Mag. Nat. Hist. (4) vii. 1871, p. 193. 



3 Manual of Coal-measure Palaeontology. 4 On some bones of Ctenodus. 



