A. Smith Woodward — On the Genus Notidanus. 257 



Such is a brief enumeration of the various specific types of fossil 

 Notidanidae that have hitherto come under my notice. One other 

 form has also been assigned to this genus by Winkler, 1 under the 

 name of N. Orpiensis, but this is apparently an erroneous determina- 

 tion : the type-specimens are from the Thanet Sand (Heersien) of 

 Orp-le-Grand, Belgium, but only one (I.e. fig. 15) seems referable to 

 Notidanus, and that is too imperfect to name specifically. 



And now, in conclusion, there are one or two general considera- 

 tions suggested by the foregoing facts to which it may be interesting 

 briefly to refer. On examining the various forms of teeth in strati- 

 graphical order, it is at once evident that there are distinct traces of 

 specialization. In the earliest Jurassic types hitherto known, with 

 a single exception (N. serratus), the secondary cones are only two, 

 three, or four in number, while most members of the genus of Cre- 

 taceous and Tertiary age possess at least six or seven in the lower 

 teeth, and one Pliocene species (N. anomale) has been recorded with 

 no less than fourteen : it is also noteworthy that some of the Jurassic 

 teeth are characterized by the absence of anterior serrations on the 

 principal cone. Again, it is interesting to observe that the very deep 

 compressed fibrous root of the more recent Notidanidae only began 

 to assume its peculiar characters towards later Cretaceous times, the 

 base in such species as N. serratus and N Daviesii being compara- 

 tively thick and depressed, and very suggestive of that of Hybodus. 

 The earliest species of Notidanus, indeed, have teeth so remarkably 

 similar to those of certain Hybodonts, that not only did Phillips — 

 as already stated — figure two specimens under the name of Hybodus 

 polyprion, but at least one other palaeontologist 2 appears to have 

 made the reverse mistake. A careful study of a large series of 

 specimens of the well-known Stonesfield species results in the dis- 

 covery of some teeth that may almost be regarded as links between the 

 two. The originals of Figs. 1, 2, PI. VI., for example, are such forms 

 as would be commonly associated with Hybodus 'polyprion, but a very 

 slight modification is necessary to convert them into most typical 

 Notidanidae, and I have considerable doubts as to the propriety of 

 placing them with Hybodus at all. To transform a long-coned 

 Hybodont tooth into one of Notidanus, it is only necessary to assume 

 lateral compression and the reduction or loss of the anterior secondary 

 cones, and the teeth represented in Figs. 1 and 2, PI. VI., exhibit a 

 marked approach to this condition. Fig. 1 shows a large, compressed 

 principal cone, preceded by four secondary cones, and followed by 

 two others, the former being reduced to quite small denticles, while 

 the latter are comparatively well developed : the front edge of the 

 principal cone is distinctly denticulated for half its length, and so 

 also is the first secondary ; and there are traces of vertical wrinkles 



1 T. C. Winkler, " Memoire sur quelques Eestes de Poissonsdu Systeme Heersien," 

 Archiv. Mus. Teyler, vol. iv. fasc. i. pp. 12, 13, figs. 13-17. 



3 E. Favre, "Description des Fossiles du Terrain Oxfordien des Alpes Tribour- 

 geoises," Mem. Soc. Paleont. Suisse, vol. iii. (1876), p. 16, pi. ii. fig. 1. The tooth 

 is described as Notidanus sp., but the figure agrees much more closely with that of a 

 species of Hybodus. 



DECADE III. VOL. III. NO. VI. 17 



