CERATODUS. 29 



horns of some Cervida, the tusks of Elephants, or the opercular valves of Pyryoma. 

 Mr. Darwin attributes to Mr. Waterhouse the saying that " a part developed in an 

 extraordinary degree or manner, in comparison with the same part in allied species, tends 

 to be highly variable." 1 This maxim would apply well to Ceratodus. 



The variability of the teeth is perhaps in part due to difference in their setting. 

 This is suggested by the circumstance that the signs of wear affect various parts of the 

 teeth in different cases. Sometimes the whole exposed surface is abraded ; sometimes 

 only the tips of the denticles ; often the tips of the denticles and the internal angle. 

 Many are worn along the vertical inner margin, while in a few the abraded surface looks 

 quite away from the usual grinding-plane. The flat, regular, triangular forms (PI. II, 

 figs. 2, 3, 4, 11, 12, 13, PI. IV, figs. 7, 10), were probably set horizontally ; the 

 irregular oval teeth with prominent denticles (PI. II, figs. 6, 7 ; PI. Ill, fig. 1 ; PI. IV, 

 figs. 1, 2, 4, 5, 6; PI. V, fig. 1), may be supposed to have been inclined so as to present 

 the inner edge more readily than the other to the opposed tooth. 



I cannot distinguish Quenstedt's C. cloacinus 2 (Lias-kloake of Tubingen) from 

 this species, but a larger suite of examples is needed to pronounce them identical. The 

 teeth from the German Trias seem distinct. C. Gulielmi, Plieninger, 3 is perhaps the 

 nearest, but not very near. C. Kaapii, Agassiz, is more palmate, the ridges more pro- 

 duced, and the internal border elevated much beyond what is usual in our species. 4 Of 

 the species from Maledi, provisionally named by Dr. Oldham, 5 C. Virapa reminds us of 

 polymorplms, but has the ridges longer, more elevated, and more regular. I consider it 

 distinct, but allied, perhaps more nearly than is the second species found at Aust. The 

 general form, elongate with nearly parallel sides, would be exceptional, though not unex- 

 ampled, in polymorplms. 



Formation. — Rhsetic. 



Locality. — Aust Cliff, on the Severn. 



Reference. — Agassiz, P. E., vol. iii, pp. 131 — 135. 



C. parvus, Ag. PI. V, figs. 3 a, b, 4 a, b, 6 — 10. 



Smaller and more constant than polymorplms. Triangular, with the internal angle 

 usually well-defined, opposite to the second denticle, or nearly so. Denticles generally 

 5 — 4 ; ridges prominent, often extending to near the angle. External margin deeply 

 sinuous, usually ill-defined. 



1 ' Origin of Species,' 5th ed., p. 185. 



2 ' Der Jura,' p. 34, pi. ii, fig. 128. 



3 ' Palaontologie Wurttembergs,' p. 7S, pi. x, fig. 13. 



4 The figure in the ' Poissons Fossiles,' t. iii, pi. xviii, fig. 4, shows what is apparently a slip of bone 

 attached behind the second denticle, and directed backwards, as in the recent Ceratodus. 



5 ' Mem. Geol. Survey of India,' vol. i, p. 295 (1859). 



