The Principles of PALAEONTOLoaY. 181 



maining for several years. The tendency of the evolution shows 

 itself here without a possibility of ambiguity. Unfortunately 

 the data of Palaeontology do not seem to be in accord with 

 this evidence. All the living families of Turtles are rep- 

 resented in the fossil state by forms but little differing from 

 those existing. But the most ancient of all, the PsammochelySy 

 of the Trias of Wurtemberg, belongs to the group of the 

 Pleurodira^ which is the most differentiated and farthest removed 

 from the modern type of Reptiles. 



The Turtles of the Jurassic and Cretaceous epochs would appear 

 to be less and less ossified, the oldest of the Atheca dates from 

 the Upper Cretaceous, and the lowest type of the group 

 {Dermochelys) is at present existiog. 



Consequently, Rtitimeyer, Baur, and after them Zittel, believe 

 that the evolution of known forms must have been produced in 

 an inverse order. All the known forms which have the carapace 

 partially developed, whether in fossil state or living, would be 

 regressive, derived from specialized types. According to this, 

 the true progenitors of the Turtle would be unknown to us. 

 If this is really so, Embryogeny can scarcely demonstrate it, for 

 it is difficult to conceive that an osseous carapace should exist in 

 an embryonic state, then be absorbed, and the development be 

 forcibly arrested. Nevertheless, the sum of the anatomical 

 characters tends to prove that the Atheca are really a lower form 

 of the Turtles; many traits of their organization bring them 

 near to other Eeptiles, for example, the Rhynchocephala. If 

 these are not primitive types, regression has at least so influenced 

 various parts of the skeleton that the entire animal would very 

 much resemble its remote progenitors. 



Geratology. — Acceleration not only influences the earliest 

 stages of development, but its action extends also to later acquired 

 characters. It may happen that the definitive adult state of cer- 

 tain species is only a temporary condition for other allied species. 

 This state will exist for a considerable time, during which the 

 animal continues the functions of reproduction. Later, when the 

 animal has attained an advanced age, modifications will appear. 

 These generally consist, in a more or less marked regression, in a 

 suppression of the highest characters recently acquired, and in a 

 general simplification. Cases of senile degeneracy are frequent 

 in living nature. Generally they exert no influence on the 

 evolution of the group, though this is not always the fact. 



Late- acquired or geratologic characters, to use an expression of 

 Hyatt,^ assume a great importance when they affect, not a few 



♦ Hyatt, Genesis of the Arietidce {Mem. Mus. Compar. Zool. Cambridge), 1889. 



