15G On a Case of Secondary Adaptation in a Tortoise. 



]uiniitive Jurassic and Eocene groups (Plesiochelyklaj and 

 IJivniil.'o), ill is difference is very marked. 



Evidently the great cranio-caudal loiigtli of the miildle 

 olenients ot; the phistron of tortoises ^represents tlie stage of 

 development attained by all tortoises after suppression of the 

 niesophistron. 



It is not without interest to investigate wliy the Cino- 

 sternidffi make an exception to tliis rule, for witln'n this group 

 tiie median bony elements of the plastron siiovv a very 

 remarkable shortening. Among the Dermatemyida3 (as 

 defined by Hay), apart from forms provided with a normal 

 well-develoju'd plastron, others also are to be met with 

 where the |)lastron is more or less reduced {Agoniphus, Hoplo- 

 chelys, and Baptemys tricarinata). In these cases the 

 reduction of the plastron acts in the first instance upon the 

 middle elements, which become shortened to such an extent 

 as to form a sort of cruciform phistron. 



A perfectly crnciform plastron is to be met with in the 

 family C'helydridpe and in Staurotypus. In these animals 

 the middle elements of the plastron are much shortened on 

 both sides, while tiie terminal elements (epiplastra and xiplii- 

 plastra) show a cranio-caudal stretching, as if their distal 

 parts had been fixed to something that prevented them from 

 yielding to the sliortening of the middle region. The shape 

 of such a plastron is very much the same as that of a 

 diamond-shaped piece of indiarubber that has become com- 

 pressed on either side while it was fixed at its two ends. 

 Since it is a well-known fact that in all primitive tortoises 

 (Am])liichelyida?) and also in the Dermatemyidaj the scapular 

 and pelvic arch adhere more strongly to the distal jjarts of 

 the plastron than in all the other Oryptodira, this explains 

 the cruciform shape. 



I Hrmly believe that in all these primitive and relatively 

 flat tortoises the scapular and pelvic arches formed internal 

 pillars, whose functions were mucii the same as tiie great 

 convexity of the shell in later terrestrial forms — namely, to 

 protect the shell from being crushed, 



Siebenrock demonstrated in the * Sitzungsberichte' of the 

 Vienna Academy in 1907 (pages 537-538) that all the Clno- 

 sternidaj with strongly develoj)ed plastron (group G. cruen- 

 tatum) originate with Cinosternidse where the plastron has 

 tlie sliape of a cross (group C. odoratuvi). To suj)port his 

 argument lie mentions the existence of an entoplastion in 

 Staurotypus and tiie existence of one or two flexible joints in 

 the plastron of the group C. cruentatum. These observations 



A 



