50 BAUR. [VOL. I. 



The points to which the capitula and tubercula of the ribs 

 are attached are raised into tubercles, and by degrees these 

 become elongated into distinct capitular and tubercular proc- 

 esses, -- diapophysis and epapophysis, -- between which in the 

 third to the tenth vertebrae, the neurocentral suture passes. 

 In the ninth and the tenth the diapophysis ascends to the level 

 of the neurocentral suture, and in the eleventh it ascends with 

 its upper half to the neural arch, being traversed by the neuro- 

 central suture. At the same time the epapophyses become 

 more and more elongated. In the twelfth vertebra a sudden 

 change in the character of the transverse processes takes place. 

 There is no true diapophysis any longer ; the capitulum is also 

 connected with the long epapophysis at the anterior middle 

 portion. In all the following vertebrae up to the lumbar region 

 the ribs are connected with the strong epapophyses. In all 

 the Crocodilia tJie first vertebra which carries the ribs completely 

 on the transverse process, is always the twclftli. In Gavialis, 

 with 25 presacral vertebrae, it is the thirteenth. There are 12 

 vertebrae behind this one in all the Crocodilia and the abnormal 

 Gavial; therefore in front of this one a vertebra must have 

 been added. By careful comparison it was found that this addi- 

 tional vertebra is intercalated between the ninth and tenth. 

 The vertebra resembles the ninth and tenth in having the capit- 

 ulum completely placed on the center just on the level of the 

 neurocentral suture. In all the Crocodilia there are two such 

 vertebrae; in the abnormal Gavial, three. In this case we have 

 no shifting of the pelvis, but true intercalation. 



In my paper on intercalation I made the following remark : 

 " If intercalation takes place at all, we ought to expect traces 

 of it in such forms which show a great increase in the number 

 of vertebrae, for instance, snakes, different families of lizards, 

 and plesiosaurs " ('91, p. 333). The enormous number of 435 

 of the vertebrae in PytJwn molnrus Linn, has certainly not been 

 reached by adding vertebrae at the distal end, but by the inter- 

 vertebral increase of the smaller numbers. It is evident that inter- 

 calation can only be proved by the presence of half-divided verte- 

 brae, and it is very interesting, and certainly not accidental, that 

 in snakes such half-divisions have been observed quite frequently. 



