120 ME. W. G. EIDEWOOD ON THE HTOBEANCHIA.L 



vertebral column of Pelodytes does not differ materially from the 

 normal. The first two vertebrse are not confluent as Cope states 

 (6. p. 108). 



In both Xenopus and Pipa the diapophyses of tbe third and 

 fourth vertebrse are very long, and those of the fourth are termi- 

 nated by a large backwardly-directed plate of cartilage. It has 

 always been urged as an argument against associating the Aglossa 

 with the Discoglossidse, that the latter have free ribs on the 

 second, third, and fourth vertebrae, while the former have no ribs. 

 But in ripe embryos of Pipa (Stage VI., antea) the second 

 and third diapophyses are segmented, and the peripheral portions 

 correspond exactly with the second and third ribs of Disco- 

 glossus. The first diapophysis is short and unsegmented. The 

 hinder of the two ribs is terminated by a backwardly-directed 

 plate of cartilage. Both the ribs and the transverse processes 

 proper are ossified, and the former are freely movable upon the 

 latter. The remaining vertebrse have no autogenous ribs. In 

 Xenopus larvse (Stage II., antea) there are long, ossified, movable 

 ribs attached to the third and fourth vertebrse, and a small 

 pair to the second. The transverse processes are practically 

 absent, so that the ribs arise directly from the neural arch. In 

 larvse of Peldbates and Pelodytes I have been unable to find ribs 

 difierentiated on the second, third, and fourth vertebrse, although 

 I have examined four different stages of the former and three of 

 the latter. 



As already noted by von Ihering (Morph. Jahrb. vi. 1880, 

 pp. 297-314) and Grronberg (18. p. 641), no trace of separation 

 of the first vertebra into two parts can be detected in embryos 

 of Pipa. But the fact of the first spinal nerve passing out 

 through the neural arch, and the presence of but six other pre- 

 sacral vertebrse, point to the conclusion that two vertebrse are 

 here represented. The occurrence in the embryo of diapophyses 

 in connexion with this vertebra is a serious bar to the acceptance 

 of the view propounded by von Ihering (Z. c.) that in Pipa the 

 second vertebra has been excalated, for in no anurous amphibian 

 are diapophyses known to occur on the first vertebra. 



In PalcBohatrachus the vertebrse are sometimes procoelous and 

 sometimes opisthoccelous. In Pipa and Xenopus they are opis- 

 thocoelous. But great importance cannot be attached to the 

 characters of the articular surfaces of the centra, since it is a 

 well-known fact (Duges, 10. p. 107, PI. 4. fig. 33 ; Cope, 7. p. 253 ; 



