170 



MR. G. A. BOULENGER ON 



[Feb. 17, 



nothing to do with the division of the vertebra. I have also noticed 

 a less complete division, or rather bifurcation, of the caudal neural 

 spines in Agama bibronii. The specimen further shows very clearly 

 the paired intercentral hypapophyses at the base of the tail, to 

 which attention was called by me at a previous meeting'. 



Returning to the classification of Lydekker, I can only repeat 

 what I have said above as to his division of the Rhynchocephalia 

 into two suborders Homoeosauria and Sphenodontina, that the only 





Lateral view of middle caudal vertebras, and lower view of second sacral and 

 three anterior caudal vertebrae of Lacerta ocellata. Twice natural size. 



diagnostic differences revealed by the definitions are that in the 

 former the praemaxillaries apparently do not form a beak and the ribs 

 have no uncinate processes, whilst in the latter the praemaxillaries 

 form a more or less deflected beak and the ribs have uncinate 

 processes. These characters, even if well founded, would be insuf- 

 ficient for subordinal separation ; but they are not exact, for it is 

 well known that Sphenodon has praemaxillary teeth ; it is therefore 



1 See above, p. 114. 



