322 BE. ST. GEOEGE 5IIYABT OS THE SKELETON OF [Apr. 2 , 



bounded by two ridges, the postaxial end of each of which projects 

 beyond the postaxial margin of the lamella, and correspond with 

 the two similar directed, superimposed, postaxial processes of more 

 anterior cervical vertebrae. In P. erithacus the conditions are similar 

 save that the neural spine (n)is not so well developed (though slightly 

 more so than in the tenth vertebra), while the pleurapophysial 

 lamella is quite as long antero-posteriorly as in that vertebra, 

 though the styliform, more ventral, process (st) projecting postaxiad 

 from its hinder margin is shorter than is that of the same ver- 

 tebra. The neural arch is not so preeminently shortened antero- 

 posteriorly as in the Lory. In both species the metapophyses are 

 smaller than in the preceding vertebra. 



The twelfth vertebra of L. flavopalliatus (figs. 5, 7, & 9) has the 

 neural arch about as short antero-posteriorly as in the eleventh 

 vertebra, while the pre- and postzygapophyses do not project more, 

 and the interzygapophysial ridges {iz) are similarly developed. But 

 the neural spine {n) is much more developed than in the eleventh 

 vertebra ; it is a high subquadrate process, highest preaxially. The 

 hypapophysis (Ji) is of about the same size as in the eleventh vertebra, 

 but the pleurapophysial lamella altogether makes default. There is 

 a strong transverse process {diapophysis) (d) which serially corre- 

 sponds with the ridge bounding dorsally the antero-posterior groove 

 outside the pleurapophysial lamella of the eleventh vertebra — much 

 enlarged. The ventral surface of the centrum is subquadrate in 

 shape, the hypapophysis (h) depending from the hinder half of its 

 antero-posterior middle. The ventral surface is bounded laterally 

 by two ridges, the anterior ends of which project preaxiad as two 

 processes one on either side of the preaxial central articular surface. 

 The same two processes exist in the same parts (though to a less 

 marked degree) in the tenth and eleventh vertebrae. 



In P. erithacus (figs. 6, 8, & 10) this vertebra is similar save 

 that (like the eleventh) it is not so much smaller than more 

 preaxial cervical vertebrae. In both species the metapophyses 

 have here become obsolete. 



The thirteenth vertebra, in both, is very similar to its serial 

 predecessor save that its neural spine (n) is more extended 

 antero-posteriorly, that its hypapophysis is more pointed, that its 

 zygapophyses are smaller, but with some reappearance of hyperapo- 

 physes (Jip), its diapophysis {d) more antero-posteriorly wide and 

 presenting a small concave articular surface on its ventral aspect, 

 towards its postero- external angle, for the tuberculum (tp) of the 

 first cervical rib, and another {cp) for its capitulum on the outer side 

 of the centrum, a little behind the process which projects forwards 

 outside the preaxial central articular surface. This process in 

 P. erithacus has here (figs. 13, 14, & 15, pp. 325 & 326) begun to 

 subdivide into two processes connected by a short ridge : one, more 

 dorsal {x), on one side of the middle of the central articular surface ; 

 the other, more ventral (c), at the preaxial end of the ridge bounding 

 laterally the ventral surface of the centrum. These latter processes 

 may be regarded as catapophyses which here make their appearance 



