ME, W. K. PARKER ON THE OSTEOLOGY OF THE KAGU. 513 



upper spine begins to reappear, and in the last two the ribs do not become ankylosed. 

 The last but thi-ee (the thirteenth)' has more breadth than length (PL XCI. figs. 1 & 2, 

 and PI. XCII. tigs. 17, 17 «, 17 b, 17(7); and this great development in the transverse 

 direction is still greater in the three last. The spine of the last cervical (PI. XCI. 

 figs. 1 & 2, and PI. XCTI. figs. 18, 18 ff, 18 5) is very high and also very thick; it 

 has a low inferior spine, a large trans^'erse process that projects beyond the " tuber- 

 culum " of its styloid rib, and a neat cup on each side for the corresponding " capitulum 

 costee;" the styloid feeble ribs (PI. XCI. fig. 2, is) are less than an inch in length. 

 The five dorsal vertebrae are all ankylosed together, except the last, and in the free 

 bone the inferior crest has died out (PI. XCI. fig. 1, c? o). The upper spine of the first 

 dorsal is higher than that of the second ; they then increase in height until we reach the 

 third sacro-lumbar. As the tendons are but little ossified in the Kagu, the transverse 

 processes of the dorsals are permanently distinct (PI. XCII. fig. 5 d) ; the bodies of the 

 second, third, and fourth are very thin and carinate. 



In the height of the dorsal and lumbo-sacral spines the Kagu comes near to Oci/dro- 

 mus ; in the Kagu the third sacro-lumbar attains its greatest height (PL XCI. fig. 1) ; 

 behind this the spine aborts rapidly, and the hinder two-fifths of the sacrum has no 

 spine. 



The eleventh sacral has a long, unsegmented rib on each side ; for this part of the 

 sacrum is of great wdth (PL XCII. figs. 5 & 7) ; behind the eleventh the component 

 vertebrae gradually lose the distinction in rib and transverse process. The sacrum 

 (PI. XCI. fig. 1) has its last third more bent downwards than in any bird I know, and 

 the caudal series (cd) takes the same downward direction. The caudal vertebrae of the 

 Kagu (PL XCI. fig. 1, cd, and PL XCII. figs. 5 & 7, cd) are furnished with long 

 transverse processes, as in the Eurypyga ; in this they both differ from the Psophia. 

 The last piece (PL XCII. fig. 1, cd) is sharper than in either of the related types. The 

 ribs (PL XCI. fig. 1, and PL XCII. fig. 5) are feebler than in any related type, and the 

 appendages are small, and only present on the first four. The upper part of the ribs is 

 very much enlarged in Psophia, Ocydromus, and Eurypijga, much less so in the Kagu. 

 The sternal pieces {sr) are also feeble, and that belonging to the sacro-lumbar rib 

 {s.Jr) is imperfect below; altogether, the ribs are very feeble and embryonic. 



But the shoulder-girdle and sternum show the most remarkable embryonic characters, 

 and have their only counterpart in this respect amongst the Carinatae in the feeblest- 

 winged Kail [Praclnjpferyx australis). Ocydromus has a more perfect sternum. 



The scapula (PL XCI. fig. 3, sc, and PL XCII. fig. 5, sc) is an extremely feeble and 

 very curved bone; it is only half the relative width of the scapula of Psophia and 

 Eurypyga; but in these types this bone is very much curved. The acromion process 

 (PL XCII. fig. 5, sc) is blunt and of a moderate size, and the suprascapular tip of the 



' In Plate XCI. fig. 2, the penultimate vertebra is numbered 17 by mistake ; in Plate XCII., fig. 17 repre- 

 sents the thirteenth vertebra. 



4b 2 



