116 NATURAL HISTORY OF KERGUELEN ISLAND. 



the classification of birds) "will Dot allow that the osteological charac- 

 ters are an all-sufficient guide (in classification), believing that the 

 whole structure of a bird and its corresponding habits may be pro- 

 foundly modified, while its sternum may closely resemble a common 

 form, and vice versa.'''' (See Ibis, 1864, pp. 36-41.) Chionis is a forcible 

 illustration of this sound remark. 



It seems worth while to note a generic distinction probably existing 

 between Chionis alba and the so-called G. minor. We have not had the 

 opportunity of examiniog the former, and must judge solely by the de- 

 scriptions thereof which have been published. According to De Blain- 

 ville there is even a difference in the number of the cervical vertebrae. 

 He describes G. alba as possessing one more cervical vertebra than we 

 find in G. minor. No descriptions allude to the extension of the ca- 

 runcular casque entirely across the forehead in either species. The var- 

 ious descriptions of G. alba indicate a very different arrangement of 

 the caruncular folds about the eye ; the sheath of the bill in G. alba is 

 flat and closely apposed to the upper mandible, as in Lestris, while in G. 

 minor it is canted upward anteriorly and tubular, almost as in the 

 petrels. 



These characteristics, among others, seem to us to be supra-specific ; 

 and in view of the fact that we consider Chionis minor to be undoubt- 

 edly nearest to the ancestral type, we proi)ose to callit Chionarchus. ■ Its 

 name would then be in strictness Chionarchus minor (Hartl.). 



/ 



