470 Mr. Swainson on New Australasian Birds. 



P. versicolor. P. cyanoptera. 

 Total length, from the tip of the inches. inches. 



billtotheendofthe tail 9| 7* 



Bill from the gape \\ 1^^ 



Wings in length , 5 4|: 



Tail 1| — 



Tarsi 1| 1^ 



Hallux and claw 1 | 



COLUMBID JE. 



The extensive genus of Columba, like that of Falco, has been 

 pronounced indivisible by an eminent Ornithologist of the present 

 day; who, from having made these birds his peculiar study, is in 

 one sense pre-eminently qualified to give a decided opinion. The 

 principle he has laid down, and on which this opinion conse- 

 quently is founded, is, that whenever intermediate species are 

 discovered which serve to unite two neighbouring genera, such 

 genera should invariably be united. This theory has been, 

 in my mind, completely refuted in the pages of this Journal; and 

 is so calculated to destroy, not only the valuable labours of its 

 author, but all those artificial arrangements which have been 

 invented to facilitate our acquaintance with Nature, that I shall 

 not stop to give it further consideration. It is admitted that there 

 are certain peculiarities of form, and of economy among the 

 ColumbidcB^ which point out natural divisions. Some of these have 

 been used for the construction of genera by MM. Le Vaillant, 

 Vieillot, and Cuvier, and of sections by M. Temminck ; but the 

 immense number of species already known, and the great influx 

 of new ones, renders it essential that many others should be 

 formed. As we labour under a comparative ignorance of the natural 

 economy of the vast number of tropical species recently described, 

 any attempt to throw the Coliimbidce into their natural arrange- 

 ment must be very imperfect. The basis of such a work must 

 rest on their natural habits, their food, and their geographic dis- 

 tribution. Yet, as we see in other natural families that a pecu- 



* " Longueur, sept pouees." PI. Col. pi. 218. 



