124 
president’s address. 
There is no doubt that these Pigeons have enemies. Although 
there are no mammals in the Fiji Islands, they are visited by an 
Australian Peregrine Falcon and an Australian Harrier, to say 
nothing of a Fijian Goshawk, which is probably the worst enemy 
of the three. 
It is very difficult to know how to select examples from the 
many genera which are represented in the Polynesian sub-region of 
the Australian region without trespassing too long upon your time 
and attention. Amongst the Parrots the genus Cyanorhamphus is 
specially interesting, as showing a great number of closely allied 
species, each with a definite circumscribed locality. The Fly- 
catchers are also very remarkable in this respect, especially those 
belonging to the two genera Ehipidura and Myiagra, which are 
largely represented in Polynesia. 
The Kingfishers of Polynesia have also a most interesting 
distribution. They are all closely allied, and all belong to the 
genus Halcyon. Three species have indeed been removed to a 
genus which has been called Todirhamphus by Lesson, Reichen- 
bach, Bonaparte, and some other pre-Darwinian ornithologists, on 
the ground of their somewhat flattened bills ; but there can be no 
doubt that the flattened bills have been independently acquired, 
and therefore denote analogy, whilst the similarity of colour to 
other species has been inherited from common ancestors, and 
therefore denotes affinity. No blame attaches to these old writers; 
they constructed their genera according to their lights; they placed 
a much higher value on the similarity of the shape of the bill 
than on the similarity of a complicated pattern of colour. They 
knew no better ; Darwin had not convinced the scientific world 
of the truth of the theory of evolution ; they were still working 
upon the antiquated lines of Linneus, and had only the vaguest 
conception of the difference between analogy and affinity. How 
could they have had any notion of the idea, which is now accepted 
as an axiom by all scientific men, that classification cannot be 
merely morphological but must be genetic ? There are no genera 
in nature. Genera may be divided and sub-divided to suit tho 
convenience of tho student, but they can only be divided in 
