chap. x.J THE PAL^ARCTIC REGION. 185 



the supposed regional limits, or send one or two, out of a large 

 number of species, into adjacent regions ; yet there is some 

 reason to believe that the latter are really more important as 

 characterising a zoological region than the former. In the case 

 of a single isolated species or genus we have a dying-out group ; 

 and we have so many cases of discontinuous species of such 

 groups (of which Urotrichus in Japan and British Columbia, 

 Eupetes in Sumatra and New Guinea are examples), that it is 

 quite as probable as not, that any such isolated species has only 

 become peculiar to the region by the recent extinction of an 

 allied form or forms in some other region. On the other hand, 

 a genus consisting of numerous species ranging over an entire 

 region or the greater part of one, is a dominant group, which 

 has most likely been for some time extending its range, and 

 whose origin dates back to a remote period. The slight exten- 

 sion of such a group beyond the limits of the region to which 

 it mainly belongs, is probably a recent phenomenon, and in that 

 case cannot be held in any degree to detract from its value as 

 one of the peculiar forms of that region. 



The most numerous examples of this class, are those birds of 

 the temperate regions which in winter migrate, either wholly or 

 partially, into adjacent warmer countries. This migration most 

 likely began subsequent to the Miocene period, during that 

 gradual refrigeration of the temperate zones which culminated 

 in the glacial epoch, and which still continues in a mitigated 

 form. Most of the genera, and many even of the species of 

 birds which migrate southwards in winter, have therefore, most 

 likely, always been inhabitants of our present Palsearctic and 

 Nearctic regions ; permanent residents during warm epochs, but 

 only able now to maintain their existence by migration in 

 winter. Such groups belong truly to the temperate zones, and 

 the test of this is the fact of their not having any, or very few, 

 representatives, which are permanent residents in the adjacent 

 tropical regions. When there are such representative species, we 

 do not claim them as peculiar to the Northern regions. Bearing 

 in mind these various considerations, it will be found that we have 

 been very moderate in our estimate of the number of genera 



Yol. I.— 14 



