EMIGRATION AND EVOLUTION 133 



ally is sedentary in the sense of not crossing the 

 sea. The St. Kilda Wren (^Troglodjjtes hirtensis) 

 is another instance. This island form of the 

 Common Wren (T. paruuius) succeeds in retaining 

 its distinguishing characteristics, not only because 

 it is sedentary in St. Kilda, but because its island 

 home is a long way removed from the usual fly- 

 lines of any migrant Wrens that cross from the 

 continent to our islands. This endemic race 

 shows how rapidly variation can take place when its 

 greatest check, the facility of interbreeding with the 

 parent form, is removed. On the other hand, such 

 slight local variation as is presented in the British 

 form of the Coal Tit {Pants ater britannicus) and 

 the Long-tailed Tit {Acredula caudata rosea) are 

 prevented from becoming more specialized or 

 constant by the regular influx of individuals from 

 adjoining continental areas, which visit our islands 

 and interbreed with these local forms. Again, 

 Heligoland (with Sandy Island, a tiny islet of only 

 some 250 acres area) has a wonderful record of 

 no less than 396 species reputed to have been 

 met with on its shores, but does not contain 

 one endemic bird, because it is situated on another 

 important route of migration. The British Islands, 

 being one of the best examples known of recent 

 continental islands, furnish wonderful evidence of 

 the emigration of species. At a period no more 

 remote than the latter part of the Post-Pliocene 

 Glacial Epoch, nearly if not all their area was sub- 

 merged to a depth of some 2000 feet, only our 



