io8 THE MIGRATION OF BRITISH BIRDS 



across Canada and winter in Mongolia ! That the 

 Wheatear is a species of great colonizing power is 

 proved by the enormous extent of its breeding area, 

 which extends from the south temperate zone to the 

 highest known limits of land.^ The Knots that reach us 

 from Greenland never pass south of say lat. 65" in that 

 country, and come via Iceland to the British Islands. 



Here I may remark that the fact of the Great 

 Northern Diver {Colymbus glacialis) breeding in the 

 Faroes — and possibly within the British Isles (Shetlands) 

 — is to my mind overwhelming and positive proof that 

 some portion, if comparatively but a few individuals, of 

 this species dwelt in Refuge Area I. during the Glacial 

 Epoch, and from which their descendants passed north 

 with the return of a milder climate and open water. 

 It may seem rather an anomalous fact that a species 

 dwelling in Refuge Area I. should not breed in Scandi- 

 navia, but this Diver is thoroughly a marine bird, or one 

 only able to exist on or near to open water, and its Emi- 

 grations northwards appear to have been restricted to the 

 western portions of the British Area ; at the time they 

 were taking place the English Channel and the North 

 Sea were dry land, and possibly still covered with ice 

 and snow, whilst a land connection between Scotland 

 and Iceland, by barring the progress of the warm ocean 

 currents, kept the climate of Scandinavia and the seas 

 adjoining too severe for the existence of any birds at all. 

 The disappearance of the ancient isthmus or peninsula 



' Strong evidence that the Wheatears breeding in Circenland do 

 not reach the country via Iceland is furnished in the fact that birds 

 have been observed in autumn to pass the extreme south of Green- 

 land as late as the 5th of October, after having apparently been 

 entirely absent from that district for several weeks. 



