146 Proceedings of the Royal Physical Society. 
in Northern Sweden and Finland, while Northern Asia 
forms its principal breeding area (vide ‘Heligoland as 
an Ornithological Observatory,’ p. 533. D. Douglas, Edin- 
burgh).” 
From the above remarks I conclude that the original 
trend of autumn or winter migration of this species was—as 
is the case with most ducks and swimming birds—from north 
to south; but that in course of time considerable deviation 
from the due north to south direction took place, as areas 
along the old route became congested. Also I think we 
detect the tendency on the Continent to disperse and extend 
the breeding areas on lines from easterly to westerly direc- 
tions, leaving out, however, the “turning points”! (Gatke) 
before reaching Heligoland. The distribution along the 
British coasts seems to take place along the south to north 
or south-west to north-east spring fly-lines. 
I think we can scarcely expect to find it extend to Ferée, 
Iceland, or Greenland as a breeding species for a long time 
to come, if at all, as that would entail an autumn or winter 
migration from localities still farther east and north, and a 
return again in spring. Thus, recorded occurrences of the 
birds in Ferde even as late as June (Feilden), and others 
in Shetland and Orkney in winter after stormy or excep- 
tionally severe weather, indicate an expansion from the east 
or north-east. But Feerde is not, perhaps, suitable, because 
most of the lakes are rocky and very bare of vegetation, 
which conditions, even in more favoured climes, are not 
those which the species affects for nesting localities. 
I cannot think that we at present get great accessions 
to the host in winter from Norway, as the bird is rarer along 
that seaboard; but that Finland and the districts east of the 
1 Herr Gitke clearly explains that many species find their turning points 
from east to west migration before reaching Heligoland, passing down the 
Continental coast, and thence, no doubt, by the Rhine valley, and by the 
coasts of France and Spain, and it seems likely that England first, and after- 
wards Scotland, have been indebted for their first pioneers of the species to 
individuals which have overshot their north to south line in autumn, but 
have again returned by the same or nearly the same route in spring. In 
this connection, I recommend a perusal of Mr Wm. Brewster’s appropriate 
article in the Memoirs of the Nuttall Ornithological Club, No, I., Bird 
Migration, part ii., pp. 11-22. 
