540 J. H. Fleming: 



and North Pacific, but its great centre of abundance is in the 

 Arctic Seas. 



To find a cause for the migrations we must look to where 

 Brünnich's Murre is unaccompanied by the other species, and 

 Hudson Bay would seem to be in the most likely place 

 whence the migrations had their origin. Mr. A. P. Lowe, 

 who has travelled so extensively in Hudson Bay in the 

 interests of the Geological Survey of Canada, and lately as 

 Administrator for the Canadian Government, writing to me, 

 says : " The most southern breeding places in Hudson Bay 

 are at Digges Islands and Cape Wolstenholni, unless a few 

 breed on the Arctic islands off the east coast, which is 

 improbable. The birds remain throughout the year in the 

 waters of Hudson Bay, between the moving and shore ice. 

 We killed numbers of them during the winter 1903-4 at Cape 

 Fullerton, in the north-west part of the bay. 



"The Eskimos reported them as plentiful about the outer or 

 western side of the Belcher Islands, on the east side of the 

 bay, during the winter, and there is little doubt that they go 

 south to the edge of the solid ice in James Bay, which 

 extends almost to the northern end of that bay." 



A series of migrations confined to one species, extending 

 over territory much of which was misuited to sustain the 

 life of a bird whose food is marine, must have some definite 

 cause, and in relation to cause, several facts connected with 

 these migrations must be noted. 



While a great proportion of the birds examined were 

 young, some at least were adult. 



It is extremely probable that none of the birds survived 

 to return to their place of origin. Only in isolated cases 

 were birds found with food. I have quoted all such records, 

 some six in number, five of which are not bej'ond the 

 supposed normal range of winter migration, the remaining* 

 one, of a Toronto bird, taken February 23rd, 1895, shows- 

 that it is not impossible to sustain life on the food obtainable 

 in the Great Lakes, where the Long-tailed Duck (Harelda 

 cjlacialis) winters in considerable numbers. 



It seems likely that no food was obtained from the place 

 where the necessity of migration arose to where it ended, or 

 where the individual birds dropped exhausted. Even when 



