408 BULLETIN OF THE 



I. Cosmopolitan Species. A large proportion of ornithologists 

 have of late been unwilling to admit that any bird has what is usually 

 termed a " cosmopolitan " range, while others recognize only about 

 twenty such species, taking into account, of course, their total range. 

 These embrace two or three species each of hawks and owls, the rest 

 being either Grallce or Natatores. Very few of them, however, breed 

 within both the tropic and the polar zones ; many of those which 

 visit the shores of all lands in their migratory journeys being restricted 

 in the breeding season to comparatively limited areas. Pandion 

 haliaetus and Otus brachyotus are the only examples of commonly so- 

 cahed cosmopolitan species which appear to breed from the Arctic Circle 

 southward through the tropics to the southern extremity of the southern 

 continents. Falco peregrinus may form a third, but its peculiar breed- 

 ing habits give it a very irregular dispersion at that season. Slrix 

 flammed appears to be also everywhere resident, except in the arctic and 

 cold-temperate zones. Colyle riparia and Hirundo rufa (including 

 under the latter name the several slightly differing geographical races 

 of this group, which have of late been regarded as species), seem also to 

 be nearly cosmopolite. The list of species which are permanently cos- 

 mopolitan will hence not exceed half a dozen, and are those above 

 enumerated. 



II. Circumpolar Species. Regarding as circumpolar species only 

 those numerously represented in both the eastern and western hemi- 

 spheres, nearly one hundred species* can be included in the list of 



the interior, as has been previously explained; so that an indication of only the average 

 boreal and austral limits of the species at this season has been aimed at, and only so far 

 is their winter range is circumscribed within the region under special consideration. 

 The blanks in the third column of the tables hence indicate that the species winter 

 entirely within the American Tropical Realm; those in the fourth, that the austral limit 

 is within that realm. The few occurring in the second column of the tables indicate 

 that the species in question also ranges southward in the breeding sea>on into the Tropi- 

 cal Realm. A [?] in place of a blank indicates that the southward range of the species 

 is supposed to be limited to the Eastern Province, but as being too vaguely known to 

 warrant a specification of its limit in the direction indicated by the column in which the 

 query stands. 



* Dr. Richardson, in 1831 (in the " Fauna Boreali-Americana," p. xxxix), gave thirty- 

 two species of land birds and "upwards of sixty-two species of water birds" (ninety- 

 four in all) as "common to the Old World and the Vm- Countries." A few truly cir- 

 cumpolar species were not included in this list, and others were included which were 

 merely accidental vi-itors from one continent to the other. Since the date of that list 

 the identity of the greater part of the species therein mentioned has been questioned by 



