DISTRIBTTTION OF THE GTJLLS AND TEENS. 401 



Chatham Island) is indicated ; the plumage seems to be that of 

 maturity, and the date accords with what that plumage ought to 

 be. The other specimen, whicb is in the Paris Museum, is stated 

 by Neboux to have been obtained at Monterey, California, during 

 the cruise of the ' Venus '; but that frigate also visited the Gala- 

 pagos, and there may be a mistake in the locality, as Mr. O. Salvin 

 has shown that such errors have occurred with other birds. This 

 supposition is favoured by the fact that the American naturalists 

 have kept a keen but unavailing look-out for it during many 

 years past ; and as the Gralapagos group is seldom visited except 

 by whalers and an occasional British man-of-war, it seems probable 

 that this is another of those forms which are not merely confined 

 to that archipelago, but even to a few islands of it. Under these 

 circumstances it is interesting to find that its nearest ally comes 

 at times so close to its domain ; and this approximation in the 

 Pacific is another link in the chain of evidence respecting the 

 centre of dispersion. 



These Fork-tailed Grulls lead in a manner to the subfamily of 

 Terns (Sternince), although there is a tolerably wide gulf between 

 them, as shown by the shape of the bill, the short feet and tarsi, 

 and the long wings, the latter pointing to increased adaptation 

 for prolonged flight. Accordingly we find that, as a rule, there 

 are fewer specialized forms than in the Larince, and that the range 

 of the majority of the species is wider than in the same proportion 

 of the Gulls. This is mainly due to the conditions of their exist- 

 tence, which depend on fish and aquatic productions ; but even 

 under conditions so favourable to dispersion, there are not wanting 

 some remarkable instances of isolation. Of the larger and heavier 

 species, the largest, Sterna cassia, although nowhere numerically 

 abundant, has an immense range, being found breeding from the 

 Nearctic and Palsearctic regions down to New Zealand, although 

 replaced throughout intertropical America and on the west 

 coast of Africa by the somewhat smaller and more elegant S. 

 maxima, Bodd. >S^. cantiaca has a western Palsearctic and eastern 

 Nearctic range, going to the Cape of Good Hope in winter, as do 

 also both S. fluviatilis and S. macrura, Naum., our Common and 

 Arctic Terns, which have a more extended range in the north, 

 whilst none of them are know^n to breed in the southern hemi- 

 sphere. Prom the Mediterranean to the Malay Archipelago and 

 Torres Straits is found S. media, the Old- World representative of 

 S. elegans and S. eurygnathus of tropical America ; whilst from 



