.=i28 MR. H. SAUNDERS ON LARID^ [ JunC 6, 



hemisphere ; hut inasmuch as the Afcicfce proper are now generally 

 admitted to be connected with the LaridcB on the one side, as the 

 LaridcB are connected with the Limicolce on the other, and as the 

 Alcidce are only known to inhabit the northern hemisphere, it would 

 appear more probable that the LaridcB, or at least the Larince, had 

 their origin there also. There are, however, some very remarkable 

 points about the Gulls of the southern hemisphere, especially with 

 relation to the Pacific, of which M. Milne-Edwards seems to be 

 unaware, and upon which I should like to offer a few observations. 



Throughout the northern hemisphere, exclusive of the shores 

 washed by the Pacific, the Gulls (with the exception of the three 

 arctic genera Rissa, Pagophila, and Rhodostethid) fall into two 

 well-marked groups — those in which the adults bear a coloured 

 hood during the breeding-season, and those which never have a hood 

 at any time. In all the members of both of these groups the imma- 

 ture birds have a dark band across the rectrices, which disappears as 

 they approach maturity. Several representatives of each of these 

 groups also inhabit the southern hemisphere — Larus cirrhocephalus 

 of South America, and its close ally L. phceocephalus of South 

 Africa, L. maculipennis, L. glaucodes, and L. serranus of South 

 America, belonging to the former ; L. dominicanus, found from New 

 Zealand to South America (by way of the Cape of Good Hope, not 

 through the South Pacific), L. bulleri, L. scopidinus, L. novce-hol- 

 landiee, and L. hartlaubii, found at or between New Zealand and the 

 Cape of Good Hope, belonging to the latter group. But it is only 

 in the Pacific (merely including in that area the desolate islands of 

 South Shetland, the Falklands, and a portion of Patagonia near the 

 eastern entrance of the Straits of Magellan) that we find several 

 species of Gulls which agree with the other forms in having a band 

 across the rectrices in the immature stage, but differ from them 

 in many other respects. Most of these species are of a coarse 

 h.avy build, and have a tendency to a sooty hue on the under- 

 parts ; but their principal characteristic is the presence of a more or 

 less defined hood in the immature stage, which is genejally lost in 

 the adult plumage. L. scoresbii, which extends beyond the Pacific 

 to the Falklands, varies least in this respect from the general type 

 of Gulls ; it has a well-defined hood in youth, but loses the hood 

 with the disappearance of the bar on the tail, which becomes 

 white. Its range is but little beyond that of the Cape-Horn 

 current. Larus belcheri, distributed along the whole coast-line 

 of Chili and Peru '.vashed by Humboldt's current of cold water, 

 has 11 hood in youth only ; but it has a barred tail at all ages, the 

 black I'.redoniinatiug over the white. The species most nearly re- 

 sembling it in the adult plumage is Larus crassirostris, of the 

 Pacific coasts of Japan and China, in which, however, the amount 

 of black and white in the rectrices is nearly equal ; nor has the 

 your.g so well-marked a hood ; its mantle also is lighter, and its 

 otiier characteristics are so far modified as to make it intermediate 

 between the Old-World and the South-Pacific Gulls. The line of 

 communication of the ancestor of these forms, which now constitute 



