[381] Bull. Mus. C. Z. Harv. xiv. p. 202; xxiii. p. 34 f.
[382] See papers in P. Z. S. 1878–85.
[383] A break in this uniformity may be found underneath the course of a great oceanic current like the Gulf Stream, which rains upon the bottom a large amount of food. A. Agassiz (Bull. Mus. C. Z. Harv. xxi. p. 185 f.) explains in this way the richness of the fauna of the Gulf of Mexico as compared with that of the west coast of tropical America.
[384] On the western coasts of Europe and America, where the change in surface temperature is very gradual, Purpura lapillus (the west American ‘species’ are at best only derivatives) is able to creep as far south as lat. 32° (Mogador) in the former case, and lat. 24° (Margarita Bay) in the latter, the mean annual temperature of the surface water being 66° off Mogador, with an extreme range of only 8°, and that of Margarita Bay 73°, with an extreme range of only 5°. On the eastern coasts, where the Pacific and Atlantic gulf-streams cause a sudden change of temperature, the Purpura is barred back at points many degrees farther north, viz. at lat. 41° (Hakodadi), surface temperature 52°, extreme range 25°; and at lat. 42° (Newhaven), surface temperature 52°, extreme range 30°.
[385] E. A. Smith, P. Z. S. 1890, pp. 247, 317.
[386] A. H. Cook, Ann. Mag. Nat. Hist. (5) xviii. (1886) p. 380 f; E. A. Smith, P. Z. S. 1891, p. 391 f.
[387] C. Keller, Neue denksch. Schw. Gesell. xxviii. 1883, pt. 3.
[388] According to Tate (Trans. Roy. Soc. S. Austr. 1887–88, p. 70), ‘Australian’ species predominate at Freemantle (32°), but Tenison-Woods (J. Roy. Soc. N. S. Wales, xxii. p. 106) holds that the tropical fauna extends as far south as Cape Leeuwin (34°), and that the Australian forms are not predominant until the extreme south. Tenison-Woods regards Cape Byron (31°) as the limit of the tropical fauna on the east coast, while some characteristic tropical genera reach Port Jackson, and a few (e.g. Cypraea annulus) Tasmania.
[389] A full account of the distribution of Voluta is given by Crosse, Journ. de Conchyl. (3) xix. p. 263.
[390] Usually known as ‘Patagonian,’ but since the Magellanic Sub-region includes a considerable part of Patagonia, and since the greater part of sub-region (6) lies out of Patagonia, it has been thought advisable to change the name.