190 REPORT— 1845. 



a new edition, or in any other work that embraces the same objects : and in 

 regard to the extent of range of the described species, the alterations will be 

 no less important. 1 shall not therefore attempt more in this paper in refer- 

 ence to the geographical distribution of fish than merely to mention one or 

 two facts that have some bearing on opinions at present entertained by geo- 

 logists. Much stress has been laid upon the existence of tropical forms of 

 fish in the ancient deposits of northern latitudes as a proof of the high tem- 

 perature of the earth in former ages; but I believe that the range of inter- 

 tropical species is less restricted than it has been supposed to be. Among 

 the Bermudas, on the 32nd parallel, the Chcetodontidce are so abundant that 

 they are preserved in basins inclosed from the sea as an important article of 

 food for the garrison and inhabitants ; and a considerable number of fish 

 range northwards from the Brazils to the coasts of the United States, some 

 of them even to the banks of Newfoundland. It is probable that the gulf- 

 stream has something to do with this, as fewer tropical forms seem to reach 

 the same parallels on the coasts of Europe. If so, there is probably a cur- 

 rent of a similar kind setting to the northward on the coasts of China, for 

 many species which abound in the Indian ocean range as far north as Japan. 

 M. Agassiz says, " Les Xiphio'ides de Sheppy ont tous le bee arrondi comme 

 le Tetrapture et les Histiophores ; or ces derniers ne quittent jamais les mers 

 du Sud." (Rep. Br. Ass. for 1844, p. 305.) Yet M. Burger has discovered 

 a Histiophorus on the south-west coasts of the Japanese isles, and the same 

 or another species exists in the seas of New Zealand. 



Several remarkable generic forms described in the ' Fauna Japonica,' such 

 as Hoplegnathus or Scarodon, Histiopterus, Melcmichthys or Crenidens and 

 others, have been detected also in the Australian seas. In short, from the 

 42nd degree of south latitude to the same parallel north of the equator, be- 

 tween the meridians which include Australia, New Zealand, the Malay ar- 

 chipelago, China and Japan, there is but one ichthyological province, though 

 towards the respective extremes there is a mingling of antarctic and arctic 

 forms with a corresponding diminution in the numbers of the intertropical 

 ones. But in the middle portion of this province its dimensions in longitude 

 are vastly extended. Very many species of the Red sea, the eastern coast 

 of Africa, Madagascar and the Mauritius, range to the Indian ocean, the 

 southern seas of China, the Malay archipelago, the northern coasts of Au- 

 stralia, and the whole of Polynesia, — the almost continuous ranges of islands 

 apparently favouring their distribution. A comparatively small number of 

 these species enter the Atlantic, and such as do are mostly Scomberoids, 

 Scopelines, Lophobranchs, Plectognathes or Sharks. It is repeatedly re- 

 marked in the ' Histoire des Poissons,' that few species of fish cross the 

 Atlantic. From this observation, the Scomberoids which skim the surface 

 of the high seas ought perhaps to be excluded ; and some allowance must 

 also be made for South American species discovered on the African coasts 

 and islands since the time that the passages in the i Histoire des Poissons/ to 

 which I allude, were written. But with these qualifications, the remark ap- 

 pears to be well-founded, and the great bulk of species on different sides of 

 the Atlantic are different. When we seek for some cause which may explain 

 this difference in the distribution of the fish of the two oceans, we observe 

 that the bounding shores of the Atlantic run north and south, with a deep 

 sea between them, and no transverse chains of islands. " On the other hand, 

 we have from Africa eastward, within the warmer districts of the ocean, a 

 continuous range through the Indian ocean and archipelago, the Malay ar- 

 chipelago and Polynesia, which embraces three-fourths of the circumference 

 of the globe; there being no points of continent which cut through that 



