The similarity of the crayfish of the Amurland and of Japan is a fact of the same order as the identity of the English crayfish with the _Astacus torrentium_ of the European Continent, and is to be explained in an analogous fashion. For there can be no doubt that the Asiatic continent formerly extended much further to the eastward than it does at present, and included what are now the islands of Japan. Even with this alteration of the geographical conditions, however, it is not easy to see how crayfishes can have got into the Amur-Japanese fresh waters. For a north-eastern prolongation of the Asiatic highlands, which ends to the north in the Stanovoi range, shuts in the Amur basin on the west; while the Amur debouches into the sea of Okhotsk, and the Pacific ocean washes the shores of the Japanese islands. But there are many grounds for the conclusion that, in the latter half of the tertiary epoch, eastern Asia and North America were connected, and that the chain of the Kurile and Aleutian islands may indicate the position of a great extent of submerged land. In that case, the sea of Okhotsk and Behring’s sea may occupy the site of inland waters which formerly placed the mouth of the Amur in direct communication with the Northern Ocean, just as the Black Sea, at present, brings the basin of the {326} Danube into connection, first with the Mediterranean and then with the western Atlantic; and, as in former times, it gave access from the south to the vast area now drained by the Volga. When the Black Sea communicated with the Aralo-Caspian sea, and this opened to the north into the Arctic sea, a chain of great inland waters must have skirted the eastern frontier of Europe, just such as would now lie on the eastern frontier of Asia if the present coast underwent elevation. Supposing, however, that the ancestral forms of the _Potamobiidæ_ obtained access to the river basins in which they are now found, from the north, the hypothesis that a mass of fresh water once occupied a great part of the region which is now Siberia and the Arctic Ocean, would be hardly tenable, and it is, in fact, wholly unnecessary for our present purpose. The vast majority of the stalk-eyed crustaceans are, and always have been, exclusively marine animals; the crayfishes, the _Atyidæ_, and the fluviatile crabs (_Thelphusidæ_), being the only considerable groups among them which habitually confine themselves to fresh waters. But even in such a genus as _Penæus_, most of the species of which are exclusively marine, some, such as _Penæus brasiliensis_, ascend rivers for long distances. Moreover, there are cases in which it cannot be doubted that the descendants of marine _Crustacea_ have gradually accustomed themselves to fresh water conditions, and have, at the same time, become more or less modified, {327} so that they are no longer absolutely identical with those descendants of their ancestors which have continued to live in the sea.[39]