The resemblance consists in the presence in the two regions of certain striking-looking fishes not found in other parts of the world. An analysis of these resemblances takes away much of their impressiveness. Most of the forms in question are widely distributed, ranging from Japan through India to the Cape of Good Hope. Only three genera are restricted to Japan and the Mediterranean. Resemblances equally strong exist between Japan and the West Indies, or between Japan and Australia. The differences are equally marked. The types regarded as of Japanese origin are all wanting in the Mediterranean. Those of Mediterranean origin are wanting in Japan. There are two main reasons why one fish fauna may resemble another; the one, actual connection, so that fishes migrate from one region to another; the other, similarity of physical conditions, favoring in each region the development of similar kinds of fishes. The evidence points toward the theory that similarity of physical conditions is the chief source of resemblance between Japan and the Mediterranean. The resemblance between Japan and the West Indies is due to this cause, while that of Japan to the East Indies is due largely to direct connection. If Japan and the Mediterranean were ever connected, the Red Sea must have been a region of junction. Yet, while the Red Sea in its fishes closely resembles southern Japan, it has almost nothing in common with the Mediterranean. Except a few shallow water or brackish water types, the shore fishes of the two regions are wholly distinct, none of the characteristic genera of either sea being found in the other.
Yet, geologists affirm that in Pliocene or Post-Pliocene times the Isthmus of Suez was submerged. It is made up of Pliocene, deposits with alluvium from the Nile and drifting sand-hills. Admitting this to be true, the nature of the fishes shows that this channel must have been very shallow and probably in part occupied by fresh water. No bottom-fish or rock-fish has crossed it—only sting-rays, torpedoes, eels and mullets appear to have passed from one side to the other. It must have been impossible for Japan and the Mediterranean ever to have exchanged their deep-water fishes in this way. The only other alternative is the Cape of Good Hope, and this barrier is, to this day, passed by many characteristic fishes of both oceans.
Four hundred and eighty-three genera of fishes are known from Japan. For the purpose of our present study we must take from this list all the fresh-water types, derived from China; all the northern types, derived from Bering Sea and the general Arctic stock; all the pelagic fishes, at home in the open sea, and all the bassalian fishes, or those inhabiting great depths below the range of climatic changes. After these are withdrawn, we have left the shore fishes of tropical, or semi-tropical, origin. Of these, Japan has 334 genera; the Mediterranean, 144; the Red Sea, 191; India, 280; Australia, 344; New