A few of the more interesting features in connection with the distribution of fresh-water Crustacea have now been touched upon. With regard to the origin of this fauna, we can see that a number of the species are comparatively recent immigrants from the sea, working their way up the estuaries of rivers, a proceeding which can be observed to be taking place to-day in a district like the Broads of Norfolk. Others, again, but these are few, appear to be true relict marine animals left stranded in arms of the sea that have been cut off from the main ocean, and have been gradually converted into fresh-water lakes and seas. Such are, perhaps, _Mysis relicta_ and the rich Gammarid fauna of Lake Baikal, a lake that, in the presence of Seals, Sponges, and other marine forms, has clearly retained some of the characters of the ocean from which it was derived. The majority of the fresh-water species, however, have probably been evolved _in situ_, and their origin from marine ancestors is lost in an obscure past. The Crustacean fauna of the Caspian Sea[177] shows us in an interesting manner the effects of isolation and changes in salinity, etc., on the inhabitants of a basin which once formed part of the ocean. The waters of the Caspian Sea are not fresh, but they are on the average about one-third as salt as that of the open ocean. The Crustacea, described by Sars, belong to undoubtedly marine groups, _e.g._ the Mysidae, Cumacea, and Amphipoda Crevettina, but the remarkable feature of these Caspian Crustacea is the great variety of peculiar species representing marine genera which are very poorly represented in the sea, thus indicating that the variety of the fauna is not due to a great variety of species having been shut up in the Caspian Sea to begin with, but rather that, after the separation from the sea, the isolated species began to vary and branch out in the most luxuriant way—whether from lack of competition or owing to the changing conditions of salinity it is difficult to say. As an example, the Cumacea of the Caspian Sea are ten in number, all belonging to peculiar genera related to _Pseudocuma_, except one species which is included in that genus. These Caspian forms make up the Family Pseudocumidae, which contains in addition only two marine forms of the genus _Pseudocuma_ (see p. 121). A very similar condition is found in the numerous Amphipods of the Caspian Sea. Considering the enormous changes that must have taken place in the distribution of land and water even during Tertiary times, it is astonishing that the fresh waters of the world do not contain more species in common with the ocean, but it must be considered that the limited area and comparatively uniform conditions of fresh-water lakes and streams would only permit a limited number of these forms to survive which could most easily adapt themselves to the changed conditions. And these would in all probability be the littoral species that were in the habit of passing up into the