We now come to the Amphibia, which are represented in the Seychelles by two tailless and two serpent-like forms. The frogs are _Rana mascareniensis_, found also in Mauritius, Bourbon, Angola, and Abyssinia, and probably all over tropical Africa; and _Megalixalus seychellensis_ a peculiar tree-frog having allies in Madagascar and tropical Africa. It is found, Dr. Wright informs me, on the Pandani or screw-pines; and as these form a very characteristic portion of the vegetation of the Mascarene Islands, all the species being peculiar and confined each to a single island or small group, we may perhaps consider it as a relic of the indigenous fauna of that more extensive land of which the present islands are the remains. The serpentine Amphibia are represented by two species of Caecilia. These creatures externally resemble large worms, except that they have a true head with jaws and rudimentary eyes, while internally they have of course a true vertebrate skeleton. They live underground, burrowing by means of the ring-like folds of the skin which simulate the jointed segments of a worm's body, and when caught they exude a viscid slime. The young have external gills which are afterwards replaced by true lungs, and this peculiar metamorphosis shows that they belong to the amphibia rather than to the reptiles. The Caecilias are widely but very sparingly distributed through all the tropical regions; a fact which may, as we have seen, be taken as an indication of the great antiquity of the group, and that it is now verging towards extinction. In the Seychelles Islands there appear to be three species of these singular animals. _Cryptopsophis multiplicatus_ is confined to the islands; _Herpele squalostoma_ is found also in Western India and in Africa; while _Hypogeophis rostratus_ inhabits both West Africa and South America.[107] This last is certainly one of the most remarkable cases of the wide and discontinuous distribution of a species; and {433} when we consider the habits of life of these animals and the extreme slowness with which it is likely they can migrate into new areas, we can hardly arrive at any other conclusion than that this species once had an almost world-wide range, and that in the process of dying out it has been left stranded, as it were, in these three remote portions of the globe. The extreme stability and long persistence of specific form which this implies is extraordinary, but not unprecedented, among the lower vertebrates. The crocodiles of the Eocene period differ but slightly from those of the present day, while a small freshwater turtle from the Pliocene deposits of the Siwalik Hills is absolutely identical with a still living Indian species, _Emys tectus_. The mud-fish of Australia, _Ceratodus forsteri_ is a very ancient type, and may well have remained specifically unchanged since early Tertiary times. It is not, therefore, incredible that this Seychelles Caecilia may be the oldest land vertebrate now living on