ty of the genera are found
on two or more of the islands and the mainland, while nearly every
species is absolutely restricted to a single island, appears to me to be
very strong testimony in favour of a former general land connection" (p.
428).

Amphibians are affected in the same manner by sea-water as slugs are.
The accidental transportal of an amphibian from the mainland to an
island is therefore almost inconceivable. And the presence of frogs,
toads, and newts in the British Islands, in Corsica and Sardinia,
indicates, if nothing else did, that all these islands were at no
distant date united with the continent of Europe.

As regards the terrestrial reptiles, the case is somewhat different.
Many of them readily take to the sea, and, as probably all snakes and
some lizards are able to swim, it is possible that sometimes, though
very rarely, they might reach islands if not too far from a continent.
Instances of accidental transportal of land-reptiles to islands have
actually been observed. But the fact of the occurrence of such instances
by no means proves that reptiles thus conveyed are able to establish
themselves permanently in their new home. Sir Charles Lyell records in
his _Principles of Geology_ that a large boa-constrictor was once seen
floating to the island of St. Vincent, twisted round the trunk of a
tree. It appeared so little injured by its long voyage from South
America, that it captured some sheep before it was killed.

Mammals might be accidentally conveyed to islands on such rafts as have
been described by Sir Charles Lyell, and there are instances on record
of their having crossed short distances of sea by swimming. Elephants
and also deer and pigs are good swimmers, the former having been known
to swim for six hours at a stretch. "But," remarks Mr. Lydekker (p. 13),
"it may be assumed that about twenty miles is the utmost limit which
mammals are likely to cross by swimming, even when favoured by currents.
Such passages as these must, however, be of very rare occurrence, for a
terrestrial mammal is not likely to take it into its head to swim
straight out to sea in an unknown direction. Moreover, supposing a
mammal, near to a particular island, to have arrived there by swimming,
unless it happen to be a pregnant female, or unless another individual
of the same species but of the opposite sex should arrive soon after (a
most unlikely event), it would in due course die without being able to
propagat