60
Cyperaceae 62 Compositae 60
Rubiaceae 57 Leguminosae 36
Euphorbiaceae 45 Rubiaceae 24
Compositae 43 Cyperaceae 24
Leguminosae 41 Euphorbiaceae 18

The cause of the great preponderance of ferns in oceanic islands has
already been discussed in my book on _Tropical Nature_; and we have seen
that Mauritius, Bourbon, and Rodriguez must be classed as such, though from
their proximity to Madagascar they have to be considered as satellites to
that great island. The abundance of orchids, the reverse of what occurs in
remoter oceanic islands, may be in part due to analogous causes. Their
usually minute and abundant seeds would be as easily carried by the wind as
the spores of ferns, and their frequent epiphytic habit affords them an
endless variety of stations on which to vegetate, and at the same time
removes them in a great measure from the competition of other plants. When,
therefore, the climate is sufficiently moist and equable, and there is a
luxuriant forest vegetation, we may expect to find orchids plentiful on
such tropical islands as possess an abundance of insects adapted to
fertilise them, and which are not too far removed from other lands or
continents from which their seeds might be conveyed.

_Concluding Remarks on Madagascar and the Mascarene Islands._--There is
probably no portion of the globe that contains within itself so many and
such varied features of interest connected with geographical distribution,
or which so well illustrates the mode of solving the problems it presents,
as the comparatively small insular region which comprises the great island
of Madagascar and the smaller islands and island-groups which immediately
surround it. In Madagascar we have a continental island of the first rank,
and undoubtedly of immense antiquity; we have detached fragments of this
island in the Comoros and {447} Aldabra; in the Seychelles we have the
fragments of another very ancient island, which may perhaps never have been
continental; in Mauritius, Bourbon, and Rodriguez we have three undoubtedly
oceanic islands; while in the extensive banks and coral reefs of Cargados,
Saya de Malha, the Chagos, and the Maldive Isles, we have indications of
the submergence of many large islands which may have aided in the
transmission of organisms from the Indian Peninsula. But between and around
all these islands we have dep