dinia, and North-western Africa, and _vice versa_.
The Balearic Islands were then connected with Spain; and we find there
many curious survivals which have long ago become extinct on the
mainland.

That the Straits of Gibraltar are only of recent formation has been
suggested on zoogeographical evidence by Bourguignat, Simroth, Kobelt,
and many others. Dr. Kobelt believes that the former land-connection
between the south of Spain and Morocco was much wider than is generally
assumed, and that the coast-line stretched from Oran in Algeria straight
across to Cartagena in Spain (_b_, ii., p. 228).

My allusions to the lands lying beyond the Lusitanian province, refer
chiefly to the Canary Islands and Madeira. Whatever doubts Dr. Wallace
had on the subject of their former connection with Morocco, it cannot be
denied that they used to be of much larger extent, especially in miocene
and pliocene times. It seems extremely probable that these islands
formed part of the mainland of North Africa until comparatively
recently, and that they are the last traces of a sunken continent which
united Africa and South America. A discussion of this problem, however,
must be deferred, as it is a complicated one, and one which would lead
me altogether outside the scope of this little volume. I hope I shall
have an opportunity to publish my views on this subject before long,
meanwhile the reader must content himself with this mere statement.

During the greater portion of the Miocene, and I think for part of the
Pliocene Epoch too, the advance of the Lusitanian species eastward was
barred on the continent of Europe by an arm of the sea which stretched
northward along the Rhone valley from the Mediterranean. The Lusitanian
forms which originated in Southern Spain were able to travel east during
these times by way of North-west Africa, Sicily, Southern Italy, and
Greece; it is possible that some may have reached the Alps in this
manner, and Eastern Europe generally. That the Lusitanian centre was
never a very active one compared with, for instance, the Oriental is
indicated by many distributional facts. It is difficult to understand,
however, why the Oriental species, on the whole, have migrated so far
west, while few Lusitanians have gone very far east. This seems to have
been noted particularly in the case of the flora. Mr. Bonnet drew
attention to the fact that in Tunis there are none of the absolutely
characteristic plants of Morocco and Sp