7^ Le Conte on the Coast Islands of California. 
parts. This colonization was probably mainly from Mexico, 
<jithcr directly or through the Sierra region; for the distinct- 
ively Californian plants, though peculiar, are more like those of 
Mexico than any other. Whencesoever it may have been col- 
onized, however, the environment was sufficiently peculiar, the 
'isolation sufficiently complete, and the time has been sufficiently 
long to make a very distinct flora. According to \Vallace, it is 
one of the primary divisions of the nearctic region. 
During the late Pliocene and early Quaternary, as already 
seen, the islands were still a part of the mainland, and the whole 
was occupied by the same flora, viz: the distinctively Califor- 
nian (with some differences doubtless), now found in both, 
together with the peculiar island-species. 
During the oscillations of the Quaternary the then western- 
most coast range was separated by subsidence, and has remained 
ever since as islands. Simultaneously with, or after, this sepa- 
ration, came the invasion of northern species, driven southward 
by glacial cold. Then followed the mingling of invaders with 
the natives, the struggle for mastery, the extermination of many 
.(viz: the peculiar island sj^ecics), and perhaps the slight modi- 
fication of all, and the final result is the California flora of 
to-day. But the island flora was saved from this invasion by 
isolation, and therefore far less changed than the flora of the 
mainland, i. c.^ the invading species are mostly wanting, and 
many species survived there which were destroyed, or else 
modified into other species on the mainland, and the remainder 
probably less modified than on the inainland. The flora of 
these islands, therefore, represents somewhat nearly the charac- 
ter of the flora of the whole country during the Pliocene times. 
Some modification they have doubtless suffered, but the time 
has been too short for any great change in the absence of severe 
competition. 
The question naturally arises, "How is it that with a separa- 
tion of only 20-30 miles the two floras — insular and mainland 
— have not become entirely similar by mutual colonization?" 
The prevailing winds being landward would, I suppose, largely 
prevent the colonization of common American forms on the 
islands, although some such colonization has in fact taken place. 
But with the prevailing winds in this direction, why have not 
