Le Conte on the Coast Islands of California. 79 
all the peculiar island species been long ago colonized on the 
mainland? According to the view above presented the answer 
is evident. These peculiar species did once inhabit the main- 
land and have been either destroyed or transformed in the 
struggle with invaders. They are, therefore, weaker species. 
The same unfitness which made them succumb then, still for- 
bids their successful colonization. This brings me to the next 
point. 
There are quite a number of rare and j^eculiar forms found 
struggling for existence in the southern counties which are 
found very abundant on the islands. This certainly looks like 
the beginnings of colonization. This is indeed Mr. Greene's 
view, and is rendered all the more probable by the fact that the 
ocean currents probably drift in that direction. But there is at 
least another explanation suggested by the view above pre- 
sented. These may be, and probably are, rcffinants of Pliocene 
indigenes still undestroyed, but ready to perish. From this 
point of view their place far south is just what we might ex- 
pect, for the main invasion was from the north. 
But there is still a last point to be explained. Lavateras are 
unknown in the New World, except on these islands, where 
there are four species. But they are found in the Old World, 
in the Mediterranean region and in Australia. Mr. Greene 
suggests, as a possible explanation, a fot'iucr connectio7i of these 
islands xvith some other continent. I think not. The substan- 
tial permanence of continental land masses and oceanic basins, 
with only marginal changes, at least during later geological 
times — taken together with the comparative recency of the 
flora of California — renders this explanation extremely improb- 
able. The above presented view suggests another and far 
more probable explanation. 
The existence of Lavateras in such widely separated localities 
as Australia, the Mediterranean region and the coast islands of 
California, shows unmistakably that existing sjDccics arc but 
remnants of an old, once very abimdant and widely spread 
genus, with numerous species. They are now dying out. 
They have been mostly destroyed and re^^laced by newer and 
stronger forms. I conclude, therefore, that in Pliocene times 
several species of Lavatera existed all over the coast region of 
