326 ORIGIN OF EXISTING VEGETABLE CREATION. 
species ; but on all these occasions it is culture that has been the cause: 
as far as I know, we possess no facts to prove that natural causes have 
produced this effect. Indeed, it is much more likely that a species should 
altogether disappear under a change in external circumstances, than that 
it should become specifically altered ; unless it happens to be of that sort 
which assumes different habits under different circumstances, as is the 
case with amphibious plants, and such as take different forms in shade and 
light. Peat lands, on drying up, produce no longer Primula farinosa, 
species of Drosera, Andromeda polifolia, Scheuchzeria, &c. ; but it is not . 
that these plants are transformed into new species. Anemone nemorosa, 
Hepatica triloba, Oxalis Acetosella, &c., disappear from lands from 
which forests have been removed; but they are not commuted into 
new species. The same takes place with Nymphaea, Sagittaria, and Stra- 
_ liotes, on the desiccation of standing waters: they are not changed. An 
argument against the creation of new species may also be derived from 
what occurs when a tract hitherto naked becomes overgrown. Thus, 
when accessions are made from the sea, it is not new species, but those 
from the nearest coast, that constitute the vegetation of such new land. 
Naked lava formations become gradually covered with vegetation, and 
coral islands, rising above the surface of the ocean, become overgrown. 
In this last case it seems that only those plants are at first produced, 
which can be brought as seeds or fruits by currents and waves; for 
instance, the Cocoa-nut, which is peculiarly fitted for that kind of 
migration. Hence it is that islands of this nature, if far away from 
land, are peculiarly poor in vegetation; such as Keeling’s Island, 
south-west of Java, according to Darwin; and various islands of this 
class in the South Sea, according to Chamisso. To the same cause 
must be ascribed, probably, the phenomenon that alluvial formations 
of great extent, and still in progress, possess, if not a poor«vegetation, 
vat least one of a very ordinary description, and not marked by any 
striking peculiarity. Examples of this are the valley of the Nile, 
Lombardy, and perhaps New Holland. 
It is upon those grounds that I look upon it as highly probable, if 
. mot absolutely demonstrated, that no species are any longer created. 
(To be continued.) 
