DE CANDOLLE ON SPECIES. 
197 
Apennines, that it once spread over the intermediate areas; the 
glacial epoch would have supplied the required climate for it to do so, 
and we thus arrive at the belief that cold caused by the northern 
glaciers, and the contiguity of Corsica and Sicily with the European 
continent were simultaneous, or more or less so. 
From this and some negative palaeontological evidence, confessedly 
however of no appreciable value, M. De Candolle concludes, that 
the Beech appeared in Europe, at the base of the Alps and Apennines, 
at the end of the tertiary epoch only, in the long interval which 
followed the elevation of the Alps, and preceded the separation of 
Sicily and Corsica from Italy; thence it spread to Holland, Nor¬ 
mandy and the British Islands, since the Homan epoch only; and lastly 
it is becoming more common in our time in Germany and Denmark. 
The history of the Chesnut is similarly treated, as follows:— 
Its history is analogous to that of the Beech in Europe. Its 
northern limit is Belgium on the west, Styria and the Crimea 
on the east; it shows no inclination to spread in England, if in¬ 
deed it was ever wild there, which is more than doubtful, # and is 
undoubtedly absent in Ireland and the Azores. It is registered as a 
native of Madeira and the Canaries, but evidently cultivated speci¬ 
mens have been taken for wild ones. It abounds in the south of 
Europe, Sicily, Sardinia, and Crete, as well as in Asia Minor; it 
must have established itself in these islands before their separation 
from the continent. Its abundance on the mountains of Grenada 
renders its absence in N. Africa most remarkable, since the climate 
of Algeria is identical in the east with that of Sicily and Sardinia, 
and in the west with that of southern Spain. Hence we arrive at 
the only two possible hypotheses, either that an arm of the sea pre¬ 
vented its migration from Europe to Africa, or that if if once did in¬ 
habit Africa, it has since been destroyed there, of which hypotheses 
the first appears to M. De Candolle the most probable. 
After alluding to the absence of the genus Castanea in the ter¬ 
tiary rocks, and more recent European deposits, and in the island 
of Cyprus, M. De Candolle proceeds to state, that the European 
Chesnut inhabits two other regions, China and Japan, and (a variety 
of it) the United States, but that it is absent in N. W. America, 
and over a large area of Central Asia. It thus has three centres, 
of which our own and the Eastern Asiatic are ancient; for the first 
antedates the present condition of the Mediterranean basin, and the 
second antedates the separation of Japan from the American conti¬ 
nent ; and further the existence of four distinct varieties of the tree 
in Japan, besides the European form (also present), indicates its anti¬ 
quity there. Lastly, the American form extends from the moun¬ 
tains of Carolina to the plains of Maine and Michigan, and there are 
no grounds for speculating on its relative antiquity. 
* We suppose that there can be no doubt that it never was indigenous in the 
British Islands. 
