while we sometimes possess of fossil animals entire skeletons, and nearly always parts on which we can justly base a classification, we ordinarily get of the ancient trees only isolated specimens of leaves, or more rarely of fruits and seeds. We, nevertheless, generally succeed in determining these relics, and, by comparing them with analogous living ones, in forming conclusions, the probability of which carries conviction. Thus, with the aid of the data furnished by stratigraphy, we can not only reconstitute the forests of former days, but can also arrange them chronologically, grasp their mutual relations, establish their filiations, and finally explain how they have in the past been displaced and renewed.
It is necessary to take account of the peculiarity that trees are enrooted or fixed in the soil, so that only their seeds can leave them and be carried away, but never to a very great distance. This fixity is certainly one of the causes of the regularity and relative slowness of the modifications to which arborescent vegetation has been subjected in the periods anterior to ours. The new-comers of each region can never have rapidly traversed distances. It has been rather by slow steps, and by the aid of at first partial introductions, that the flora of all the epochs has been transformed. Instead of leaps, we meet with modifications aided by time, which were worked out through a long duration before becoming definitive. An attentive examination of the vegetable impressions collected over many successive levels and at points distributed along the course formerly followed by the vegetation, and marking its advance, should therefore enable us to recover the partial terms of the presumed filiation of the types whose origin we are investigating.
One phenomenon has been remarked in intimate relation with this gradual and successive substitution of plants; it is the cooling of the globe, operating insensibly, but subject to a general movement, the progress of which, although extremely slow, has never been arrested. Plants have pursued their migrations under the rule of this phenomenon, moving toward the south and gradually abandoning the north, beginning with the extreme north, or the immediate environs of the pole. The discovery of numerous vegetable fossils at different points in the arctic regions, in Spitzbergen, Greenland, Grinnell Land, etc., has been sufficient to give rise to terms of comparison and demonstrate what was the character of fossil vegetation when that of Europe more or less resembled the present vegetation of countries near the tropics. Hence it has been possible to establish with fair probability not only the general march but also the filiation of a number of plants; and it has been ascertained that the direct ancestors of part of our trees originally inhabited the interior of the polar circle, while many others, confined now to southern countries, once had European