If we pass from the consideration of details of structure, and of botanical affinities, to a general survey of the mineralized remains of the vegetable kingdom, we perceive that from the palæozoic deposits, to those which are contemporaneous with the human race,--from the coal-measures to the peat-bogs of modern times,--vast accumulations of vegetable matter, in various states of carbonization, have been produced from the imbedded relics of the terrestrial floras that flourished during the respective periods of their formation; petrifaction, or the transmutation of vegetable tissues into stone, from the infiltration of siliceous, calcareous, or metallic solutions, being an accidental process, dependent on the physical conditions under which the trees and plants were submerged, and entombed in the strata. Although the entire system of vegetable life which prevailed during the earlier ages of the world is but partially revealed by the fossil remains which geological researches have brought under the examination of the naturalist,--for numerous tribes of plants may have existed of which no traces have been detected, while of species of delicate tissues all vestiges may have perished,--yet a review of the facts hitherto obtained, presents some highly important and unexpected results, as to the characters of the successive floras which prevailed during the palæozoic, secondary, and tertiary epochs. And though deductions of this nature must, in the present state of our knowledge, be regarded in the light of shifting hypotheses to be modified or abandoned with the progress of discovery,--yet the predominating types which characterize the flora of one system of formations, differ so essentially from those of another, that it may be reasonably inferred such apparent distinctions are the effect of organic laws, and not illusions arising from our misinterpretation of the natural records of former conditions of the vegetable world. The absence in the most ancient deposits of the entire class of Angiosperms, or flowering plants, which constitutes the leading features of the floras with which we are familiar,--the abundance of unknown types of Cryptogamia, and the extinction or disappearance of those tribes in the succeeding formations, and the prevalence of new species and genera belonging to another class;--the predominance in one flora, both in number and variety, of certain tribes, and their decadence in the next period; while a family subordinate in the antecedent epoch, and known but by a small number of species, suddenly acquires a pre-eminence both in numbers and variety;--are phenomena, which the facts brought before us in the course of this argument, present in a striking point of view.