FLORA. 223 
extension of the dicotyledons. The region in which the 
plants of this class had their first cradle, without being 
situated in the immediate vicinity of the Pole, could touch 
it, however, and communicate at the same time with the 
zones further south. Some day, it is to be hoped, we 
shall be able to fix the geographical location and probable 
limits of this mother region of the first dicotyledon; at 
present the data are too vague to allow us to think of 
insisting further. on this point. I have been desirous, 
however, of investigating whether the families of the most 
ancient dicotyledons, and those the presence of which in 
the chalk age have been determined in the manner the 
least doubtful, present in themselves any character which 
would prove their antiquity. In regard to this the fre- 
quency and the diffusion of the polycarpic plants, magno- 
laceae, menispermeae, perhaps berberidaceae, heliboreae, 
nympheaceae and malvaceae, have not passed unobserved, 
as the excessive development of certain parts in many of 
these families have not. ceased to reproduce and multiply 
the types and the subtypes within each of the species, 
This elaboration has gone on across the last part of the 
chalk period and throughout the whole of the tertiary, and 
it continues still in the heart of the polymorphic and 
floating groups which drive botanists to despair.’ 
Thus are we brought to the close of one of the great 
divisions of geologic eras. 
The rocks in which are found the fossil remains which 
have been under consideration latterly have been called 
secondary in contradistinction to the granite, gneiss, and 
other underlying rocks which have been: designated primary 
rocks. But there are rocks of later formation, which it 
has been found convenient to distinguish from those again ; 
and to these the designation ‘ tertiary’ has been given. 
The fossils found in the older rocks presented little 
analogy, often no resemblance, to existing plants and 
animals; here, however, the similitude is frequently so 
complete, that the naturalist can scarcely point out a 
