70 MESSRS. NEWELL ARBER AND J. PARKIN ON 
What, then, is the origin of this type of foliage? We believe the solution 
to the question is to be sought for in a study of the branch-habit. 
Wieland* has shown clearly that the Bennettiteæ possessed stems of restricted 
vertical growth, either unbranched or branched only to a limited extent. 
The same also seems to have been true of the Pteridospermes, their ancestors. 
On the other hand, one of the great characteristics of the Angiosperms, as a 
whole, is their free branching, whether of the monopodial or thesympodial system. 
With this change in habit was probably correlated a. general alteration in the 
character of the foliar organs. The Pteridospermez, with their unbranched, 
or tree-fern-like habit, obtained a considerable assimilatory surface by means 
of very large leaves. — Probably, for mechanical reasons, the inerease in the 
size of the leaf as а whole would have to be accompanied by much subdivision 
of the lamina. Hence the highly compound fronds of the Paleozoic period. 
The large, but simpler foliage of the Bennettiteæ, and of the Cycadophyta 
generally, can easily be derived from this type of leaf, and is likewise corre- 
lated with a non-branched or feebly branched habit. The association of 
megaphylly and a simple stem is found in certain living Angiosperms, 
е. g., the Palms, where it may perhaps be regarded as an ancient 
feature. 
Thus in the Angiosperms as a class, free branching and small leaves have 
been substituted for a simple unbranched habit and large leaves. One can 
readily understand how, as the tendency to branching increased, the necessity 
for mierophylly would arise and smaller foliage be evolved. In the one case 
branching takes place, as it were, in the leaf, in the other, in the stem. 
Both represent efficiency from a physiological standpoint attained by 
different methods. 
The theory that the origin of the Angiospermous type of branching and 
consequently the prevailing leaf-form, took place some considerable time after 
the evolution of the primitive flower is in harmony with the axiom (see p. 35) 
that corresponding stages in the evolution of the various organs of a seed- 
plant are not reached contemporaneously. It also explains certain facts 
which have hitherto been regarded as highly mysterious. When we attempt 
to summarise the existing data relative to the first appearance of Angiosperms 
in Neocomian rocks, we are led to three remarkable conclusions. In the 
first place, the Angiosperms appear toarise very abruptly or suddenly. In the 
second, judging by their detached leaf-impressions, our sole evidence at present, 
they belong to highly evolved, and still existing natural orders. There 
appears to be nothing primitive about these early forms. In the third place, 
from their first incoming, they are the dominant types in the vegetation of 
the Cretaceous and Tertiary periods. 
* Wieland (1906) Chapter II. 
