Floral Evolution. 3$ 7 
nection with our ranalian branch-system could only be traced 
through a pre-dicotyledonous, or even pro-angiospermous ancestor 
—the product of such lines as Dd x , Dd 2 , Aa, in the diagram ? 
It is scarcely likely that all the known Archichlamydeae have 
descended from a ranalian ancestor ; although not a few represent 
in all probability the reduced and specialized progeny of familar 
groups of both Polypetalae and Sympetalas. Indications of the 
progress to “ apetaly ” are to be found in many typically hetero- 
chlamydeous families : one such case we have already noticed in 
connection with the Oleacete ( Fvaxinus ); and there are many 
others, from among Ranales upwards. 
We have found places for all the Sympetalae in the branches 
of the evolutionary tree of the ranalian ancestry; but in an investi¬ 
gation based so largely upon mere speculation and so little upon 
actual fact the scope for error is great indeed. As we have 
insisted from time to time in these chapters, we cannot aspire to 
more than a few not unreasonable suggestions as to the general 
mode in which evolution has operated. 
It is almost certain that some of the Archichlamydeae at least 
have diverged from an ancestry other than the ranalian. Perhaps 
the most convincing evidence of this is afforded by the priority, 
in the fossil-record, of the catkin-bearing trees, the so-called 
Amentiferae. In Permian and Triassic times coniferous trees 
probably stood in the van of the vegetable kingdom. Cycads appear 
in the Oolite, accompanied, perhaps, by screw-pines, grasses and 
other wind-pollinated monocotyledonous plants. Dicotyledons are 
extremely rare even in strata at the base of the Cretaceous system, 
in many cases deposited, like the Wealden, under the most favour¬ 
able conditions for the preservation of organic remains ; and it is 
not until we reach the later Cretaceous beds that dicotyledonous 
plant-remains occur in any quantity. Our chief concern is with 
the fact that these earliest remains are prevailingly of amentiferous 
trees. 
Their priority in the geological succession has been used as an 
argument to support the idea of the relative primitiveness of the 
Amentiferas; and a similar argument has been applied even to the 
Monocotyledons. 1 The probable position of these groups in the 
history of descent will, it is hoped, be clear from our foregoing 
1 It is probable, however, that the first dicotyledonous fossil is 
as old geologically, if not older, than the first Monocotyledon 
(see Reward, Notes on the Geological Histovy of Monocotyledons, 
Ann. Bot., X, 2t)5). 
