MR. R. A. ROLFE ON THE APOSTASIER. 223 
or obscured in those stages where the greatest amount of specia- 
lization is developed. Adaptive characters may (and do) become 
ancestral ones if sufficiently beneficial to give rise to a dominant 
group of organisms, the adaptive characters being handed down 
to all the descendants in common. But when once a group 
becomes dominant, and therefore widely diffused, some of its 
members invariably come under new conditions of environment ; 
still newer adaptations arise ; the group begins again to diverge 
in various directions; and the non-variable characters are now 
easily recognized as the ancestral ones. 
We now proceed to apply these principles in discussing the 
affinities of the Apostasiee, and by the aid of the two accompanying 
diagrams (figs. 1 and 2, page 224) to show their relationship 
with surrounding groups. 
lt is very probable that the ancestral Monocotyledonous 
prototype was an apocarpous plant of very simple structure, 
destitute of perianth, and probably more nearly allied to Panda- 
nacea than to any other existing order—a conclusion based upon 
structural grounds and supported by paleontological evidence. 
Commencing from this common starting-point, the broad features 
of the evolution of existing Monocotyledones may be pretty 
closely traced ; though the exact point of divergence of many of 
the branches from the primary Monocotyledonous stem, and from 
each other, is a point on which much difference of opinion exists, 
and the real affinities of a few Orders are not yet at all conclusively 
settled. At the base of the series occurs the Nudiflore, a com- 
paratively simple group which has not departed far in its essential 
characteristies from the primary Monocotyledonous type, and 
which, together with the Apocarpe and the natural orders they 
comprise, probably represent diverging ramifications of the same 
early branch. From a point somewhere near the angle of diver- 
gence of the previous group may be traced another branch which 
afterwards separated into three ramifications, the Glumales on 
the one hand, the Calycine and Coronariee on the other. The 
passage between these groups, and their subsequent ramifications 
into Orders, is, for the most part, so gradual that it seems tolerably 
clear they had one common origin, afterwards diverging in various 
directions. Lastly may be mentioned the Epigyne, though it is 
doubtful if this group had one common origin. The Amaryllide 
and Bromeliads at least appear to have arisen from the same 
branch which produced the Liliaceae; and it seems probable 
