Floral Evolution. 
39 1 
varying relative position of the ovary among the Calycifloraae, 
variability which has brought epigyny into such sad disrepute as a 
classificatory character, is replaced by the constant epigyny of 
Rubiales, Cucurbitales, Composite, etc. 
Inconstant characters which are phyletically important for a 
group as revealing its tendencies may have to be disregarded in 
sub-dividing that group. A good example of this is the oft-quoted 
one of the Rosales; for the same genus—e.g., Saxifraga —may 
include both hypogynous and epigynous forms; so that the relative 
position of the ovary, important as shewing the fundamental 
tendency of the Rosales, is often disregarded in the internal 
classification of this order. 
We conclude, in any case, that characters which, by reason of 
their inconstancy in one group, must be disregarded for the purposes 
of classification, may become critical characters among the 
descendants of that group. Herein lies an explanation of the 
differing value of the same character in different circles of affinity ; 
and also an important modification of the principle that the critical 
value of a character is in direct proportion to its constancy. 
Progress. By continuance of the process of uniting groups by 
lines of descent, we may endeavour to build up the evolutionary 
tree, and so construct a natural scheme of classification. 
Now, if Variation and Natural Selection be actual and operative 
phenomena of the organic world, it follows that the course of 
evolution has been, speaking generally, in the direction of progress. 
But the question at once presents itself—What is involved in the 
idea of evolutionary progress ? 
In view of the great difficulties and the host of side-issues which 
this question presents in its application to the Vegetable Kingdom 
at large, it will be well at the outset to confine our considerations 
to that branch of flowering plants which have formed the subject of 
our studies, namely, the Sympetalae and their immediate ancestors. 
We may conclude from these studies that, limited to this field, at 
any rate, the idea of general progress involves, first, progressively 
greater efficiency for the business of the vital functions, coupled 
with, second, progressively diminishing “ effort ” or output to secure 
that efficiency. 
The combination of these two factors is reflected in our two 
fundamental tendencies of Economy in production of reproductive 
parts and of Adaptation to Insect Visits. It is reflected also in the 
