Floral Evolution. 
8 1 
parts of oligomerous ancestors, in a manner analogous to the 
partition of the vegetative leaves. 
It is here suggested that this tendency may be due to the 
extension of the active branching of the floral axes to form dense 
aggregates of flowers to the parts of the individual flower itself. 
It may, of course, be objected against such a supposition that in 
the groups where the highest expression of floral aggregation is 
reached (the Compositte, for example) there is no indication of 
chorisis. In reply to such an objection, we may premise that it is 
in the less specialised Archichlamydeae that the principle of 
economy, and the tendencies which immediately subserve it, is 
seen in the fullest activity of its working ; while in the Sympetala?, 
as we hope to show, oligomery is virtually a fixed character, pro¬ 
gression taking principally the line of adaptation to insect visitors. 
In other words, the polypetalous forms constitute, as it were, an 
experimental ground for evolution ; in the Sympetalae, on the other 
hand, the tendencies are more fixed and definite; and in the 
Composite, which comprise considerably more than one quarter 
of the total number of species of Sympetalac, the climax of oli¬ 
gomery combined with insect-adaptation has been attained, the 
enormous number of species of this very natural group affording 
indisputable evidence of its success in the struggle for existence. 
Returning now to the enumeration of the subsidiary tendencies, 
we may dismiss the first, conspicuousness, with the comment 
that it subserves, of course, the second fundamental principle, 
that of adaptation to insect-visitors. 
(b). Devices of floral structure and habit, in obvious relation 
to insect-visits, represent a second tendency subsidiary to our 
second fundamental principle; and pre-eminent among these, as 
being of the most general occurrence, is the tendency to 
zygomorphy. This, we shall find, is a condition relatively rare in 
the Archichlamydeae as compared with the Sympetalse—as we 
should reasonably expect, according to the lines we have laid down. 
Zygomorphy may be the direct result, so to speak, of insect-visits, 
as is the case with zygomorphic flowers which are relatively large 
and solitary, or, at any rate, members of more or less lax inflor¬ 
escences. More generally, however, zygomorphy appears to follow 
upon aggregation into dense inflorescences, the outer florets 
becoming more or less strongly zygomorphic, largely as the direct 
result of the mechanical conditions of development. Examples of 
this will present themselves readily to the reader’s mind ; but the 
