PART IV. | GENERAL RETROSPECT, “ 589 
how perfect is their structure. Senecio vulgaris, Veronica 
hederefolia, Stellaria media, Lamium purpureum are examples of 
the one condition, and Pedicularis silvatica, Malva silvestris, and 
Echium vulgare of the other. It must by no means be supposed 
that all floral mechanisms are equal; in many flowers distinct 
imperfections have already been pointed out (Posoqueria fragrans, 
Faramea, Malva silvestris, Euphrasia Odontites, Geum rivale, 
etc.). There are also species which, after abandoning the power of 
self-fertilisation, have suffered so from the competition of other 
flowers that they obtain very few insect-visits (¢.9., Ophrys 
muscifera). Such instances prove still more forcibly that the 
uniform perfection which Axell supposes to exist in Nature has 
no real existence, 
It may be urged that Axell proceeds from a totally different 
conception of perfection; for he looks upon every economy of 
space, time, and material as a step towards perfection ; and he 
must therefore look upon the reversion of monoclinic flowers to 
diclhny, or the passage of homogamic flowers to dichogamy, as 
retrograde even when those changes are of distinct advantage for 
the sexual propagation of the plant. But such a definition of 
perfection is unnatural. 
But even apart from any definition of perfection in a floral 
mechanism, nothing can be more unnatural than to assert that 
there is only one course of evolution or only one path towards 
perfection in the structure of flowers. 
As in regard to conspicuousness, odour, or production and 
concealment of food material, so also in favouring or insuring 
cross-fertilisation, there is an infinite variety of means to the 
end.. In flowers which receive abundant insect-visits cross-fertilisa- 
tion has been insured sometimes by reversion to dicliny, sometimes 
by dichogamy, sometimes by distant separation of the stigma and 
pollen in the flower, sometimes by a special mechanism causing 
the pollen to adhere to the insect and afterwards to the stigma of 
the next flower. Dichogamy, though usually proterandrous,, is 
proterogynous even in many entomophilous flowers (Aristolochia, — 
Hwonymus, many Rosacece), in spite of Axell’s statement to the 
contrary. ‘The separation in space of the essential organs may 
be of the same kind in all the flowers of a species, or the relations 
of the parts may be different in different flowers; but in all cases 
the relative positions of the parts are closely connected with in- 
sects visits. In the former case, when all the flowers are alike, 
eross-fertilisation is insured either if the same part of the insect 
