586 THE FERTILISATION OF FLOWERS. [PART IY. 
say in all generally accessible flowers, the pollen is rough or viscid, 
so that it adheres readily to the usually hairy surfaces of insect- 
visitors. In Cypripedium it forms a sticky mass, which adheres to 
the insect as it forces its way underneath, and which is afterwards 
removed by the rough stigmatic surface of the next flower. In 
Orchis it coheres in little packets which are united into “ pollinia ” 
or “pollen-masses” by elastic threads, and the whole pollen- 
mass attaches itself by a special cement to the insect-visitor; 
the stigma is sticky enough to cause the little packets of pollen 
which come in contact with it to adhere firmly, so that as the 
insect moves away the elastic threads are broken through. In 
Asclepias all the pollen-grains in an anther-lobe cohere to form 
a flattened mass, which becomes attached by its “corpusculum ” 
to an insect’s claw, and is inserted through a narrow slit into 
the stigmatic chamber of another flower, and is torn away and 
left there. 
The size of the pollen-grains also varies in relation to the 
length of the style which the po!len-tubes must traverse, as is seen 
in the varying size of pollen-grains from the different whorls of 
anthers in dimorphic and trimorphic plants (cf. Lythrum),. 
Characters which insure Cross-Fertilisation in case of Insect- Visits 
and Self-Fertilisation in Absence of Insects. 
If the adaptations of flowers to insect-visits are really due to 
the cross-fertilisation induced by the visitors, all characters which 
render cross-fertilisation inevitable in case of insect-visits, and 
which render self-fertilisation impossible, must be of special advan- 
tage to the plant ; but only so far as insect-visits occur in sufficient 
abundance to insure cross-fertilisation. If this condition is not 
regularly fulfilled, it is obviously much better for the plant 
to have the power of reproducing itself by self-fertilisation 
while leaving only a possibility of cross-fertilisation in case of 
insect-visits than to be cross-fertilised with absolute certainty when 
insect-visits occur, but to be not fertilised at all in absence of 
insects. This is the simple explanation of the law already laid 
down, that “When closely-allied flowers, alike in other respects 
in structure, differ in the abundance of their insect-visits, and at 
the same time in the degree to which cross-fertilisation is insured 
in case of insect-visits and self-fertilisation in absence of insects, 
then under otherwise similar conditions those flowers which are 
most visited are those in which cross-fertilisation is best insured, and 
EEE Ee —s 
