310 THE FERTILISATION OF FLOWERS. [PART IIL. 
the tube widens out superiorly, and the more so the longer it is, so 
that many short-lipped insects can reach the honey by creeping a 
greater or less distance down the tube. The honey is secreted by 
the upper surface of the ovary, and is lodged in the base of the 
tube; in spite of the funnel-shaped widening of the tube, it is 
sheltered from rain by hairs lining the tube. The pollen also is 
easily accessible, for the stamens, which are at first curled up in 
the bud, straighten themselves after the flowers open and protrude 
4 to 5 mm. beyond the tube, with the pollen-covered faces of the 
anthers directed upwards. Thus the flower offers great attractions 
to insects whether in quest of honey or of pollen, and all the more 
that the close aggregation of the flowers prevents waste of time in 
passing from one to another. Honey-seeking insects are induced 
to make repeated visits to the same capitulum by the gradual 
development of the florets and by the length of time during which 
they secrete honey; pollen-seeking insects are led likewise to 
repeat their visits, because the anthers in each flower do not 
develop simultaneously but one by one (Fig. 108, 1, 3, ¢, d). 
The style elongates after all the stamens have fully developed ; 
its immature stigma stood previously in the mouth of the flower, 
but now protrudes 4 to 5 mm. from the corolla and becomes 
mature (a, 2). The empty anthers, if they have not been removed — 
by the insects, as frequently happens, are retracted to the mouth 
of the flower by the filaments shrivelling up (d, 2). In each 
capitulum the florets open and the anthers develop in succession 
from the circumference towards the centre; but the styles only — 
begin to elongate and the stigmas to develop after all the stamens — 
on the capitulum have completed their development, and then 
almost simultaneously throughout the capitulum. Since here the 
whole capitulum is at first male, and later, if insect-visits have — 
taken place, wholly female, cross-fertilisation takes place regularly — 
between separate capitula in case of insect-visits; and since all the — 
stigmas on a capitulum develop simultaneously, while the anthers — 
succeed each other very gradually, cross-fertilisation en masse may 
be effected by a single insect’s visit, and the possibility of cross- 
fertilisation is extended over a long period of time. So even a — 
period of fine weather very short in comparison with the whole 
flowering period of the plant may suffice for complete fertilisation 
of the flowers. 
Besides plants with hermaphrodite flowers, others occur whose 
flowers all possess more or less aborted anthers which do not dehisce 
and which contain small and often very deformed pollen-grains. 
x 
' ~~ 
Pee ee I I 6 Reger OA we Sk - 
