FERTILISATION OF VARIOUS FLOWERS EY INSECTS. 
109 
faces downwards, that is, away from the cavity of the bell. The- 
style rises from a superior ovary, under which is a disk that 
secretes a fluid highly attractive to all kinds of bees. From 
beneath this spring eight stamens, which soon converge and end 
in a circlet of anthers set closely round the style (fig. 8). Each 
anther consists of two cells, adherent in their lower and middle- 
parts, slightly divergent at their apices. From the base of each 
cell a curious flat process is thrown out almost horizontally, 
towards the inner wall of the corolla, which it nearly touches. 
Thus when one looks into the mouth of the bell one sees sixteen 
processes radiating from the centre like the spokes of a wheels 
and forming an imperfect barrier between the upper and lower 
halves of the bell. The other ends of the two cells of each 
anther, as already said, diverge slightly, and near the apex of 
each, on the external lateral aspect, is an oval opening or pore 
(fig. 9), which gives issue to the pollen. As the whole corolla, 
hangs upside down, the pollen grains would at once fall out from 
these pores as soon as dehiscence occurred, were there nothing 
to prevent it. But even then it is clear that the grains would 
not light on the viscid surface, for this faces in the wrong direc- 
tion. The pollen, however, is prevented from so falling ; for 
each anther cell adheres just in the part where its opening i& 
situated to the corresponding part of the adjoining cell of the 
next placed anther in the circlet. Thus the pore of a cell, say 
the light cell of an anther, is, so to speak, closed by the pore of 
the left cell of the next adjoining anther; and so on all the way 
round. Thus the pollen is kept imprisoned in the cells by their 
mutual adherence. A very little force, however, is enough to 
dislocate this chain of anther cells. If a slight pressure be 
made down upon one of the radiating processes, which form the- 
long arm of a lever of which the filament is the fulcrum, the 
upper ends of the corresponding anther cells, which form the- 
other arm, are lifted up, and break from their union with their 
neighbours. The pores are then disclosed, and the pollen 
grains fall out in obedience to gravity. When a bee visits a 
flower it pushes its head against the mouth, which is too small 
to admit it, but which it obstructs completely. In so doing the- 
head necessarily strikes against the viscid stigma, and will leave 
on this any pollen grains that it may have. The bee then 
extends its proboscis down to the disk to suck the nectar. The 
proboscis can scarcely get there without striking one or other of. 
the processes, and as soon as it does this the jointed chain of 
anther cells, as before explained, is dislocated, and a shower 
of pollen falls from the exposed pores on the insect’s head. This- 
the bee carries of, and will, of course, give up to the stigma of' 
the next flower it visits. 
In our botany manuals the structure of the individual anther- 
