118 THE NATURE-STUDY OF PLANTS 
rate clear that self-pollination must not be ruled out 
altogether from the life-history of the Herb Robert. 
Incidentally they teach us that we must not infer 
from what does or does not take place in June 
and July what will happen at a later stage, but 
that we must go on experimenting patiently until we 
have exhausted every possibility. Since self-pollina- 
tion as a rule does not take place until the life of the 
flower is nearly at an end, it is perhaps possible that 
in some species it is withheld until the plant itself is 
nearing old age, and, as already pointed out, the 
Herb Robert does not long survive its one and only 
flowering period from April to October. 
However much or little self-pollination there may 
be, however, the arrangements, as a whole, leave 
little to be desired, for the plant sets an abundance 
of seed. 
After the pollination of the stigma the sepals part 
company to a small extent and the petals fall, whole 
and unwithered, and then the sepals close up again 
and help to protect the maturing ovary : the stamens, 
or rather their antherless filaments, do not fall, but 
they also enwrap the ovary and, like the calyx, keep 
pace with its increasing size. 
Before I leave the flower I must point out the three 
light lines upon the petals; they form, as it were, the 
prongs of a white long-handled fork, and they are 
supposed to guide the insect visitor in the right way 
to the honey at the bottom of the calyx, but that is 
a very difficult supposition to prove, although I am 
by no means concerned to dispute it. 
I may mention, too, that the blossom is so compact 
and its entrance so narrow, that raindrops cannot 
enter and spoil the pollen or the honey, while it is 
