GRASSES AND SEDGES 127 
from a neighbouring flower, and the whole mechanism of 
the plant is arranged to secure that end, self-fertilisation 
being avoided by the simple provision that before the 
anthers shed their pollen the stigmas are too old to 
receive it. In the first place, the carrion smell of the 
plant attracts flies, which come hoping for a feed on their 
particular luxury. Let us suppose that they have visited 
an Arum plant, been dusted with pollen in the way 
I shall soon describe to you, and have flown off to 
another. The open sheath shows them the way in, and 
they easily push their way downwards through the 
bristles, and look for their expected carrion. Round 
and round the bottom they run and of course leave 
the pollen they have brought upon the stigmas, over 
which they climb. Thus the first part is secured, but 
the second remains. These flies are wanted to carry 
away pollen to other plants, and, if the stigmas were 
ripe to receive pollen, the anthers could not be ripe to 
shed it. Now comes the office of the bristles. The flies, 
in disgust at finding no carrion, try to get up, but the 
bristles point against them and they cannot force their 
way up. There they must remain until the anthers ripen. 
When these have split, and dusted them thoroughly, two 
modifications are seen at the same time in the flower. 
Honey is produced at the stigmas, which encourages the 
flies to seek other arum flowers in spite of their experi- 
ence, and the hairs or bristles, now their work is done, 
wither away! So the flies are released to repeat their 
experiment. 
Now, whether these flowers were created with these 
devices at the very beginning, or whether through count- 
less generations and by infinite experiment these various 
forms have been evolved from one common origin, this is 
