V4 
‘e 
| 4 
D f 
? 
5 
: 
y 
PART III. } THE MECHANISMS OF FLOWERS. 341 
weather they never fail) the pollen is in a short time all removed. 
But if insects do not visit the flower until the stigmatic edges of 
the branches of the style have already bent outwards, then self- 
fertilisation is possible, since in this case the hairs of the style 
_ still remain covered with pollen. If insects do not visit the flower 
at all, some of the pollen-grains, which hang in little clumps 
Fie. 114,.—Cnicus arvensis, L. 
1.—Flower, in first stage ; pollen issuing from the anther-eylinder. 
2.—Pollen-grain (x 400). 
3.—Ditto, in optical section. 
4.—Flower, in second stage. The pollen has been removed by insects and only a few grains 
remain upon the style; the style projects far beyond the anther-eylinder, and its stigmatic branches 
are closely applied together. 
__ 5.—Upper part of style with its ring of long hairs (bb), and lower part of its two stigmatic 
¢ peeceee; the latter are covered outside with short hairs, and on the edges with stigmatic papille 
a ). 
_. 4, anther-cylinder; b, ring of long hairs; ¢, short hairs; d, stigmatic papil’e; e, pollen.~ 
~~ 
to the hairs, may easily fall of themselves upon the stigmatic 
_ papillee. In fine weather, and in the open air, this can scarcely 
_ ever take place, for Cnicus arvensis is one of the most abundantly 
visited of all our native plants. ‘As the following list shows, very 
_ few insects resort to it for the sake of its pollen, but very many for 
its honey :— 
