140 THE NATURE-STUDY OF PLANTS 
it provides examples, as we pass from the large to 
the small flowered species, of arrangements that ex- 
clude the possibility of spontaneous self-pollination, 
of others in which both kinds occur commonly, 
and of still others in which self-pollination is the 
rule rather than the exception. Thus the Meadow 
Cranesbill cannot be self-pollinated except by the 
merest and most unlikely accident because there is 
no overlap, or, in other words, the pollen is all shed 
before the stigmas of the same flower are mature; in 
the small-flowered species there is a decided overlap 
and a good deal of self-pollination ; while in the Small 
Dovesfoot the five anthers mature almost at the 
same time as the stigmas, and self-pollination seems 
to be the more usual mode. ‘There is a short-lived 
separation of the sexes, but in this and in some other 
of the annual species, all of which are small-flowered, 
it is the stigmas that mature first, whereas in the 
perennials, including the somewhat small-flowered 
Mountain Cranesbill, it is the anthers. The annual 
species are the less stable of the two in this respect. 
It is not easy to see or to suggest any advantage 
in the earlier maturity of either of the organs, unless 
it be that the laggard has the shorter period of 
functional activity, and so is less likely to miss an 
opportunity because its partner is not yet ready. IH, 
for example, it could be proved that in the Bloody 
Cranesbill the life of the pollen grains is longer than 
that of the stigmas, whereas in the Small Dovesfoot 
it is shorter, one might admit an advantage in the 
earlier maturity of the anthers of the former and the 
stigmas of the latter, but I have scarcely a particle 
of evidence in support of this possibility, nor would 
it be easy to put it to the test without the skill of 
