114 
But it may be the office of these tiny insects to effect self- 
fertilization in flowers that do not often receive visits from larger 
‘insects, for whose visits they are, however, functionally contrived. 
They bring about results which, though inferior, are greatly 
superior to failure of pollenation. Though the flowers are proter- 
androus, the round and unbroken grains of pollen are still visible 
on the outer side of the style some time after its parts have 
expanded and exposed the stigmatic surfaces. The Thrips will 
be useful in carrying this pollen across to the other side, and in 
this way overcome a difficulty which the pupil of Linnzus 
imagined must be obviated by his supposed canals. 
Both the ability and desirability of self-fertilization in case of 
a lack of insect-visitation have been recognized by Miiller and other 
writers as probable in species of Campanula.* He indeed calls it 
a makeshift (méglichkeit), but as such it is far better than failure 
in accomplishing the development of seed. He mentions C. 
barbata, L., as an example of this kind, where, if cross-fertilization 
fails, self-fertilization may take place. The method of securing this 
is for the parts of the style to curve around till their stigmatic faces 
come in contact with the lower part of the brush formed by the 
undivided portion of the style, or by the pollen falling on the 
stigmas when they are in this position. The parts of the style 
would have to curve into a circle to bring the two sets of hairs, 
those on the brush and those on the stigma, into this relation. 
But they do not appear to curve so far in the flowers of C. 
aparinoides. This condition of affairs also implies that the pollen 
of a flower is available for its own fertilization when the parts of 
the style expand and long enongh afterward to allow them to 
curve to make the required contact. And this is corroborated 
by finding fresh pollen on the brush and on the Thrips, which, 
though it might have been brought from other flowers, is more 
likely to have been left for use by their own anthers. And 
it seems much easier and more satisfactory to make the Thrips 
‘The Campanulacee have been inquired into by many investigators as a supposed 
case of necessary self-fertilization, some of whom, Cassini, Treviranus, Hartig, and 
Gaertner, decide for it ; others, Sprengel, Wiegman and Henschel, against. Accord- 
ing to my investigations, which, however, are to be extended by more exact experi- 
ments, self-fertilization does not take place.”—Hildebrand, Geschlechter Vertheilung 
bei den Pflanzen, p. 64, (Leipzig, 1867.) d 
