I have repeatedly seen Bombus hortorum, lapidarius, and a third species, as well as hive-bees, sucking the flowers of this violet: I marked six which were thus visited, and four of them produced fine capsules; the two others were gnawed off by some animal. I watched Bombus hortorum for some time, and whenever it came to a flower which did not stand in a convenient position to be sucked, it bit a hole through the spur-like nectary. Such ill-placed flowers would not yield any seed or leave descendants; and the plants bearing them would thus tend to be eliminated through natural selection.
The seeds produced by the cleistogamic and perfect flowers do not differ in appearance or number. On two occasions I fertilised several perfect flowers with pollen from other individuals, and afterwards marked some cleistogamic flowers on the same plants; and the result was that 14 capsules produced by the perfect flowers contained on an average 9.85 seeds; and 17 capsules from the cleistogamic ones contained 9.64 seeds,--an amount of difference of no significance. It is remarkable how much more quickly the capsules from the cleistogamic flowers are developed than those from the perfect ones; for instance, several perfect flowers were cross-fertilised on April 14th, 1863, and a month afterwards (May 15th) eight young cleistogamic flowers were marked with threads; and when the two sets of capsules thus produced were compared on June 3rd, there was scarcely any difference between them in size.
Viola odorata (WHITE-FLOWERED, SINGLE, CULTIVATED VARIETY).
The petals are represented by mere scales as in the last species; but differently from in the last, all five stamens are provided with diminutive anthers. Small bundles of pollen-tubes were traced from the five anthers into the somewhat distant stigma. The capsules produced by these flowers bury themselves in the soil, if it be loose enough, and there mature themselves. (8/6. Vaucher says 'Hist. Phys. des Plantes d'Europe' tome 3 1844 page 309, that V. hirta and collina likewise bury their capsules. See also Lecoq 'Geograph. Bot.' tome 5 1856 page 180.) Lecoq says that it is only these latter capsules which possess elastic valves; but I think this must be a misprint, as such valves would obviously be of no use to the buried capsules, but would serve to scatter the seeds of the sub-aerial ones, as in the other species of Viola. It is remarkable that this plant, according to Delpino, does not produce cleistogamic flowers in one part of Liguria, whilst the perfect flowers are there abundantly fertile (8/7. 'Sull' Opera, la Distribuzione dei Sessi nelle Piante' etc. 1867 page 30.); on the other hand, cleistogamic flowers are produced by it near Turin. Another fact is worth giving as an instance of correlated development: I found on a purple variety, after it had produced its perfect double flowers, and whilst the white single variety was bearing its cleistogamic flowers, many bud-like bodies which from their position on the plant were certainly of a cleistogamic nature. They consisted, as could be seen on bisecting them, of a dense mass of minute scales closely folded over one another, exactly like a cabbage-head in miniature. I could not detect any stamens, and in the place of the ovarium there was a little central column. The doubleness of the perfect flowers had thus spread to the cleistogamic ones, which therefore were rendered quite sterile.
Viola hirta.
The five stamens of the cleistogamic flowers are provided, as in the last case, with small anthers, from all of which pollen-tubes proceed to the stigma. The petals are not quite so much reduced as in V. canina, and the short pistil instead of being hooked is merely bent into a rectangle. Of several perfect flowers which I saw visited by hive-and humble-bees, six were marked, but they produced only two capsules, some of the others having been accidentally injured. M. Monnier was therefore mistaken in this case as in that of V. odorata, in supposing that the perfect flowers always withered away and aborted.