Hermann Muller gives an elaborate description of the flowers 'Befruchtung' etc. page 247.) Notwithstanding these manifest provisions for cross-fertilisation, varieties which have been cultivated for very many successive generations in close proximity, although flowering at the same time, remain pure. I have elsewhere given evidence on this head, and if required could give more. (5/13. 'Variation of Animals and Plants under Domestication' chapter 9 2nd edition volume 1 page 348.) There can hardly be a doubt that some of Knight's varieties, which were originally produced by an artificial cross and were very vigorous, lasted for at least sixty years, and during all these years were self-fertilised; for had it been otherwise, they would not have kept true, as the several varieties are generally grown near together. Most of the varieties, however, endure for a shorter period; and this may be in part due to their weakness of constitution from long-continued self-fertilisation.
It is remarkable, considering that the flowers secrete much nectar and afford much pollen, how seldom they are visited by insects either in England, or, as H. Muller remarks, in North Germany. I have observed the flowers for the last thirty years, and in all this time have only thrice seen bees of the proper kind at work (one of them being Bombus muscorum), such as were sufficiently powerful to depress the keel, so as to get the undersides of their bodies dusted with pollen. These bees visited several flowers, and could hardly have failed to cross-fertilise them. Hive-bees and other small kinds sometimes collect pollen from old and already fertilised flowers, but this is of no account. The rarity of the visits of efficient bees to this exotic plant is, I believe, the chief cause of the varieties so seldom intercrossing. That a cross does occasionally take place, as might be expected from what has just been stated, is certain, from the recorded cases of the direct action of the pollen of one variety on the seed-coats of another. (5/14. 'Variation of Animals and Plants under Domestication' chapter 11 2nd edition volume 1 page 428.) The late Mr. Masters, who particularly attended to the raising of new varieties of peas, was convinced that some of them had originated from accidental crosses. But as such crosses are rare, the old varieties would not often be thus deteriorated, more especially as plants departing from the proper type are generally rejected by those who collect seed for sale. There is another cause which probably tends to render cross-fertilisation rare, namely, the early age at which the pollen-tubes are exserted; eight flowers not fully expanded were examined, and in seven of these the pollen-tubes were in this state; but they had not as yet penetrated the stigma. Although so few insects visit the flowers of the pea in this country or in North Germany, and although the anthers seem here to open abnormally soon, it does not follow that the species in its native country would be thus circumstanced.
Owing to the varieties having been self-fertilised for many generations, and to their having been subjected in each generation to nearly the same conditions (as will be explained in a future chapter) I did not expect that a cross between two such plants would benefit the offspring; and so it proved on trial. In 1867 I covered up several plants of the Early Emperor pea, which was not then a very new variety, so that it must already have been propagated by self-fertilisation for at least a dozen generations. Some flowers were crossed with pollen from a distinct plant growing in the same row, and others were allowed to fertilise themselves under a net. The two lots of seeds thus obtained were sown on opposite sides of two large pots, but only four pairs came up at the same time. The pots were kept in the greenhouse. The seedlings of both lots when between 6 and 7 inches in height were equal. When nearly full-grown they were measured, as in Table 5/57.
TABLE 5/57. Pisum sativum.
Heights of plants measured in inches.