Other facts could be given, but these will suffice for us. Naturalists formerly attributed all these results to the difference between species being fundamentally distinct from that between the varieties of the same species; and this is still the verdict of some naturalists.
The results of my experiments in self-fertilising and cross-fertilising the individuals or the varieties of the same species, are strikingly analogous with those just given, though in a reversed manner. With the majority of species flowers fertilised with their own pollen yield fewer, sometimes much fewer seeds, than those fertilised with pollen from another individual or variety. Some self-fertilised flowers are absolutely sterile; but the degree of their sterility is largely determined by the conditions to which the parent plants have been exposed, as was well exemplified in the case of Eschscholtzia and Abutilon. The effects of pollen from the same plant are obliterated by the prepotent influence of pollen from another individual or variety, although the latter may have been placed on the stigma some hours afterwards. The offspring from self-fertilised flowers are themselves more or less sterile, sometimes highly sterile, and their pollen is sometimes in an imperfect condition; but I have not met with any case of complete sterility in self-fertilised seedlings, as is so common with hybrids. The degree of their sterility does not correspond with that of the parent-plants when first self-fertilised. The offspring of self-fertilised plants suffer in stature, weight, and constitutional vigour more frequently and in a greater degree than do the hybrid offspring of the greater number of crossed species. Decreased height is transmitted to the next generation, but I did not ascertain whether this applies to decreased fertility.
I have elsewhere shown that by uniting in various ways dimorphic or trimorphic heterostyled plants, which belong to the same undoubted species, we get another series of results exactly parallel with those from crossing distinct species. (12/18. 'Journal of the Linnean Society Botany' volume 10 1867 page 393.) Plants illegitimately fertilised with pollen from a distinct plant belonging to the same form, yield fewer, often much fewer seeds, than they do when legitimately fertilised with pollen from a plant belonging to a distinct form. They sometimes yield no seed, not even an empty capsule, like a species fertilised with pollen from a distinct genus. The degree of sterility is much affected by the conditions to which the plants have been subjected. (12/19. 'Journal of the Linnean Society Botany' volume 8 1864 page 180.) The pollen from a distinct form is strongly prepotent over that from the same form, although the former may have been placed on the stigma many hours afterwards. The offspring from a union between plants of the same form are more or less sterile, like hybrids, and have their pollen in a more or less aborted condition; and some of the seedlings are as barren and as dwarfed as the most barren hybrid. They also resemble hybrids in several other respects, which need not here be specified in detail,--such as their sterility not corresponding in degree with that of the parent plants,--the unequal sterility of the latter, when reciprocally united,--and the varying sterility of the seedlings raised from the same seed-capsule.
We thus have two grand classes of cases giving results which correspond in the most striking manner with those which follow from the crossing of so-called true and distinct species. With respect to the difference between seedlings raised from cross and self-fertilised flowers, there is good evidence that this depends altogether on whether the sexual elements of the parents have been sufficiently differentiated, by exposure to different conditions or by spontaneous variation. It is probable that nearly the same conclusion may be extended to heterostyled plants; but this is not the proper place for discussing the origin of the long-styled, short-styled and mid-styled forms, which all belong to the same species as certainly as do the two sexes of the same species. We have therefore no right to maintain that the sterility of species when first crossed and of their hybrid offspring, is determined by some cause fundamentally different from that which determines the sterility of the individuals both of ordinary and of heterostyled plants when united in various ways.