Bentham. The flowers are so constructed that hive and humble-bees, which visit them incessantly, almost always alight on the left wing-petal, as they can best suck the nectar from this side. Their weight and movements depress the petal, and this causes the stigma to protrude from the spirally-wound keel, and a brush of hairs round the stigma pushes out the pollen before it. The pollen adheres to the head or proboscis of the bee which is at work, and is thus placed either on the stigma of the same flower, or is carried to another flower. (5/4. The flowers have been described by Delpino, and in an admirable manner by Mr. Farrer in the 'Annals and Magazine of Natural History' volume 2 4th series October 1868 page 256. My son Francis has explained 'Nature' January 8, 1874 page 189, the use of one peculiarity in their structure, namely, a little vertical projection on the single free stamen near its base, which seems placed as if to guard the entrance into the two nectar-holes in the staminal sheath. He shows that this projection prevents the bees reaching the nectar, unless they go to the left side of the flower, and it is absolutely necessary for cross-fertilisation that they should alight on the left wing-petal.) Several years ago I covered some plants under a large net, and these produced on one occasion about one-third, and on another occasion about one-eighth, of the number of pods which the same number of uncovered plants growing close alongside produced. (5/5. 'Gardeners' Chronicle' 1857 page 725 and more especially ibid 1858 page 828. Also 'Annals and Magazine of Natural History' 3rd series volume 2 1858 page 462.) This lessened fertility was not caused by any injury from the net, as I moved the wing-petals of several protected flowers, in the same manner as bees do, and these produced remarkably fine pods. When the net was taken off, the flowers were immediately visited by bees, and it was interesting to observe how quickly the plants became covered with young pods. As the flowers are much frequented by Thrips, the self-fertilisation of most of the flowers under the net may have been due to the action of these minute insects. Dr. Ogle likewise covered up a large portion of a plant, and "out of a vast number of blossoms thus protected not a single one produced a pod, while the unprotected blossoms were for the most part fruitful." Mr. Belt gives a more curious case; this plant grows well and flowers in Nicaragua; but as none of the native bees visit the flowers, not a single pod is ever produced. (5/6. Dr. Ogle 'Popular Science Review' 1870 page 168. Mr. Belt 'The Naturalist in Nicaragua' 1874 page 70. The latter author gives a case 'Nature' 1875 page 26, of a late crop of Phaseolus multiflorus near London which "was rendered barren" by the humble-bees cutting, as they frequently do, holes at the bases of the flowers instead of entering them in the proper manner.)
From the facts now given we may feel nearly sure that individuals of the same variety or of different varieties, if growing near each other and in flower at the same time, would intercross; but I cannot myself advance any direct evidence of such an occurrence, as only a single variety is commonly cultivated in England. I have, however, received an account from the Reverend W.A. Leighton, that plants raised by him from ordinary seed produced seeds differing in an extraordinary manner in colour and shape, leading to the belief that their parents must have been crossed. In France M. Fermond more than once planted close together varieties which ordinarily come true and which bear differently coloured flowers and seeds; and the offspring thus raised varied so greatly that there could hardly be a doubt that they had intercrossed. (5/7. 'Fécondation chez les Végétaux' 1859 pages 34-40. He adds that M. Villiers has described a spontaneous hybrid, which he calls Phaseolus coccineus hybridus, in the 'Annales de la Soc. R. de Horticulture' June 1844.) On the other hand, Professor H. Hoffman does not believe in the natural crossing of the varieties; for although seedlings raised from two varieties growing close together produced plants which yielded seeds of a mixed character, he found that this likewise occurred with plants separated by a space of from 40 to 150 paces from any other variety; he therefore attributes the mixed character of the seed to spontaneous variability.