All who have tried have found much difficulty in crossing papilionaceous flowers.² Several years ago, Dr. Herbert remarked to me that with the exception of Erythrina no hybrids had been made in this great family.³ Gärtner crossed 36 flowers of different varieties of the common Pea, and he did not get a single pod perfectly developed and with the full complement of seed; he crossed 10 flowers of Kidney Beans, and did not get a single pod.⁴ Some years ago I crossed the varieties of the Sweet Pea, and many more flowers dropped off unimpregnated than were fertilised.⁵ The difficulty arises from the anthers opening at so early an age that they must be removed long before the flower expands. After the operation the immature stigma is liable to exposure to the air; and it is difficult to judge when to apply the pollen. Moreover there is some reason to suspect that the stigma requires successive applications of pollen. To show the difficulty of fertilising papilionaceous flowers, I may mention that I lately removed all the pollen that I could with a soft brush from six recently expanded flowers of Lupinus pilosus⁶ protected from the visits of insects, and then applied pollen from a distinct individual of the same species. Although in this case there was no operation at an early age, yet five flowers out of the six dropped off unimpregnated. Had these flowers remained untouched, all, judging from the others, probably would have set, and the only difference would have been that their stigmas would have been surrounded by a mass of pollen as long as the flowers continued in bloom. This case is worth mentioning as showing how erroneous the belief is that fertilisation usually takes place in unopened flowers, in which the pollen is shed at an early age. These trials on the Lupines, and others formerly on Sweet Peas, led me to try the following plan. I rolled up thin paper into a cylinder, rather thinner than a knitting needle. I then tied a thread tight round, and cut off the cylinder beneath the thread, so that a little pipe closed at one end or cap, about the fifth of an inch in length, was left. This was easily filled with pollen from the keel-petal of any desired variety, and could then be placed on the pistil and secured below the stigma by being tied with a thread. I then castrated four flower-buds of the Sweet Pea, and placed on the young stigmas caps filled with pollen from another variety, and four fine pods were soon formed. I also fertilised eight castrated flowers of two species of Lupins with pollen from distinct plants of the same species, but from these I have got only four pods. I may add, that as an experiment I filled one of the little caps with pollen of Lathyrus grandiflorus and placed it on the stigma of a Sweet Pea (Lathyrus odoratus), and to my great surprise, considering how distinct these species are, a fine pod has been formed. I am certain no pollen could have been left in the flower of the Sweet Pea, as the anthers were removed whilst quite immature; and if these hybrid seeds grow, a curious hybrid will be produced. I should not have thought this plan of fertilising papilionaceous flowers worth mentioning had it not been applicable in all cases in which early castration is necessary, and likewise in certain cases mentioned by Gärtner, in which the stigma requires, or is benefited by, successive applications of pollen. In all such cases some trouble would be saved and certainty gained by the use of the little caps filled with the desired kind of pollen.

Charles Darwin

Down, Bromley, Kent