I have looked at the seeds; I never saw in the British orchids nearly so many empty testae; but this goes for nothing, as unnatural conditions would account for it. I suspect, however, from the variable size and transparency, that a good many of the seeds when dry (and I have put the capsule on my chimney-piece) will shrivel up. So I will wait a month or two till I get the capsule of some large Vandeae for comparison. It is more likely that I have made some dreadful blunder about Acropera than that it should be male yet not a perfect male. May there be some sexual relation between A. Loddigesii and luteola; they seem very close? I should very much like to examine the capsule of the unimpregnated flower of A. Loddigesii. I have got both species from Kew, but whether we shall have skill to flower them I know not. One conjectures that it is imperfect male; I still should incline to think it would produce by seed both sexes. But you are right about Primula (and a very acute thought it was): the long-styled P. sinensis, homomorphically fertilised with own-form pollen, has produced during two successive homomorphic generations only long-styled plants. (635/1. In "Forms of Flowers," Edition II., page 216, a summary of the transmission of forms in the "homomorphic" unions of P. sinensis is given. Darwin afterwards used "illegitimate" for homomorphic, and "legitimate" for "heteromorphic" ("Forms of Flowers," Edition i., page 24).) The short-styled the same, i.e. produced short-styled for two generations with the exception of a single plant. I cannot say about cowslips yet. I should like to hear your case of the Primula: is it certainly propagated by seed?
LETTER 636. TO J. SCOTT. Down, December 3rd, [1862?].
What a capital observer you are! and how well you have worked the primulas. All your facts are new to me. It is likely that I overrate the interest of the subject; but it seems to me that you ought to publish a paper on the subject. It would, however, greatly add to the value if you were to cover up any of the forms having pistil and anther of the same height, and prove that they were fully self-fertile. The occurrence of dimorphic and non- dimorphic species in the same genus is quite the same as I find in Linum. (636/1. Darwin finished his paper on Linum in December 1862, and it was published in the "Linn. Soc. Journal" in 1863.) Have any of the forms of Primula, which are non-dimorphic, been propagated for some little time by seed in garden? I suppose not. I ask because I find in P. sinensis a third rather fluctuating form, apparently due to culture, with stigma and anthers of same height. I have been working successive generations homomorphically of this Primula, and think I am getting curious results; I shall probably publish next autumn; and if you do not (but I hope you will) publish yourself previously, I should be glad to quote in abstract some of your facts. But I repeat that I hope you will yourself publish. Hottonia is dimorphic, with pollen of very different sizes in the two forms. I think you are mistaken about Siphocampylus, but I feel rather doubtful in saying this to so good an observer. In Lobelia the closed pistil grows rapidly, and pushes out the pollen and then the stigma expands, and the flower in function is monoecious; from appearance I believe this is the case with your plant. I hope it is so, for this plant can hardly require a cross, being in function monoecious; so that dimorphism in such a case would be a heavy blow to understanding its nature or good in all other cases. I see few periodicals: when have you published on Clivia? I suppose that you did not actually count the seeds in the hybrids in comparison with those of the parent-forms; but this is almost necessary after Gartner's observations. I very much hope you will make a good series of comparative trials on the same plant of Tacsonia. (636/2. See Scott in "Linn. Soc. Journal," VIII.) I have raised 700-800 seedlings from cowslips, artificially fertilised with care; and they presented not a hair's-breadth approach to oxlips.