Your observations seem to have been chiefly directed to the A. luteola, mine to the A. Loddigesii, which, however, as you remark, is in a very similar constructural condition with the former; having the same narrow stigmatic chamber, abnormally developed placenta, etc. In regard to the former point—contraction of stigmatic chamber—I may remark that it does not appear to be absolutely necessary that the pollen-masses penetrate this chamber for effecting fecundation. Thus a raceme was produced upon a plant of A. Loddigesii in the Botanic Gardens here lately; upon this I left only six flowers. These I attempted to fertilise, but with two only of the six have I been successful: I succeeded in forcing a single pollen-mass into the stigmatic chamber of one of the latter, but I failed to do this on the other; however, by inserting a portion of a pedicel with a pollinium attached, I caused the latter to adhere, with a gentle press, to the mouth of the stigmatic chamber. Both of these, as I have already remarked, are nevertheless fertilised; one of them I have cut off for examination, and its condition I will presently describe; the other is still upon the plant, and promises fair to attain maturity. In regard to the other four flowers, I may remark that though similarly fertilised—part having pollinia inserted, others merely attached—they all withered and dropped off without the least swelling of the ovary. Can it be, then, that this is really an {andro-monoecious} species?—part of the flowers male, others truly hermaphrodite.
In making longitudinal sections of the fertilised ovary before mentioned, I found the basal portion entirely destitute of ovules, their place being substituted by transparent cellular ramification of the placentae. As I traced the placentae upwards, the ovules appeared, becoming gradually more abundant towards its apex. A transverse section near the apex of the ovary, however, still exhibited a more than ordinary placental development—i.e. {congenitally?} considered—each end giving off two branches, which meet each other in the centre of the ovary, the ovules being irregularly and sparingly disposed upon their surfaces.
In regard to the mere question of fertilisation, then, I am perfectly satisfied, but there are other points which require further elucidation. Among these I may particularly refer to the contracted stigmatic chamber, and the slight viscidity of its disk. The latter, however, may be a consequence of uncongenial conditions—as you do not mention particularly its examination by any author in its natural habitat. If such be the case, the contracted stigmatic chamber will offer no real difficulty, should the viscous exudations be only sufficient to render the mouth adhesive. For, as I have already shown, the pollen-tubes may be emitted in this condition, and effect fecundation without being in actual contact with the stigmatic surface, as occurs pretty regularly in the fertilisation of the Stapelias, for example. But, indeed, your own discovery of the independent germinative capabilities of the pollen-grains of certain Orchidaceae is sufficiently illustrative of this.
I may also refer to the peculiar abnormal condition that many at least of the ovaries present in a comparative examination of the placentae, and of which I beg to suggest the following explanation, though it is as yet founded on limited observations. In examining certain young ovaries of A. Loddigesii, I found some of them filled with the transparent membranous fringes of more or less distinctly cellular matter, which, from your description of the ovaries of luteola, appears to differ simply in the greater development in the former species. Again, in others I found small mammillary bodies, which appeared to be true ovules, though I could not perfectly satisfy myself as to the existence of the micropyle or nucleus. I unfortunately neglected to apply any chemical test. The fact, however, that in certain of the examined ovaries few or none of the latter bodies occurred—the placenta alone being developed in an irregular membranous form, taken in conjunction with the results of my experiments—before alluded to—on their fertilisation, leads me to infer that two sexual conditions are presented by the flowers of this plant. In short, that many of the ovaries are now normally abortive, though Nature occasionally makes futile efforts for their perfect development, in the production of ovuloid bodies; these then I regard as the male flowers. The others that are still capable of fertilisation, and likewise possessing male organs, are hermaphrodite, and must, I think, from the results of your comparative examinations, present a somewhat different condition; as it can scarcely be supposed that ovules in the condition you describe could ever be fertilised.
This is at least the most plausible explanation I can offer for the different results in my experiments on the fertilisation of apparently similar morphologically constructed flowers; others may, however, occur to you. Here there is not, as in the Catasetum, any external change visible in the respective unisexual and bisexual flowers. And yet it would appear from your researches that the ovules of Acropera are in a more highly atrophied condition than occurs in Catasetum, though, as you likewise remark, M. Neumann has never succeeded in fertilising C. tridentatum. If there be not, then, an arrangement of the reproductive structures, such as I have indicated, how can the different results in M. Neumann's experiments and mine be accounted for? However, as you have examined many flowers of both A. luteola and Loddigesii, such a difference in the ovulary or placental structures could scarcely have escaped your observation. But, be this as it may, the—to me at least—demonstrated fact still remains, that certain flowers of A. Loddigesii are capable of fertilisation, and that, though there are good grounds for supposing that important physiological changes are going on in the sexual phenomena of this species, there is no evidence whatever for supposing that external morphological changes have so masked certain individuals as to prevent their recognition.
I would now, sir, in conclusion beg you to excuse me for this infringement upon your valuable time, as I have been induced to write you in the belief that you have had negative results from other experimenters, before you ventured to propose your theoretical explanation, and consequently that you have been unknowingly led into error. I will continue, as opportunities present themselves, to examine the many peculiarities you have pointed out in this as well as others of the Orchid family; and at present I am looking forward with anxiety for the maturation of the ovary of A. Loddigesii, which will bear testimony to the veracity of the remarks I have ventured to lay before you.
LETTER 632. TO J.D. HOOKER. Down, 18th {November 1862}.
Strange to say, I have only one little bother for you to-day, and that is to let me know about what month flowers appear in Acropera Loddigesii and luteola; for I want extremely to beg a few more flowers, and if I knew the time I would keep a memorandum to remind you. Why I want these flowers is (and I am much alarmed) that Mr. J. Scott, of Bot. Garden of Edinburgh (do you know anything of him?) has written me a very long and clever letter, in which he confirms most of my observations; but tells me that with much difficulty he managed to get pollen into orifice, or as far as mouth of orifice, of six flowers of A. Loddigesii (the ovarium of which I did not examine), and two pods set; one he gathered, and saw a very few ovules, as he thinks, on the large and mostly rudimentary placenta. I shall be most curious to hear whether the other pod produces a good lot of seed. He says he regrets that he did not test the ovules with chemical agents: does he mean tincture of iodine? He suggests that in a state of nature the viscid matter may come to the very surface of stigmatic chamber, and so pollen-masses need not be inserted. This is possible, but I should think improbable. Altogether the case is very odd, and I am very uneasy, for I cannot hope that A. Loddigesii is hermaphrodite and A. luteola the male of the same species. Whenever I can get Acropera would be a very good time for me to look at Vanda in spirits, which you so kindly preserved for me.
LETTER 633. TO J. SCOTT.
(633/1. The following is Darwin's reply to the above letter from Scott. In the first edition of "Fertilisation of Orchids" (page 209) he assumed that the sexes in Acropera, as in Catasetum, were separate. In the second edition (page 172) he writes: "I was, however, soon convinced of my error by Mr. Scott, who succeeded in artificially fertilising the flowers with their own pollen. A remarkable discovery by Hildebrand (633/2. "Bot. Zeitung," 1863 and 1865.), namely, that in many orchids the ovules are not developed unless the stigma is penetrated by the pollen-tubes...explains the state of the ovarium in Acropera, as observed by me." In regard to this subject see Letter 608.)