The Natural History Review 14: 243-248 (April, 1864)


  1. ElNIGE BEOBACHTUNGEN UEBER DlMORPHE BLUETHEN.—By H. von Mohl. Botanische Zeitung, 1863, pp. 309, 321.
  2. TREVIRANUS, UEBER DICHOGAMIE. Bot. Zeit. 1863, pp. 1, 9.
  4. ALEFELD, UEBER LINUM. Ib. p. 284.
  7. A. GRAY. STRUCTURE AND FERTILISATION OF CERTAIN ORCHIDS. Am. Journ. Science, Ser. ii. xxxvi, p. 292.
  8. TRIMEN. ON THE FERTILISATION OF Disa grandiflora. Journ. Proc. Linn. Soc. Bot. vii. 144.*

*The papers above cited, suggested by the researches of Mr. Darwin, were all published in 1863.

PROFESSOR VON MOHL’S observations refer principally to those flowers of abnormal structure which occur upon plants bearing, besides abnormal flowers, others which present the structure usual in the genera to which they belong. These 'abnormal' dimorphic flowers are structurally hermaphrodite; and necessarily so, since a prominent character which they present is the total, or almost total, occlusion of their sexual organs, so that if these flowers be fertilised at all, they must necessarily, it would seem, be self-fertilised. In a previous number of this Journal (1862, p. 235), we have described the general features of these dimorphic flowers, and indicated the different groups of plants in which they occur. At the time our communication was penned, some of the best marked and most easily accessible instances had never been accurately described, so that we could indicate only their general features. The chief value of von Mohl's essay is the account which it gives of his own recent examination of a few of these cases, thus supplying a gap which much needed to be filled, and which no one could be more competent to supply than this accomplished and experienced observer. The general characters of the flowers to which we refer are briefly these: they are usually very small; they remain unopened, and often hermetically sealed up, at least until after fecundation has taken place; the corolla is more or less arrested or wholly suppressed; the stamens are often fewer than in the normal flowers, and the quantity of pollen which their anthers contain is very small indeed; their fruit is developed like that of the normal flower, or sometimes even more freely. They may either precede or be contemporary with, or follow the flowers of normal structure borne by the same plant; they are developed either above or below the surface of the ground; they may either be produced annually with the normal flowers, or at a different period, or, according to von Mohl, in some cases to the exclusion of the ordinary flowers.

Von Mohl gives copious references to what has been already written about these curious flowers; instances of which were not unobserved by Dillenius and Linnaeus. As in our previous communication referred to above, we made some reference to their literature, we shall content ourselves here with noting the points in their (structure which von Mohl's observations have now finally set at rest. He proves that in Oxalis, Viola, etc. pollen-tubes are really developed from the pollen-grains while still contained in the anthers, which, in order to get access to the stigma (to which the anthers are closely applied) develope their tubes through the wall of the anther. We previously entertained no doubt but that pollen-tubes were actually developed, binding the anthers to the stigma like so many Lilliputian cables, but we had satisfied ourselves of their existence by direct observation, only in the genera Campanula and Krascheninikovia.

We observe in a recent number of the Bulletin of the Botanical Society of France (Vol. x. p. 194), an interesting account of the flowering and fruiting of Leersia oryzoides by M. Duval-Jouve. It would seem as though this grass might be classed along with the other species cited by von Mohl as bearing dimorphic flowers. From what M. Duval-Jouve tells us, it may, perhaps, prove to be a plant with self-fertilising flowers only. The Leersia is an inconspicuous grass, growing in a few of our southern counties, especially liable to be overlooked, from the circumstance that the flowering-panicle is rarely protruded from the sheath of the upper leaf of the stem. Occasionally it is exserted, and when this is the case the flowers borne upon the exserted portion are sterile, those which are included in the sheath being fertile. So much Schreber pointed out in his u. Beschreibung d. Gräsern, ii. p. 8. In the exserted flowers the glumes spread out so as to allow the stamens and stigma to get free. These organs appear normally developed, nevertheless they do not produce any seed. In the included flowers, fertilisation is found by M. Duval-Jouve to take place at an extremely early period. The fertile flowers are found not only in the uppermost sheaths, but in the sheaths of the lower leaves as well. The glumes of the included fertile flowers are so closely applied that it is almost impossible to separate them without tearing, and the cavity which the envelopes of the flower enclose is constantly filled with a transparent and slightly viscous fluid which bathes the essential organs. Curiously enough the anthers of the fertile flowers are described as being very small, containing but very little pollen. The pollen-grains are extremely delicate and easily torn, and look as though they were abortive.

The very small number of pollen-grains in the anthers of the closed dimorphic flowers, not only of Leersia, but also, as stated above, of Oxalis and Viola, is another interesting point and singularly in contrast with their profusion, and often apparent waste in many flowers of normal structure. In the latter, however, as we learn especially from Mr. Darwin's observations, foreign and adventitious aids are of essential importance, and often absolutely necessary in transferring the pollen from flower to flower, and the supply is no doubt proportioned to the average risk of its miscarriage or careless delivery by the casual or indifferent agents employed.

*Darwin. Origin of Species, 101.

There can be no doubt but that the imperfect dimorphic flowers described by von Mohl are as evidently designed for self-fertilisation as are those of normal form very often designed to insure constant or partial crossing with other flowers. But, von Mohl says, we must not forget " that these dimorphous flowers are not the only ones designed for self-fertilisation; there occur other plants which bear only homomorphic flowers, the structure of which is such that they can only be self-fertilised." In the present state of our knowledge it is quite out of our power to indicate, with any fair confidence, what rôle the small dimorphic flowers occupy in the general economy of plants. We find them occurring in species widely removed from each other in their natural affinity, although in infinitely the largest proportion of flowering plants they do not occur at all, so far as we know. Von Mohl does not venture an explanation of the part which they supply. "We cannot help thinking, however, that he ascribes an undue importance to them, especially when he proceeds to couple with them the self-fertilising homomorphic flowers referred to above, and opposes them collectively to the alleged 'law of nature,' that both in the vegetable and animal kingdoms* there shall be an occasional intercross of distinct individuals.

But what are these cases of self-fertilising homomorphic flowers? "Will they bear investigation? The instance of self-fertilising homomorphic flowers specially adduced by von Mohl is that of the Fumitories. He says, "in all the living species examined by me, self-fertilisation must necessarily take place, and the transport of pollen from one flower to another appears to me to be impossible, owing to the intimate union of the inner petals which enclose the anthers and stigma."

†"Gänzlich unzulässig ist es aber, zu Gunsten der angeblichcn Allgemeinheit eines Naturgesetzes, in welches sich bestimmte Thatsachen nicht fügen wollen, zu verlangen, dass da und dort cinmal, wenn auch nur in Jahrhunderten oder Jahrtausenden, Ausnahmen von dem gewöhnlichen Gange der Functionen der Organe vorkommen, welche bei normaler Ausbildung nicht vorkommen können and fur deren wirkliches Vorkommen keine Beobachtung spricht."

Now without having forlorn recourse to some casual cross once in a century or a chiliad,† we have reason to believe that, in the case of the Fumitories, fertility is favoured and increased by the not unfrequent visits of insects. We venture to state this upon the authority of the admirable observer to whom we already owe so much in this direction, and whose accuracy is unimpeachable. Mr. Darwin has not overlooked the Fumitories in relation to the question of their being self-fertilising or otherwise, and he has been led to the opinion that although they are fertile to a large degree without insect aid, yet they "are all (or nearly all), manifestly adapted to visits of insects." "When a case like that of the Fumitories in which the possibility of a cross has been apparently most vigilantly guarded against breaks down, we feel but little hope of discovering any hermaphrodite flower habitually self-fertilising. We do not, indeed, recall any plant in which self-fertilisation has been shown to obtain with greater general certainty than the Bee Orchis, so well described in Mr. Darwin's work on the "Fertilisation of Orchids."

Professor von Mohl in conclusion, recommends to further examination those plants of warm countries, which, when removed to colder climates, develop their fruit without the flower first expanding in the normal way. Experiments are also needed upon the influence which temperature exercises upon the development of the sexual organs. The observation of Knight is recalled, that a Water Melon in a very warm house bore only male flowers, while Cucumbers growing in a much cooler place bore only female flowers; also the case of Specularia, which develops its apetalous flowers early in the season, while in Viola and Oxalis they generally succeed the normal flowers, when the summer is further advanced and the weather warmer.

Treviranus refers at some length to Mr. Darwin's essays upon the various contrivances of nature to insure heteromorphic fertilisation, naming several additional species presenting more or less of a dimorphic condition with respect to the essential organs. He says that in Primula longiflora the stigma always projects beyond the anthers. Prof. Treviranus distinguishes two categories both of homomorphic and of heteromorphic flowers. Of homomorphic: 1st, those in which self-fertilisation is insured by juxtaposition and synchronous maturity of the sexual organs, no mechanical difficulty interfering to prevent the access of pollen to own-flower stigma. In this class he includes "most flowers, Crucifers, Rosaceae, and especially the irregular-flowered Papilionaceae, Labiatae, &c." 2nd, those which require some kind of aid to insure fertilisation, either the motion or change of position of organs during development, or the visits of insects. Of this category he names Proteaceae, Asclepiadeae, Compositae, Campanulaceae, Lobeliaceae, &c. Of heteromorphic flowers: 1st. Cases in which structural hermaphrodites are physiologically more or less disqualified for self-fertilisation, as in the instances given by Mr. Darwin of Primula, the Orchids, &c.; and, 2nd, unisexual (diclinous) flowers. This classification is not satisfactory, and appears to us to rest upon an imperfect basis of observation. It is probable that future research will prove that nearly all hermaphrodite flowers are more or less heteromorphic, from the stigma favouring, though in the most different degrees, the pollen of other flowers of the same species.

Extreme instances of heteromorphism, in which the pollen of a distinct species would appear to have the advantage over own-flower pollen, we find recorded by Mr. Scott, in his paper cited on page 243. He has been trying to fertilise species of Oncidium and Maxillaria with their own pollen artificially applied, but without success. He found, however, that they might be fertilised with the pollen of distinct species. If these cases do not break down on extended scrutiny, Professor Treviranus will find a fifth category on his hands.

Dr. Hildebrand, writing upon the Orchids, calls attention to the twofold influence exercised by the pollen, first in determining an enlargement of the ovary and development of the ovules upon the placentas; and, second, in the actual fertilisation of the ovules. He finds that in Dendrobium nobile some four months intervene between the period of application of the pollen to the stigma, and 'the incipient formation of the embryo;' in Cymbidium sinense, perhaps even six months intervene. At the time of flowering, the placentary lines, though undulated, do not bear even the rudiments of ovules. He gives the corresponding intervals in several other species. It is generally stouter than in the Dendrobium, and in the different species of Orchis, varies from eight or nine days to a month. This is a curious and difficult subject, though not a new one, and it would be well worth while to collect more information upon it, both with regard to the supposed double effect of the pollen, and the length of time intervening between its application and the actual fertilisation of the ovules. In some Monochlamydeae and in Gymnosperms, the interval between the application of the pollen and fertilisation is wonderfully long. With regard to the influence of pollen in determining the enlargement of the ovary, we recall the curious case noted by M. Duchartre (Bull. Soc. Bot. ix. p. 531) of the ovules of Cycas revoluta, to which the pollen of a Ceratozamia had been applied, enlarging to well-developed fruits without containing any trace of an embryo. He does not attempt to settle how far this enlargement of the fruit is due to the foreign pollen. No doubt many analogous cases must be familiar to our experienced horticulturists, besides those recorded by Herbert and by the German Hybridologists, and something might issue were they judiciously correlated.