The Various Contrivances by which Orchids are Fertilised by Insects/Chapter 9

by Charles Darwin

 

CHAPTER IX.

GRADATION OF ORGANS, &C.—CONCLUDING REMARKS.

Gradation of organs, of the rostellum, of the pollen-masses—Formation of the caudicle—Genealogical affinities—Secretion of nectar—Mechanism of the movement of the pollinia—Uses of the petals—Production of seed—Importance of trifling details of structure—Cause of the great diversity of structure in the flowers of Orchids—Cause of the perfection of the contrivances—Summary on insect-agency—Nature abhors perpetual self-fertilisation.

This chapter will be devoted to the consideration of several miscellaneous subjects which could not well have been introduced elsewhere.

On the gradation of certain Organs.—The rostellum, the pollinia, the labellum, and, in a lesser degree, the column, are the most remarkable points in the structure of Orchids. The formation of the column and labellum, by the confluence and partial abortion of several organs, has been discussed in the last chapter. With respect to the rostellum, no such organ exists in any other group of plants. If the homologies of Orchids had not been pretty well made out, those who believe in the separate creation of each organism might have advanced this as an excellent instance of a perfectly new organ having been specially created, and which could not have been developed by successive slow modifications of any pre-existing part. But, as Robert Brown long ago remarked, it is not a new organ. It is impossible to look at the two groups of spiral vessels (fig. 36) running from the bases of the midribs of the two lower sepals to the two lower stigmas, which are sometimes quite distinct, and then to look at the third group of vessels running from the base of the mid-rib of the upper sepal to the rostellum, which occupies the exact position of a third stigma, and doubt its homological nature. There is every reason to believe that the whole of this upper stigma, and not merely a part, has been converted into the rostellum; for there are plenty of cases of two stigmas, but not one of three stigmatic surfaces being present in those Orchids which have a rostellum. On the other hand, in Cypripedium and Apostasia (the latter ranked by Brown in the Orchidean order), which are destitute of a rostellum, the stigmatic surface is trifid.

As we know only those plants which are now living, it is impossible to follow all the gradations by which the upper stigma has been converted into the rostellum; but let us see what are the indications of such a change having been effected. With respect to function the change has not been so great as it at first appears. The function of the rostellum is to secrete viscid matter, and it has lost the capacity of being penetrated by the pollen-tubes. The stigmas of Orchids, as well as of most other plants, secrete viscid matter, the use of which is to retain the pollen when brought to them by any means, and to excite the growth of the pollen-tubes. Now if we look to one of the simplest rostellums,—for instance, to that of Cattleya or Epidendrum,—we find a thick layer of viscid matter, not distinctly separated from the viscid surface of the two confluent stigmas: its use is simply to affix the pollen-masses to a retreating insect, which are thus dragged out of the anther and transported to another flower, where they are retained by the almost equally viscid stigmatic surface. So that the office of the rostellum is still to secure the pollen-masses, but indirectly by means of their attachment to an insect's body.

The viscid matter of the rostellum and of the stigma appear to have nearly the same nature; that of the rostellum generally has the peculiar property of quickly drying or setting hard; that of the stigma, when removed from the plant, apparently dries more quickly than gum-water of about equal density or tenacity. This tendency to dry is the more remarkable, as Gärtner^[1] found that drops of the stigmatic secretion from Nicotiana did not dry in two months. The viscid matter of the rostellum in many Orchids when exposed to the air changes colour with remarkable quickness, and becomes brownish-purple; and I have noticed a similar but slower change of colour in the viscid secretion of the stigmas of some Orchids, as of Cephalanthera grandiflora. When the viscid disc of an Orchis, as Bauer and Brown have observed, is placed in water, minute particles are expelled with violence in a peculiar manner; and I have observed exactly the same fact in the layer of viscid matter covering the stigmatic utriculi in an unopened flower of Mormodes ignea.

In order to compare the minute structure of the rostellum and stigma, I examined young flower-buds of Epidendrum cochleatum and floribundum, which, when mature, have a simple rostellum. The posterior parts of both organs were quite similar. The whole of the rostellum at this early age consisted of a mass of nearly orbicular cells, containing spheres of brown matter, which resolve themselves into the viscid fluid. The stigma was covered with a thinner layer of similar cells, and beneath them were the coherent spindle-formed utriculi. These are believed to be connected with the penetration of the pollen-tubes; and their absence in the rostellum probably accounts for its not being penetrated. If the structure of the rostellum and of the stigma is as here described, their only difference consists in the layer of cells which secrete the viscid matter being thicker in the rostellum than in the stigma, and in the utriculi having disappeared from the former. There is therefore no great difficulty in believing that the upper stigma, whilst still in some degree fertile or capable of penetration by the pollen-tubes, might have gradually acquired the power of secreting a larger amount of viscid matter, losing at the same time its capacity for fertilisation; and that insects smeared with this viscid matter removed and transported the pollen-masses in a more and more effective manner to the stigmas of other flowers. In this case an incipient rostellum would have been formed.

In the several tribes, the rostellum presents a marvellous amount of diversity of structure; but most of the differences can be connected without very wide breaks. One of the most striking differences is, that either the whole anterior surface to some depth, or only the internal parts become viscid; and in this latter case the surface retains, as in Orchis, a membranous condition. But these two states graduate into each other so closely, that it is scarcely possible to draw any line of separation between them: thus, in Epipactis, the exterior surface undergoes a vast change from its early cellular condition, for it becomes converted into a highly elastic and tender membrane, which is in itself slightly viscid, and allows the underlying viscid matter readily to exude; yet it acts as a membrane, and its under surface is lined with much more viscid matter. In Habenaria chlorantha the exterior surface is highly viscid, but still closely resembles, under the microscope, the exterior membrane of Epipactis. Lastly, in some species of Oncidium, &c., the exterior surface, which is viscid, differs, as far as appearance under the microscope goes, from the underlying viscid layer only in colour; but it must have some essential difference, for I find that, until this very thin exterior layer is disturbed, the underlying matter remains viscid; but, after it has been disturbed, the underlying matter rapidly sets hard. The gradation in the state of the surface of the rostellum is not surprising, for in all cases the surface is cellular in the bud; so that an early condition has only to be retained more or less perfectly.

The nature of the viscid matter differs remarkably in different Orchids: in Listera it sets hard almost instantly, more quickly than plaster of Paris; in Malaxis and Angræcum it remains fluid for several days; but these two states pass into each other by many gradations. In an Oncidium I have observed the viscid matter to dry in a minute and a half; in some species of Orchis in two or three minutes; in Epipactis in ten minutes; in Gymnadenia in two hours; and in Habenaria in over twenty-four hours. After the viscid matter of Listera has set hard, neither water nor weak spirits of wine has any effect on it; whereas that of Habenaria bifolia, after having been dried for several months, when moistened became as adhesive as ever it was. The viscid matter in some species of Orchis, when remoistened, presented an intermediate condition.

One of the most important differences in the state of the rostellum is, whether or not the pollinia are permanently attached to it. I do not allude to those cases in which the upper surface of the rostellum is viscid, as in Malaxis and some Epidendrums, and simply adheres to the pollen-masses; for these cases present no difficulty. But I refer to the so-called congenital attachment of the pollinia by their caudicles to the rostellum or viscid disc. It is not, however; strictly correct to speak of congenital attachment, for the pollinia are invariably free at an early period, and become attached either earlier or later in different Orchids. No actual gradation is at present known in the process of attachment; but it can be shown to depend on very simple conditions and changes. In the Epidendreæ the pollinia consist of a ball of waxy pollen, with a long caudicle (formed of elastic threads with adherent pollen-grains), which never becomes spontaneously attached to the rostellum. In some of the Vandeæ, as in Cymbidium giganteum, on the other hand, the caudicles are congenitally (in the above sense) attached to the pollen-masses, but their structure is the same as in the Epidendreæ, with the sole difference, that the extremities of the elastic threads adhere to, instead of merely lying on, the upper lip of the rostellum.

In a form allied to Cymbidium, namely, Oncidium unguiculatum, I studied the development of the caudicles. At an early period the pollen-masses are enclosed in membranous cases, which soon rupture at one point. At this early period, a layer of rather large cells, including remarkably opaque matter, may be detected within the cleft of each pollen-mass. This matter can be traced as it gradually changes into a translucent substance which forms the threads of the caudicles. As the change progresses, the cells themselves disappear. Finally the threads at one end adhere to the waxy pollen-masses, and at the other end after protruding through a small opening in the membranous case in a semi-developed state, they adhere to the rostellum, against which the anther is pressed. So that the adhesion of the caudicle to the back of the rostellum seems to depend solely on the early rupturing of the anther-case, and on a slight protrusion of the caudicles, before they have become fully developed and hardened.

In all the Orchideæ a portion of the rostellum is removed by insects when the pollinia are removed; for the viscid matter, though conveniently spoken of as a secretion, is in fact part of the rostellum in a modified condition. But in those species which have their caudicles attached at an early period to the rostellum, a membranous or solid portion of its exterior surface in an unmodified condition is likewise removed. In the Vandeæ this portion is sometimes of considerable size (forming the disc and pedicel of the pollinium), and gives to their pollinia their remarkable character; but the differences in the shape and size of the removed portions of the rostellum can be finely graduated together, even within the single tribe of the Vandeæ; and still more closely by commencing with the minute oval atom of membrane to which the caudicle of Orchis adheres, passing thence to that of Habenaria bifolia, to that of H. chlorantha with its drum-like pedicel, and thence through many forms to the great disc and pedicel of Catasetum.

In all the cases in which a portion of the exterior surface of the rostellum is removed together with the caudicles of the pollen-masses, definite and often complicated lines of separation are formed, so as to allow of the easy separation of the removed portions. But the formation of these lines of separation does not differ much from the process by which certain portions of the exterior surface of the rostellum assume a condition intermediate between that of unaltered membrane and of viscid matter, which has been already alluded to. The actual separation of portions of the rostellum depends in many cases on the excitement from a touch; but how a touch thus acts is at present inexplicable. Such sensitiveness, in the stigma to a touch (and the rostellum, as we know, is a modified stigma), and indeed in almost every other part, is by no means a rare quality in plants.

In Listera and Neottea, if the rostellum is touched, even by a human hair, two points rupture and the loculi containing the viscid matter instantly expel it. Here we have a case towards which as yet no gradation is known. But Dr. Hooker has shown that the rostellum is at first cellular, and that the viscid matter is developed within the cells, as in other Orchids.

The last difference which I will mention in the state of the rostellum of various Orchids is the existence in many Ophreæ of two widely-separated viscid discs, sometimes included in two separate pouches. Here it appears at first sight as if there were two rostella; but there is never more than one medial group of spiral vessels. In the Vandeæ we can see how a single viscid disc and a single pedicel might become divided into two; for in some Stanhopeas the heart-shaped disc shows a trace of a tendency to division; and in Angræcum we have two distinct discs and two pedicels, either standing close together or removed only a little way apart.

It might be thought that a similar gradation from a single rostellum into what appears like two distinct rostella was shown still more plainly in the Ophreæ; for we have the following series,—in Orchis pyramidalis a single disc enclosed in a single pouch—in Aceras two discs touching and affecting each other's shapes, but not actually joined—in Orchis latifolia and maculata two quite distinct discs but with the pouch still showing plain traces of division; and, lastly, in Ophrys we have two perfectly distinct pouches, including of course two perfectly distinct discs. But this series does not indicate the former steps by which a single rostellum became divided into two distinct organs; on the contrary, it shows how the rostellum, after having been anciently divided into two organs, has now in several cases been reunited into a single organ.

This conclusion is founded on the nature of the little medial crest, sometimes called the rostellate process, between the bases of the two anther-cells (see fig. 1, B and D, p. 8). In both divisions of the Ophreæ—namely the species having naked discs and those having discs enclosed in a pouch—whenever the two discs come into close juxta-position, this medial crest or process appears.^[2] On the other hand, when the two discs stand widely apart, the summit of the rostellum between them is smooth, or nearly smooth. In the Frog Orchis (Peristylus viridis) the overarching summit is bent like the roof of a house; and here we see the first stage in the formation of the folded crest. In Herminium monorchis, however, which has two separate and large discs, a crest, or solid ridge, is rather more plainly developed than might have been expected. In Gymnadenia conopsea, Orchis maculata, and others, the crest consists of a hood of thin membrane; in O. mascula the two sides of the hood partly adhere; and in O. pyramidalis and in Aceras it is converted into a solid ridge. These facts are intelligible only on the view, that, whilst the two discs were gradually brought together, during a long series of generations, the intermediate portion or summit of the rostellum became more and more arched, until a folded-crest, and finally a solid ridge was formed.

Fig. 37.

Rostellum of Catasetum.

ped.an. antennæ of rostellum.

ped.d. viscid disc.

ped. pedicel of rostellum, to which the pollen-masses are attached.

Whether we compare together the state of the rostellum in the various tribes of the Orchideæ, or compare the rostellum with the pistil and stigma of an ordinary flower, the differences are wonderfully great. A simple pistil consists of a cylinder surmounted by a small viscid surface. Now, see what a contrast the rostellum of Catasetum, when dissected from all the other elements of the column, presents; and as I traced all the vessels in this Orchid, the drawing may be trusted as approximately accurate. The whole organ has lost its normal function of being fertilised. Its shape is most singular, with the upper end thickened, bent over and produced into two long tapering and sensitive antennæ, each of these being hollow within, like an adder's fang. Behind and between the bases of these antennæ we see the large viscid disc, attached to the pedicel; the latter differs in structure from the underlying portion of the rostellum, and is separated from it by a layer of hyaline tissue, which spontaneously dissolves when the flower is mature. The disc, attached to the surrounding parts by a membrane which ruptures as soon as it is excited by a touch, consists of strong upper tissue, with an underlying elastic cushion, coated with viscid matter; and this again in most Orchids is overlaid by a film of a different nature. What an amount of specialisation of parts do we here behold! Yet in the comparatively few Orchids described in this volume, so many and such plainly-marked gradations in the structure of the rostellum have been described, and such plain facilities for the conversion of the upper pistil into this organ, that, we may well believe, if we could see every Orchid which has ever existed throughout the world, we should find all the gaps in the existing chain, and every gap in in many lost chains, filled up by a series of easy transitions.

We now come to the second great peculiarity in the Orchideæ, namely their pollinia. The anther opens early, and often deposits the naked masses of pollen on the back of the rostellum. This action is prefigured in Canna, a member of a family nearly related to the Orchideæ, in which the pollen is deposited on the pistil, close beneath the stigma. In the state of the pollen there is great diversity: in Cypripedium and Vanilla single grains are embedded in a glutinous fluid; in all the other Orchids seen by me (except the degraded Cephalanthera) the grains are united three or four together.^[3] These compound grains are tied one to the other by elastic threads, but they often form packets which are tied together in like manner, or they are cemented into the so-called waxy masses. The waxy masses graduate in the Epidendreæ and Vandeæ from eight to four, to two, and, by the cohesion of the two, into a single mass. In some of the Epidendreæ we have both kinds of pollen within the same anther, namely, large waxy masses, and caudicles formed of elastic threads with numerous compound grains adhering to them.

I can throw no light on the nature of the cohesion of the pollen in the waxy masses; when they are placed in water for three or four days, the compound grains readily fall apart; but the four grains of which each is formed still firmly cohere; so that the nature of the cohesion in the two cases must be different. The elastic threads by which the packets of pollen are tied together in the Ophreæ, and which run far up inside the waxy masses of the Vandeæ, are also of a different nature from the cementing matter; for the threads are acted on by chloroform and by long immersion in spirits of wine; whilst these fluids have no particular action on the cohesion of the waxy masses. In several Epidendreæ and Vandeæ the exterior grains of the pollen-masses differ from the interior grains, in being larger, and in having yellower and much thicker walls. So that in the contents of a single anther-cell we see a surprising degree of differentiation in the pollen, namely, grains cohering by fours, then being either tied together by threads or cemented together into solid masses, with the exterior grains different from the interior ones.

In the Vandeæ, the caudicle, which is composed of fine coherent threads, is developed from the semi-fluid contents of a layer of cells. As I find that chloroform has a peculiar and energetic action on the caudicles of all Orchids, and likewise on the glutinous matter which envelopes the pollen-grains in Cypripedium, and which can be drawn out into threads, we may suspect that in this latter genus,—the least differentiated in structure of all the Orchideæ,—we see the primordial condition of the elastic threads by which the pollen-grains are tied together in other and more highly developed species.^[4]

The caudicle, when largely developed and destitute of pollen-grains, is the most striking of the many peculiarities presented by the pollinia. In some Neotteæ, especially in Goodyera, we see it in a nascent condition, projecting just beyond the pollen-mass, with the threads only partially coherent. In the Vandeæ by tracing the gradation from the ordinary naked condition of the caudicle, through Lycaste in which it- is almost naked, through Calanthe, to Cymbidium giganteum, in which it is covered with pollen-grains, it seems probable that its ordinary condition has been arrived at by the modification of a pollinium like that of one of the Epidendreæ; namely, by the abortion of the pollen-grains which primordially adhered to separate elastic threads, and afterwards by the cohesion of these threads.

In the Ophreae we have better evidence than is afforded by gradation, that their long, rigid and naked caudicles have been developed, at least partially, by the abortion of the greater number of the lower pollen-grains and by the cohesion of the elastic threads by which these grains were tied together. I had often observed a cloudy appearance in the middle of the translucent caudicles in certain species; and on carefully opening several caudicles of Orchis pyramidalis, I found in their centres, fully half-way down between the packets of pollen and the viscid disc, many pollen-grains (consisting, as usual, of four united grains), lying quite loose. These grains, from their embedded position, could never by any possibility have been left on the stigma of a flower, and were absolutely useless. Those who can persuade themselves that purposeless organs have been specially created, will think little of this fact. Those on the contrary, who believe in the slow modification of organic beings, will feel no surprise that the changes have not always been perfectly effected,—that, during and after the many inherited stages of the abortion of the lower pollen-grains and of the cohesion of the elastic threads, there should still exist a tendency to the production of a few grains where they were originally developed; and that these should consequently be left entangled within the now united threads of the caudicle. They will look at the little clouds formed by the loose pollen-grains within the caudicles of Orchis pyramidalis, as good evidence that an early progenitor of this plant had pollen-masses like those of Epipactis or Goodyera, and that the grains slowly disappeared from the lower parts, leaving the elastic threads naked and ready to cohere into a true caudicle.

As the caudicle plays an important part in the fertilisation of the flower, it might have been developed from one in a nascent condition, such as we see in Epipactis, to any required length merely by the continued preservation of varying increments in its length, each beneficial in relation to other changes in the structure of the flower, and without any abortion of the lower pollen-grains. But we may conclude from the facts just given, that this has not been the sole means,—that the caudicle owes much of its length to such abortion. That in some cases it has subsequently been largely increased in length by natural selection, is highly probable; for in Bonatea speciosa the caudicle is actually more than thrice as long as the elongated pollen-masses; and it is highly improbable that so lengthy a mass of grains, slightly cohering together by the aid of elastic threads, should ever have existed, as an insect could not have safely transported and applied a mass of this shape and size to the stigma of another flower.

We have hitherto considered gradations in the state of the same organ. To any one with more knowledge than I possess, it would be an interesting subject to trace the gradations between the several species and groups of species in this great and closely-connected order. But to make a perfect gradation, all the extinct forms which have ever existed, along many lines of descent converging to the common progenitor of the group, would have to be called back into life. It is due to their absence, and to the consequent wide gaps in the series, that we are enabled to divide the existing species into definable groups, such as genera, families, and tribes. If there had been no extinction, there would still have been great lines or branches of special development,—the Vandeæ, for instance, would still have been distinguishable as a great body, from the great body of the Ophreæ; but ancient and intermediate forms, very different probably from their present descendants, would have rendered it utterly impossible to separate by distinct characters the one great body from the other.

I will venture on only a few more remarks. Cypripedium, in having three stigmas developed, and therefore in not possessing a rostellum, in having two fertile anthers with a large rudiment of a third, and in the state of its pollen, seems a remnant of the order whilst in a simpler or more generalised condition. Apostasia is a related genus, placed by Brown amongst the Orchideæ, but by Lindley in a small distinct family. These broken groups do not indicate to us the structure of the common parent-form of all the Orchideæ, but they serve to show the probable state of the order in ancient times, when none of the forms had become so widely differentiated from one another and from other plants, as are the existing Orchids, especially the Vandeæ and Ophreæ; and when, consequently, the order made a nearer approach in all its characters, than it does at present, to such allied groups as the Marantaceæ.

With respect to other Orchids, we can see that an ancient form, like one of the sub-tribe of the Pleurothallidæ, some of which have waxy pollen-masses with a minute caudicle, might have given rise, by the entire abortion of the caudicle, to the Dendrobiæ, and by an increase of the caudicle to the Epidendreæ. Cymbidium shows us how simply a form like one of our present Epidendreæ could be modified into one of the Vandeæ. The Neotteæ stand in nearly a similar relation to the higher Ophreæ, which the Epidendreæ do to the higher Vandeæ. In certain genera of the Neotteæ we have compound pollen-grains cemented into packets and tied together by elastic threads, which project and thus form a nascent caudicle. But this caudicle does not protude from the lower end of the pollinium as in the Ophreæ, nor does it always protrude from the extreme upper end in the Neotteæ, but sometimes at an intermediate level; so that a transition in this respect is far from impossible. In Spiranthes, the back of the rostellum, lined with viscid matter, is alone removed: the front part is membranous, and ruptures like the pouch-formed rostellum of the Ophreæ. An ancient form combining most of the characters, but in a less developed state, of Goodyera, Epipactis, and Spiranthes, all members of the Neotteæ, could by further slight modifications have given birth to the tribe of the Ophreæ.

Hardly any question in Natural History is more vague and difficult to answer than what forms ought to be considered as the highest in a large group;^[5] for all are well adapted to their conditions of life. If we look to successive modifications, with differentiation of parts and consequent complexity of structure, as the standard of comparison, the Ophreæ and Vandeæ will stand the highest among the Orchideæ. Are we to lay much stress on the size and beauty of the flower, and on the size of the whole plant? if so, the Vandeæ are pre-eminent. They, have, also, rather more complex pollinia, with the pollen-masses often reduced to two. The rostellum, on the other hand, has apparently been more modified from its primordial stigmatic nature in the Ophreæ, than in the Vandeæ. In the Ophreæ the stamens of the inner whorl are almost entirely suppressed,—the auricles—mere rudiments of rudiments—being alone retained; and even these are sometimes lost. These stamens, therefore, have suffered extreme reduction; but can this be considered as a sign of highness? I should doubt whether any member of the Orchidean order has been more profoundly modified in its whole structure than Bonatea spedosa, one of the Ophreæ. So again, within this same tribe, nothing can be more perfect than the contrivances in Orchis pyramidalis for its fertilisation. Yet an ill-defined feeling tells me to rank the magnificent Vandese as the highest. When we look within this tribe at the elaborate mechanism for the ejection and transportal of the pollinia of Catasetum, with the sensitive rostellum so wonderfully modified, with the sexes borne on distinct plants, we may perhaps give the palm of victory to this genus.

SECRETION OF NECTAR.

Many Orchids, both our native species and the exotic kinds cultivated in our hothouses, secrete a copious supply of nectar. I have found the horn-like nectaries of Aerides filled with fluid; and Mr. Rodgers, of Sevenoaks, informs me that he has taken crystals of sugar of considerable size from the nectary of A. cornutum. The nectar-secreting organs of the Orchideæ present great diversities of structure and position in the various genera; but are almost always situated towards the base of the labellum. In Disa, however, the posterior sepal alone, and in Disperis the two lateral sepals together with the labellum, secrete nectar. In Dendrobium chrysanthum the nectary consists of a shallow saucer; in Evelyna, of two large united cellular balls; and in Bolbophyllum cupreum, of a medial furrow. In Cattleya the nectary penetrates the ovarium. In Angræcum sesquipedale it attains the astonishing length of above eleven inches; but I need not enter on further details. The fact, however, should be recalled, that in Coryanthes the nectar-secreting glands pour forth an abundance of almost pure water, which drips into a bucket formed by the distal part of the labellum; and this secretion serves to prevent the bees which come to gnaw the surface of the labellum from flying away, and thus compels them to crawl out through the proper passage.

Although the secretion of nectar is of the highest importance to Orchids by attracting insects, which are indispensable for the fertilisation of most of the species, yet good reasons can be assigned for the belief^[6] that nectar was aboriginally an excretion for the sake of getting rid of superfluous matter during the chemical changes which go on in the tissues of plants, especially whilst the sun shines. The bracteæ of some Orchids have been observed^[7] to secrete nectar, and this cannot be of any use to them for their fertilisation. Fritz Müller informs me that he has seen such secretion from the bracteæ of an Oncidium in its native Brazilian home, as well as from the bracteæ and from the outside of the upper sepal of a Notylia. Mr. Rodgers has observed a similar and copious secretion from the base of the flower-peduncles of Vanilla. The column of Acropera and Gongora likewise secretes nectar, as previously stated, but only after the flowers have been impregnated, and when such secretion could be of no use by attracting insects. It is in perfect accordance with the scheme of nature, as worked out by natural selection, that matter excreted to free the system from superfluous or injurious substances should be utilised for highly useful purposes. To give an example in strong contrast with our present subject, the larvæ of certain beetles (Cassidæ, &c.) use their own excrement to make an umbrella-like protection for their tender bodies.

It may be remembered that evidence was given in the first chapter proving that nectar is never found within the spur-like nectaries of several species of Orchis, but that various kinds of insects penetrate the tender inner coat with their proboscides, and suck the fluid contained in the inter-cellular spaces. This conclusion has been confirmed by Hermann Müller, and I have further shown that even Lepidoptera are able to penetrate other and tougher tissues. It is an interesting case of co-adaptation that in all the British species, in which the nectary does not contain free nectar, the viscid matter of the disc of the pollinium requires a minute or two in order to set hard; and it would be an advantage to the plant if insects were delayed thus long in obtaining the nectar by having to puncture the nectary at several points. On the other hand, in all the Ophreæ which have nectar ready stored within the nectary, the discs are sufficiently viscid for the attachment of the pollinia to insects, without the matter quickly setting hard; and there would therefore be no advantage to these plants in insects being delayed for a few minutes whilst sucking the flowers.

In the case of cultivated exotic Orchids which have a nectary, without any free nectar, it is of course impossible to feel absolutely sure that it would not contain any under more natural conditions. Nor have I made many comparative observations on the rate of the setting hard of the viscid matter of the disc in exotic forms. Nevertheless it seems that some Vandeæ are in the same predicament as our British species of Orchis; thus Calanthe masuca has a very long nectary, which in all the specimens examined by me was quite dry internally, and was inhabited by powdery Cocci; but in the intercellular spaces between the two coats there was much fluid; and in this species the viscid matter of the disc, after its surface had been disturbed, entirely lost its adhesiveness in two minutes. In an Oncidium the disc, similarly disturbed, became dry in one minute and a half; in an Odontoglossum in two minutes; and in neither of these Orchids was there any free nectar. On the other hand, in Angræcum sesquipedale, which has free nectar stored within the lower end of the nectary, the disc of the pollinium, when removed from the plant and with its surface disturbed, was strongly adhesive after forty-eight hours.

Sarcanthus teritifolius offers a more curious case. The disc quite lost its viscidity and set hard in less than three minutes. Hence it might have been expected that no fluid would have been found in the nectary, but only in the intercellular spaces; nevertheless there was fluid in both places, so that here we have both conditions combined in the same flower. It is probable that insects would sometimes rapidly suck the free nectar and neglect that between the two coats; but even in this case I strongly suspect that they would be delayed by a totally different means in sucking the free nectar, so as to allow the viscid matter to set hard. In this plant, the labellum with its nectary is an extraordinary organ. I wished to have had a drawing made of its structure; but found that it was as hopeless as to give a drawing of the wards of a complicated lock. Even the skilful Bauer, with numerous figures and sections on a large scale, hardly makes the structure intelligible. So complicated is the passage, that I failed in repeated attempts to pass a bristle from the outside of the flower into the nectary; or in a reversed direction from the cut-off end of the nectary to the outside. No doubt an insect with a voluntarily flexible proboscis could pass it through the passages, and thus reach the nectar; but in effecting this, some delay would be caused; and time would be thus allowed for the curious square viscid disc to become securely cemented to an insect's head or body.

As in Epipactis the cup at the base of the labellum serves as a nectar-receptacle, I expected to find that the analogous cups in Stanhopea, Acropera, &c., would serve for the same purpose; but I could never find a drop of nectar in them. According, also, to M. Ménière and Mr. Scott^[8] this is never the case in these genera, or in Gongora, Cirrhæa, and many others. In Catasetum tridentatum, and in the female form Monachanthus, we see that the upturned cup cannot possibly serve as a nectar-receptacle. What then attracts insects to these flowers? That they must be attracted is certain; more especially in the case of Catasetum, in which the sexes stand on separate plants. In many genera of Vandeæ there is no trace of any nectar-secreting organ or receptacle; but in all these cases (as far as I have seen), the labellum is either thick and fleshy, or is furnished with extraordinary excrescences, as in the genera Oncidium and Odontoglossum. In Phalænopsis grandiflora there is a curious anvil-shaped projection on the labellum, with two tendril-like prolongations from its extremity which turn backwards and apparently serve to guard the sides of the anvil, so that insects would be forced to alight on its crown. Even in our British Cephalanthera grandiflora, the labellum of which never contains nectar, there are orange-coloured ribs and papillæ on the inner surface which faces the column. In Calanthe (fig. 26) a cluster of odd little spherical warts projects from the labellum, and there is an extremely long nectary, which does not include nectar; in Eulophia viridis the short nectary is equally destitute of nectar, and the labellum is covered with longitudinal, fimbriated ridges. In several species of Ophrys, there are two small shining protuberances, at the base of the labellum, beneath the two discs. Innumerable other cases could be added of the presence of singular and diversified excrescences on the labellum; and Lindley remarks that their use is quite unknown.

From the position, relatively to the viscid discs, which these excrescences occupy, and from the absence of any free nectar, it formerly seemed to me highly probable that they afforded food and thus attracted either Hymenoptera or flower-feeding Coleoptera. There is no more inherent improbability in a flower being habitually fertilised by an insect coming to feed on the labellum, than in seeds being habitually disseminated by birds attracted by the sweet pulp in which they are embedded. But I am bound to state that Dr. Percy, who had the thick and furrowed labellum of a Warrea analysed for me by fermentation over mercury, found that it gave no evidence of containing more saccharine matter than the other petals. On the other hand, the thick labellum of Catasetum and the bases of the upper petals of Mormodes ignea, have a slightly sweet, rather pleasant, and nutritious taste. Nevertheless, it was a bold speculation that insects were attracted to the flowers of various Orchids in order to gnaw the excrescences or other parts of their labella; and few things have given me more satisfaction than the full confirmation of this view by Dr. Crüger, who^[9] has repeatedly witnessed in the West Indies humble-bees of the genus Euglossa gnawing the labellum of Catasetum, Coryanthes, Gongora, and Stanhopea. Fritz Müller also has often found, in South Brazil, the prominences on the labellum of Oncidium gnawed. We are thus enabled to understand the meaning of the various extraordinary crests and projections on the labellum of many Orchids; for they invariably stand in such a position that insects, whilst gnawing them, would be almost sure to touch the viscid discs of the pollinia and thus remove them, afterwards effecting the fertilisation of another flower.

MOVEMENTS OF THE POLLINIA.

The pollinia of many Orchids undergo a movement of depression, after they have been removed from their places of attachment and have been exposed for a few seconds to the air. This is due to the contraction of a portion, sometimes to an exceedingly minute portion, of the exterior surface of the rostellum, which retains a membranous condition. This membrane, as we have seen, is likewise sensitive to a touch, so as to rupture in certain definite lines. In a Maxillaria the middle part of the pedicel, and in Habenaria the whole drum-like pedicel contracts. The point of contraction in all the other cases seen by me, is either close to the surface of attachment of the caudicle to the disc, or at the point where the pedicel is united to the disc; but both the disc and pedicel are parts of the exterior surface of the rostellum. In these remarks I do not refer to the movements which are simply due to the elasticity of the pedicel, as in the Vandeæ.

The long strap-formed disc of Gymnadenia conopsea is well adapted to show the mechanism of the movement of depression. The whole pollinium, both in its upright and depressed (but not closely depressed) position, has been shown (p. 65) in fig. 10. The disc, in its uncontracted condition with the caudicle removed, is seen from above highly magnified in the upper of the two adjoining figures; and in the lower figure we have a longitudinal section of the uncontracted disc, together with the base of the attached and upright caudicle. At the broad end of the disc there is a deep crescent-shaped depression, bordered by a slight ridge formed of longitudinally elongatedFig. 38.Disc of Gymnadenia conopsea. cells. The end of the caudicle is attached to the steep sides of this depression and ridge. When the disc is exposed to the air for about thirty seconds, the ridge contracts and sinks flat down; in sinking, it drags with it the caudicle, which then lies parallel to the elongated tapering part of the disc. If placed in water the ridge rises, re-elevating the caudicle, and when re-exposed to the air it sinks again, but each time with somewhat enfeebled power. During each sinking and rising of the caudicle, the whole pollinium is of course depressed and elevated.

That the power of movement lies exclusively in the surface of the disc is well shown in the case of the saddle-shaped disc of Orchis pyramidalis; for whilst it was held under water I removed the attached caudicles and the layer of viscid matter from the inferior surface, and immediately that the disc was exposed to the air the proper contraction ensued. The disc is formed of several layers of minute cells, which are best seen in specimens that have been kept in spirits of wine, for their contents are thus rendered more opaque. The cells in the flaps of the saddle are a little elongated. As long as the saddle is kept damp, its upper surface is nearly flat, but when exposed to the air (see fig. 3, E, p. 18) the two flaps or sides contract and curl inwards; and this causes the divergence of the pollinia. By a kind of contraction two valleys are likewise formed in front of the caudicles, so that the latter are thrown forwards and downwards, almost in the same way as if trenches were dug in front of two upright poles, and then carried on so as to undermine them. As far as I could perceive, an analogous contraction causes the depression of the pollinia in Orchis mascula. With O. hircina both pollinia are attached to a single rather large square disc, the whole front of which, after exposure to the air, sinks down and is then separated from the hinder part by an abrupt step. By this contraction both pollinia are carried forwards and downwards.

Some pollinia which had been gummed on card for several months, when placed in water, rose up and afterwards underwent the movement of depression. A fresh pollinium, on being alternately damped and exposed to the air, rises and sinks several times alternately. Before I had ascertained these facts, which show that the movement is simply hygrometric, I thought that it was a vital action, and tried vapour of chloroform and of prussic acid, and immersion in laudanum; but these reagents did not check the movement. Nevertheless, there are some difficulties in understanding how the movement can be simply hygrometric. The flaps of the saddle in Orchis pyramidalis (see fig. 3, D, p. 18) curl completely inwards in nine seconds, which is a surprisingly short time for mere evaporation to produce an effect;^[10] and the movement is apparently due to the drying of the under surface, although this is covered with a thick layer of viscid matter. The edges, however, of the saddle might become slightly dry in the nine seconds. When the saddle-formed disc is placed in spirits of wine it contracts energetically; and this is probably due to the attraction of alcohol for water. When replaced in water it opens again. Whether or not the contraction is wholly hygrometric, the movements are admirably regulated in each species, so that the pollen-masses, when transported by insects from flower to flower, assume a proper position for striking the stigmatic surface.

These various movements would be quite useless, unless the pollinia were attached in a uniform position to the insects which visit the flowers so as to be always directed in the same manner after the movement of depression; and this necessitates that the insects should be forced to visit the flowers of the same species in a uniform manner. Hence I must say a few words on the sepals and petals. Their primary function, no doubt, is to protect the organs of fructification in the bud. After the flower is fully expanded, the upper sepal and two upper petals often continue the same office. We cannot doubt that this protection is of service, when we see in Stelis the sepals so neatly reclosing and reprotecting the flower some time after its expansion; in Masdevallia the sepals are permanently soldered together, with two little windows alone left open; and in the open and exposed flowers of Bolbophyllum, the mouth of the stigmatic chamber after a time closes. Analogous facts with respect to Malaxis, Cephalanthera, &c., could be given. But the hood formed by the upper sepal and two upper petals, besides affording protection, evidently forms a guide, compelling insects to visit the flowers in front. Few persons now doubt the correctness of C. K. Sprengel's view,^[11] that the bright and conspicuous colours of flowers serve to attract insects from a distance. Nevertheless some Orchids have singularly inconspicuous and greenish flowers, perhaps in order to escape some danger; but many of these are strongly scented, which would equally well serve to attract insects.

The labellum is by far the most important of the external envelopes of the flower. It not only secretes nectar, but is often modelled into variously shaped receptacles for holding this fluid, or is itself rendered attractive so as to be gnawed by insects. Unless the flowers were by some means rendered attractive, most of the species would be cursed with perpetual sterility. The labellum always stands in front of the rostellum, and its outer portion often serves as a landing-place for the necessary visitors. In Epipactis palustris this part is flexible and elastic, and apparently compels insects in retreating to brush against the rostellum. In Cypripedium the distal portion is folded over like the end of a slipper, and compels insects to crawl out of the flower by one of two special passages. In Pterostylis and a few other Orchids the labellum is irritable, so that when touched it shuts the flower, leaving only a single passage by which an insect can escape. In Spiranthes, when the flower is fully mature, the column moves from the labellum, space being thus left for the introduction of the pollen-masses attached to the proboscis of a humble-bee. In Mormodes ignea the labellum is perched on the summit of the column, and here insects alight and touch a sensitive point, causing the ejection of the pollen-masses. The labellum is often deeply channelled, or has guiding ridges, or is pressed closely against the column; and in a multitude of cases it approaches closely enough to render the flower tubular. By these several means insects are forced to brush against the rostellum. We must not, however, suppose that every detail of structure in the labellum is of use: in some instances, as with Sarcanthus, its extraordinary shape seems to be partly due to its development in close apposition to the curiously shaped rostellum.

In Listera ovata the labellum stands far from the column, but its base is narrow, so that insects are led to stand exactly beneath the middle of the rostellum, In other cases, as in Stanhopea, Phalænopsis, Gongora, &c., the labellum is furnished with upturned basal lobes, which manifestly act as lateral guides. In some cases, as in Malaxis, the two upper petals are curled backwards so as to be out of the way; in other cases as in Acropera, Masdevallia, and some Bolbophyllums, these upper petals plainly serve as lateral guides, compelling insects to visit the flowers directly in front of the rostellum. In other cases, wings formed by the margins of the clinandrum or of the column, serve as lateral guides, both in the withdrawal of the pollinia and in their subsequent insertion into the stigmatic cavity. So that there can be no doubt that the petals, sepals and rudimentary anthers do good service in several ways, besides affording protection to the bud.

The final end of the whole flower, with all its parts, is the production of seed; and these are produced by Orchids in vast profusion. Not that such profusion is anything to boast of; for the production of an almost infinite number of seeds or eggs, is undoubtedly a sign of lowness of organisation. That a plant, not being an annual, should escape extinction, chiefly by the production of a vast number of seeds or seedlings, shows a poverty of contrivance, or a want of some fitting protection against other dangers. I was curious to estimate the number of seeds produced by some few Orchids; so I took a ripe capsule of Cephalanthera grandiflora, and arranged the seeds on a long ruled line as equably as I could in a narrow hillock; and then counted the seeds in an accurately measured length of one-tenth of an inch. In this way the contents of the capsule were estimated at 6020 seeds, and very few of these were bad; the four capsules borne by the same plant would have therefore contained 24,080 seeds. Estimating in the same manner the smaller seeds of Orchis maculata, I found the number nearly the same, viz., 6200; and, as I have often seen above thirty capsules on the same plant, the total amount would be 186,300. As this Orchid is perennial, and cannot in most places be increasing in number, one seed alone of this large number yields a mature plant once in every few years.

To give an idea what the above figures really mean, I will briefly show the possible rate of increase of O. maculata: an acre of land would hold 174,240 plants, each having a space of six inches square, and this would be just sufficient for their growth; so that, making the fair allowance of 400 bad seeds in each capsule, an acre would be thickly clothed by the progeny of a single plant. At the same rate of increase, the grandchildren would cover a space slightly exceeding the island of Anglesea; and the great grandchildren of a single plant would nearly (in the ratio of 47 to 50) clothe with one uniform green carpet the entire surface of the land throughout the globe. But the number of seeds produced by one of our common British orchids is as nothing compared to that of some of the exotic kinds. Mr. Scott found that the capsule of an Acropera contained 371,250 seeds; and judging from the number of flowers, a single plant would sometimes yield about seventy-four millions of seeds. Fritz Müller found 1,756,440 seeds in a single capsule of a Maxillaria; and the same plant sometimes bore half-a-dozen such capsules. I may add that by counting the packets of pollen (one of which was broken up under the microscope) I estimated that the number of pollen-grains, each of which emits its tube, in a single anther of Orchis mascula was 122,400. Amici^[12] estimated the number in O. morio at 120,300. As these two species apparently do not produce more seed than the allied O. maculata, a capsule of which contained 6200 seeds, we see that there are about twenty pollen-grains for each ovule. According to this standard, the number of pollen-grains in the anther of a single flower of the Maxillaria which yielded 1,756,440 seeds must be prodigious.

What checks the unlimited multiplication of the Orchideæ throughout the world is not known. The minute seeds within their light coats are well fitted for wide dissemination; and I have several times observed seedlings springing up in my orchard and in a newly-planted wood, which must have come from a considerable distance. This was especially the case with Epipactis latifolia; and an instance has been recorded by a good observer^[13] of seedlings of this plant appearing at the distance of between eight and ten miles from any place where it grew. Notwithstanding the astonishing number of seeds produced by Orchids, it is notorious that they are sparingly distributed; for instance, Kent appears to be the most favourable county in England for the order, and within a mile of my house nine genera, including thirteen species, grow; but of these one alone, Orchis morio, is sufficiently abundant to make a conspicuous feature in the vegetation; as is O. maculata in a lesser degree in open woodlands. Most of the other species, though not deserving to be called rare, are sparingly distributed; yet, if their seeds or seedlings were not largely destroyed, any one of them would immediately cover the whole land. In the tropics the species are very much more numerous; thus Fritz Müller found in South Brazil more than thirteen kinds belonging to several genera growing on a single Cedrela tree. Mr. Fitzgerald has collected within the radius of one mile of Sydney in Australia no less than sixty-two species, of which fifty-seven were terrestrial. Nevertheless the number of individuals of the same species is, I believe, in no country nearly so great as that of very many other plants. Lindley formerly estimated that there were in the world about 6000 species of Orchideæ, included in 433 genera.^[14]

The number of the individuals which come to maturity does not seem to be at all closely determined by the number of seeds which each species produces; and this holds good when closely related forms are compared. Thus Ophrys apifera fertilises itself and every flower produces a capsule; but the individuals of this species are not so numerous in some parts of England as those of O. muscifera, which cannot fertilise itself and is imperfectly fertilised by insects, so that a large proportion of the flowers drop off unimpregnated. Ophrys aranifera is found in large numbers in Liguria, yet Delpino estimates that not more than one out of 3000 flowers produces a capsule.^[15] Mr. Cheeseman says^[16] that with the New Zealand Pterostylis trullifolia much less than a quarter of the flowers, which are beautifully adapted for cross-fertilisation, yield capsules; whereas with the allied Acianthus sinclairii, the flowers of which equally require insect-aid for their fertilisation, seventy-one capsules were produced by eighty-seven flowers; so that this plant must produce an extraordinary number of seeds; nevertheless in many districts it is not at all more abundant than the Pterostylis. Mr. Fitzgerald, who in Australia has particularly attended to this subject, remarks that every flower of Thelymitra carnea fertilises itself and produces a capsule; yet it is not nearly so common as Acianthus fornicatus, "the majority of the flowers of which are unproductive. Phajus grandifolius and Calanthe veratrifolia grow in similar situations. Every flower of the Phajus produces seeds, only occasionally one of the Calanthe, yet Phajus is rare and Calanthe common."

The frequency with which throughout the world members of various Orchideous tribes fail to have their flowers fertilised, though these are excellently constructed for cross-fertilisation, is a remarkable fact. Fritz Müller informs me that this holds good in the luxuriant forests of South Brazil with most of the Epidendreæ, and with the genus Vanilla. For instance, he visited a site where Vanilla creeps over almost every tree, and although the plants had been covered with flowers, yet only two seed-capsules were produced. So again with an Epidendrum, 233 flowers had fallen off unimpregnated and only one capsule had been formed; of the still remaining 136 flowers, only four had their pollinia removed. In New South Wales Mr. Fitzgerald does not believe that more than one flower out of a thousand of Dendrobium speciosum sets a capsule; and some other species there are very sterile. In New Zealand over 200 flowers of Coryanthes triloba yielded only five capsules; and at the Cape of Good Hope only the same number were produced by 78 flowers of Disa grandiflora. Nearly the same result has been observed with some of the species of Ophrys in Europe. The sterility in these cases is very difficult to explain. It manifestly depends on the flowers being constructed with such elaborate care for cross-fertilisation, that they cannot yield seeds without the aid of insects. From the evidence which I have given elsewhere^[17] we may conclude that it would be far more profitable to most plants to yield a few cross-fertilised seeds, at the expense of many flowers dropping off unimpregnated, rather than produce many self-fertilised seeds. Profuse expenditure is nothing unusual under nature, as we see with the pollen of wind-fertilised plants, and in the multitude of seeds and seedlings produced by most plants in comparison with the few that reach maturity. In other cases the paucity of the flowers that are impregnated may be due to the proper insects having become rare under the incessant changes to which the world is subject; or to other plants which are more highly attractive to the proper insects having increased in number. We know that certain Orchids require certain insects for their fertilisation, as in the cases before given of Vanilla and Sarcochilus. In Madagascar Angræcum sesquipedale must depend on some gigantic moth. In Europe Cypripedium calceolus appears to be fertilised only by small bees of the genus Andrena, and Epipactis latifolia only by wasps. In those cases in which only a few flowers are impregnated owing to the proper insects visiting only a few, this may be a great injury to the plant; and many hundred species throughout the world have been thus exterminated; those which survive having been favoured in some other way. On the other hand, the few seeds which are produced in these cases will be the product of cross-fertilisation, and this as we now positively know is an immense advantage to most plants.

I have now nearly finished this volume, which is perhaps too lengthy. It has, I think, been shown that the Orchideæ exhibit an almost endless diversity of beautiful adaptations. When this or that part has been spoken of as adapted for some special purpose, it must not be supposed that it was originally always formed for this sole purpose. The regular course of events seems to be, that a part which originally served for one purpose, becomes adapted by slow changes for widely different purposes. To give an instance: in all the Ophreæ, the long and nearly rigid caudicle manifestly serves for the application of the pollen-grains to the stigma, when the pollinia are transported by insects to another flower; and the anther opens widely in order that the pollinium should be easily withdrawn; but in the Bee Ophrys, the caudicle, by a slight increase in length and decrease in its thickness, and by the anther opening a little more widely, becomes specially adapted for the very different purpose of self-fertilisation, through the combined aid of the weight of the pollen-mass and the vibration of the flower when moved by the wind. Every gradation between these two states is possible,—of which we have a partial instance in O. aranifera.

Again, the elasticity of the pedicel of the pollinium in some Vandeæ is adapted to free the pollen-masses from their anther-cases; but by a further slight modification, the elasticity of the pedicel becomes specially adapted to shoot out the pollinium with considerable force so as to strike the body of the visiting insect. The great cavity in the labellum of many Vandeæ is gnawed by insects and thus attracts them; but in Mormodes ignea it is greatly reduced in size, and serves in chief part to keep the labellum in its new position on the summit of the column. From the analogy of many plants we may infer that a long spur-like nectary is primarily adapted to secrete and hold a store of nectar; but in many Orchids it has so far lost this function, that it contains fluid only in the intercellular spaces. In those Orchids in which the nectary contains both free nectar and fluid in the intercellular spaces, we can see how a transition from the one state to the other could be effected, namely, by less and less nectar being secreted from the inner membrane, with more and more retained within the intercellular spaces. Other analogous cases could be given.

Although an organ may not have been originally formed for some special purpose, if it now serves for this end, we are justified in saying that it is specially adapted for it. On the same principle, if a man were to make a machine for some special purpose, but were to use old wheels, springs, and pulleys, only slightly altered, the whole machine, with all its parts, might be said to be specially contrived for its present purpose. Thus throughout nature almost every part of each living being has probably served, in a slightly modified condition, for diverse purposes, and has acted in the living machinery of many ancient and distinct specific forms.

In my examination of Orchids, hardly any fact has struck me so much as the endless diversities of structure,—the prodigality of resources,—for gaining the very same end, namely, the fertilisation of one flower by pollen from another plant. This fact is to a large extent intelligible on the principle of natural selection. As all the parts of a flower are co-ordinated, if slight variations in any one part were preserved from being beneficial to the plant, then the other parts would generally have to be modified in some corresponding manner. But these latter parts might not vary at all, or they might not vary in a fitting manner, and these other variations, whatever their nature might be, which tended to bring all the parts into more harmonious action with one another, would be preserved by natural selection.

To give a simple illustration: in many Orchids the ovarium (but sometimes the foot-stalk) becomes for a period twisted, causing the labellum to assume the position of a lower petal, so that insects can easily visit the flower; but from slow changes in the form or position of the petals, or from new sorts of insects visiting the flowers, it might be advantageous to the plant that the labellum should resume its normal position on the upper side of the flower, as is actually the case with Malaxis paludosa, and some species of Catasetum, &c. This change, it is obvious, might be simply effected by the continued selection of varieties which had their ovaria less and less twisted; but if the plant only afforded varieties with the ovarinm more twisted, the same end could be attained by the selection of such variations, until the flower was turned completely round on its axis. This seems to have actually occurred with Malaxis paludosa, for the labellum has acquired its present upward position by the ovarium being twisted twice as much as is usual.

Again, we have seen that in most Vandeæ there is a plain relation between the depth of the stigmatic chamber and the length of the pedicel, by which the pollen-masses are inserted; now if the chamber became slightly less deep from any change in the form of the column or other unknown cause, the mere shortening of the pedicel would be the simplest corresponding change; but if the pedicel did not happen to vary in shortness, the slightest tendency to its becoming bowed from elasticity as in Phalænopsis, or to a backward hygrometric movement as in one of the Maxillarias, would be preserved, and the tendency would be continually augmented by selection; thus the pedicel, as far as its action is concerned, would be modified in the same manner as if thad been shortened. Such processes carried on during many thousand generations in various ways, would create an endless diversity of co-adapted structures in the several parts of the flower for the same general purpose. This view affords, I believe, the key which partly solves the problem of the vast diversity of structure adapted for closely analogous ends in many large groups of organic beings.

The more I study nature, the more I become impressed with ever-increasing force, that the contrivances and beautiful adaptations slowly acquired through each part occasionally varying in a slight degree but in many ways, with the preservation of those variations which were beneficial to the organism under complex and ever-varying conditions of life, transcend in an incomparable manner the contrivances and adaptations which the most fertile imagination of man could invent.

The use of each trifling detail of structure is far from a barren search to those who believe in natural selection. When a naturalist casually takes up the study of an organic being, and does not investigate its whole life (imperfect though that study will ever be), he naturally doubts whether each trifling point can be of any use, or indeed whether it be due to any general law. Some naturalists believe that numberless structures have been created for the sake of mere variety and beauty,—much as a workman would make different patterns. I, for one, have often and often doubted whether this or that detail of structure in many of the Orchideæ and other plants could be of any service; yet, if of no good, these structures could not have been modelled by the natural preservation of useful variations; such details can only be vaguely accounted for by the direct action of the conditions of life, or the mysterious laws of correlated growth.

To give nearly all the instances of trifling details of structure in the flowers of Orchids, which are certainly of high importance, would be to recapitulate almost the whole of this volume. But I will recall to the reader's memory a few cases. I do not here refer to the fundamental framework of the plant, such as the remnants of the fifteen primary organs arranged alternately in the five whorls; for almost everyone who believes in the gradual evolution of species will admit that their presence is due to inheritance from a remote parent-form. Innumerable facts with respect to the uses of the variously shaped and placed petals and sepals have been given. So again, the importance of as light difference in the shape of the caudicle of the pollinium of the Bee Ophrys, compared with that of the other species of the same genus, has likewise been referred to; to this might be added the doubly-bent caudicle of the Fly Ophrys. Indeed, the important relation of the length and shape of the caudicle, with reference to the position of the stigma, might be cited throughout many whole tribes. The solid projecting knob of the anther in Epipactis palustris, which does not include pollen, liberates the pollen-masses when it is moved by insects. In Cephalanthera grandiflora, the upright position of the almost closed flower protects the slightly coherent pillars of pollen from disturbance. The length and elasticity of the filament of the anther in certain species of Dendrobium apparently serves for self-fertilisation, if insects fail to transport the pollen-masses. The slight forward inclination of the crest of the rostellum in Listera prevents the anther-case being caught as soon as the viscid matter is ejected. The elasticity of the lip of the rostellum in Orchis causes it to spring up again when only one of the pollen-masses has been removed, thus keeping the second viscid disc ready for action, which otherwise would be wasted. No one who had not studied Orchids would have suspected that these and very many other small details of structure were of the highest importance to each species; and that consequently, if the species were exposed to new conditions of life, and the structure of the several parts varied ever so little, the smallest details of structure might readily be acquired through natural selection. These cases afford a good lesson of caution with respect to the importance of apparently trifling particulars of structure in other organic beings.

It may naturally be inquired, Why do the Orchideæ exhibit so many perfect contrivances for their fertilisation? From the observations of various botanists and my own, I am sure that many other plants offer analogous adaptations of high perfection; but it seems that they are really more numerous and perfect with the Orchideæ than with most other plants. To a certain extent this inquiry can be answered. As each ovule requires at least one, probably several, pollen-grains,^[18] and as the seeds produced by Orchids are so inordinately numerous, we can see that it is necessary that large masses of pollen should be left on the stigma of each flower. Even in the Neotteæ, which have granular pollen, with the grains tied together by weak threads, I have observed that considerable masses of pollen are generally left on the stigmas. This circumstance apparently explains why the grains cohere in packets or large waxy masses, as they do in so many tribes, namely, to prevent waste in the act of transportal. The flowers of most plants produce pollen enough to fertilise several flowers, so as to allow of or to favour cross-fertilisation. But with the many Orchids which produce only two pollen-masses, and with some of the Malaxeæ which produce only one, the pollen from a single flower cannot possibly fertilise more than two flowers or only a single one; and cases of this kind do not occur, as I believe, in any other group of plants. If the Orchideæ had elaborated as much pollen as is produced by other plants, relatively to the number of seeds which they yield, they would have had to produce a most extravagant amount, and this would have caused exhaustion. Such exhaustion is avoided by pollen not being produced in any great superfluity owing to the many special contrivances for its safe transportal from plant to plant, and for placing it securely on the stigma. Thus we can understand why the Orchideæ are more highly endowed in their mechanism for cross-fertilisation, than are most other plants.

In my work on the "Effects of Cross and Self Fertilisation in the Vegetable Kingdom," I have shown that when flowers are cross-fertilised they generally receive pollen from a distinct plant and not that from another flower on the same plant; a cross of this latter kind doing little or no good. I have further shown that the benefits derived from a cross between two plants depends altogether on their differing somewhat in constitution; and there is much evidence that each individual seedling possesses its own peculiar constitution. The crossing of distinct plants of the same species is favoured or determined in various ways, as described in the above work, but chiefly by the prepotent action of pollen from another plant over that from the same flower. Now with the Orchideæ it is highly probable that such prepotency prevails, for we know from the valuable observations of Mr. Scott and Fritz Müller,^[19] that with several Orchids pollen from their own flower is quite impotent, and is even in some cases poisonous to the stigma. Besides this prepotency, the Orchideæ present various special contrivances—such as the pollinia not assuming a proper position for striking the stigma until some time has elapsed after their removal from the anthers—the slow curving forwards and then backwards of the rostellum in Listera and Neottia—the slow movement of the column from the labellum in Spiranthes—the diœcious condition of Catasetum—the fact of some species producing only a single lower, &c.—all render it certain or highly probable that the flowers are habitually fertilised with pollen from a distinct plant.

That cross-fertilisation, to the complete exclusion of self-fertilisation, is the rule with the Orchideæ, cannot be doubted from the facts already given in relation to many species in all the tribes throughout the world. I could almost as soon believe that flowers in general were not adapted for the production of seeds, because there are a few plants which have never been known to yield seed, as that the flowers of the Orchideæ are not as a general rule adapted so as to ensure cross-fertilisation. Nevertheless, some species are regularly or often self-fertilised; and I will now give a list of all the cases hitherto observed by myself and others. In some of these the flowers appear often to be fertilised by insects, but they are capable of fertilising themselves without aid, though in a more or less incomplete manner; so that they do not remain utterly barren if insects fail to visit them. Under this head may be included three British species, namely, Cephalanthera grandiflora, Neottia nidus-avis, and perhaps Listera ovata. In South Africa Disa macrantha often fertilises itself; but Mr. Weale believes that it is likewise cross-fertilised by moths. Three species belonging to the Epidendreæ rarely open their flowers in the West Indies; nevertheless these flowers fertilise themselves, but it is doubtful whether they are fully fertilised, for a large proportion of the seeds spontaneously produced by some members of this tribe in a hothouse were destitute of an embryo. Some species of Dendrobium, judging from their structure and from their occasionally producing capsules under cultivation, likewise come under this head.

Of species which regularly fertilise themselves without any aid and yield full-sized capsules, hardly any case is more striking than that of Ophrys apifera, which was advanced by me in the first edition of this work. To this case may now be added two other European plants, Orchis or Neotinea intacta and Epipactis viridiflora. Two North American species, Gymnadenia tridentata and Platanthera hyperborea appear to be in the same predicament, but whether when self-fertilised they yield a full complement of capsules containing good seeds has not been ascertained. A curious Epidendrum in South Brazil which bears two additional anthers fertilises itself freely by their aid; and Dendrobium cretaceum has been known to produce perfect self-fertilised seeds in a hothouse in England. Lastly, Spiranthes australis and two species of Thelymitra, inhabitants of Australia, come under this same head. No doubt other cases will hereafter be added to this short list of about ten species which it appears can fertilise themselves fully, and of about the same number of species which fertilise themselves imperfectly when insects are excluded.

It deserves especial attention that the flowers of all the above-named self-fertile species still retain various structures which it is impossible to doubt are adapted for insuring cross-fertilisation, though they are now rarely or never brought into play. We may therefore conclude that all these plants are descended from species or varieties which were formerly fertilised by insect-aid. Moreover, several of the genera to which these self-fertile species belong, include other species, which are incapable of self-fertilisation. Thelymitra offers indeed the only instance known to me of two species within the same genus which regularly fertilise themselves. Considering such cases as those of Ophrys, Disa, and Epidendrum, in which one species alone in the genus is capable of complete self-fertilisation, whilst the other species are rarely fertilised in any manner owing to the rarity of the visits of the proper insects;—bearing also in mind the large number of species in many parts of the world which from this same cause are seldom impregnated, we are led to believe that the above-named self-fertile plants formerly depended on the visits of insects for their fertilisation, and that from such visits failing they did not yield a sufficiency of seed and were verging towards extinction. Under these circumstances it is probable that they were gradually modified, so as to become more or less completely self-fertile; for it would manifestly be more advantageous to a plant to produce self-fertilised seeds rather than none at all or extremely few seeds. Whether any species which is now never cross-fertilised will be able to resist the evil effects of long-continued self-fertilisation, so as to survive for as long an average period as the other species of the same genera which are habitually cross-fertilised, cannot of course be told. But Ophrys apifera is still a highly vigorous plant, and Gymnadenia tridentata and Platanthera hyperborea are said by Asa Gray to be common plants in North America. It is indeed possible that these self-fertile species may revert in the course of time to what was undoubtedly their pristine condition, and in this case their various adaptations for cross-fertilisation would be again brought into action. We may believe that such reversion is possible, when we hear from Mr. Moggridge that Ophrys scolopax fertilises itself freely in one district of Southern France without the aid of insects, and is completely sterile without such aid in another district.

Finally, if we consider how precious a substance pollen is, and what care has been bestowed on its elaboration and on the accessory parts in the Orchideæ,—considering how large an amount is necessary for the impregnation of the almost innumerable seeds produced by these plants,—considering that the anther stands close behind or above the stigma, self-fertilisation would have been an incomparably safer and easier process than the transportal of pollen from flower to flower. Unless we bear in mind the good effects which have been proved to follow in most cases from cross-fertilisation, it is an astonishing fact that the flowers of the Orchideæ should not have been regularly self-fertilised. It apparently demonstrates that there must be something injurious in this latter process, of which fact I have elsewhere given direct proof. It is hardly an exaggeration to say that Nature tells us, in the most emphatic manner, that she abhors perpetual self-fertilisation.

---

1. 'Beiträge fur Kenntniss der Befruchtung,' 1844, p. 236.
2. Professor Babington ('Manual of British Botany,' 3rd edit.) uses the existence of this "rostellate process" as a character to separate Orchis, Gymnadenia, and Aceras from the other genera of Ophreæ. The group of spiral vessels, properly belonging to the rostellum, runs up, and even into, the base of this crest or process.
3. In several cases I have observed four tubes emitted from the four grains which form one of the compound grains. In some semi-monstrous flowers of Malaxis paludosa, and of Aceras anthropophora, and in perfect flowers of Neottia nidus-avis, I have observed tubes emitted from the pollen- grains, whilst still within the anther and not in contact with the stigma. I have thought this worth mentioning as K. Brown ('Linn. Transact.' vol. xvi. p. 729) states, apparently with some surprise, that the pollen-tubes were emitted from the pollen, whilst still within the anther, in a decaying flower of Asclepias. These cases show that the protruding tubes are, at least at first, formed exclusively at the expense of the contents of the pollen-grains.

   Having alluded to the monstrous flowers of the Aceras, I will add that I examined several (always the lowest on the spike) in which the labellum was hardly developed, and was pressed close against the stigma. The rostellum was not developed, so that the pollinia did not possess viscid discs; but the most curious feature was, that the two anther-cells had become, apparently in consequence of the position of the rudimentary labellum, widely separated, and were joined by a connective membrane, almost as broad as that of Habenaria clorantha!

4. Auguste de Saint Hilaire ('Leçons de Botanique,' &c., 1841, p. 447) says that the elastic threads exist in the early bud, after the pollen-grains have been partly formed, as a thick creamy fluid. He adds that his observations on Ophrys apifera have shown him that this fluid is secreted by the rostellum, and is slowly forced drop by drop into the anther. Had not so eminent an authority made this statement, I should not have noticed it. It is certainly erroneous. In buds of Epipactis latifolia I opened the anther. whilst perfectly closed and free from the rostellum, and found the pollen-grains united by elastic threads. Cephalanthera grandiflora has no rostellum to secrete the above thick fluid, yet the pollen-grains are thus united. In a monstrous specimen of Orchis pyramidalis the auricles, or rudimentary anthers on each side of the proper anther, had become partly developed, and they stood quite on one side of the rostellum and stigma; yet I found in one of these auricles a distinct caudicle (which necessarily had no disc at its extremity), and this caudicle could not possibly have been secreted by the rostellum or stigma. I could advance additional evidence, but it would be superfluous.
5. The fullest and the most able discussion on this difficult subject is by Professor H. G. Bronn in his 'Entwickelungs-Gesetze der Organischen Welt,' 1858.
6. This subject has been fully discussed in my work 'On the Effects of Cross and Self-fertilisation in the Vegetable Kingdom,' 1876, p. 402.
7. Kurr, 'Ueber die Bedeutung der Nektarien,' 1833, p. 28.
8. 'Bulletin Bot. Soc. de France,' tom. ii. 1855, p. 352.
9. 'Journ. Linn. Soc. Bot.' 1864, vol. viii. p. 129.
10. This fact does not now appear to me so surprising as it formerly did, for my son Francis has shown ('Transact. Linn. Soc.' 2nd series, Bot. vol. i. 1876, p. 149) with what extraordinary quickness the awn of Stipa twists and untwists when exposed to dry and damp air. These movements being due, as he has shown, to the twisting and untwisting of the separate cells.
11. This author's curious work, with its quaint title of 'Das Entdeckte Geheimniss der Natur,' until lately was often spoken lightly of. No doubt he was an enthusiast, and perhaps carried some of his ideas to an extreme length. But I feel sure, from my own observations, that his work contains an immense body of truth. Many years ago Robert Brown, to whose judgment all botanists defer, spoke highly of it to me, and remarked that only those who knew little of the subject would laugh at him.
12. Mohl, 'The Vegetable Cell,' translated by Henfrey, p. 133.
13. Mr. Bree, in 'London's Mag. of Nat. Hist,' vol. ii. 1829, p. 70.
14. 'Gardener's Chron.' 1862, p. 192.
15. 'Ult. Osservaz. sulla Dicogamia,' part i. p. 177.
16. 'Transact. New Zealand Inst.' vol. vii. 1875, p. 351.
17. 'The Effects of Cross and Self-fertilisation in the Vegetable Kingdom,' 1876.
18. Gärtner, 'Beitrage zur Kenntniss der Befruchtung,' 1844, p. 135.
19. A full abstract of these observations is given in my 'Variatlon of Animals and Plants under Domestication,' ch. xvii. 2nd edit, vol. ii. p. 114.