At the end of many of the preceding sections, and whenever the requirements of the case demanded it, a brief summary of the main facts and of the inferences to be derived from them has been given. It may be useful to give in conclusion a few general remarks on the whole subject.

It will be seen from the numerous facts herein cited, that the so-called monstrous formations (excluding morbid growths the result of disease or injury) present no peculiarities absolutely foreign to the normal organisation of plants. The difference between the natural and monstrous development is one of degree and frequency of occurrence, not of kind.

Deviations from the customary form have been shown to arise from excessive or diminished growth, or from arrested or exalted development. Even in those instances where, for convenience' sake, the term perverted development has been used, it must be understood as applying only to the particular plant or organ under consideration, as the form assumed is perfectly in accordance with the ordinary conformation of some other plant or group of plants.

The period at which malformations occur is a matter of some importance; this is, indeed, implied in the term arrest of development; evolution goes on with growth up to a certain point and is then stopped, and thus changes are brought about in the part affected of a different nature from those dependent on non-development or suppression.

Some malformations are congenital, therefore, while others are acquired—in the former instance the disturbance is coeval in origin, and contemporaneous in its growth and development, with those of the affected part; in the latter case the organ may have attained its ordinary degree of perfection, or at least may have advanced some way towards it, before any deviation shows itself. True chorisis or fission, for instance, is usually a congenital affection, arising at a very early period of development, while enation takes place from structures which are all but complete as to their organisation, even though they may not have attained their full dimensions. The date of appearance is also of consequence in determining the true nature of some changes; it does not always follow, for instance, that because one organ occupies the position of another, it is of the same nature as the one whose place it fills. The presence of anthers on petals or on such organs as the corona of Narcissus does not necessarily constitute those parts actual stamens, but rather staminodes. The true stamens are either wanting, or if present, they are in advance of their imitators as regards their development.

General morphology of the leaf and axis. Homology. Since the time when Goethe's generalisations were adopted by A. P. De Caudolle, special attention has been given to the form and mode of development of the leaf-organ; for as it was well said by Wolff, if once the course of evolution and the structure of the leaf were known, those of the parts of the flower would follow as a matter of course.

It is not necessary, in this place, to pursue the subject of the development and construction of the leaf further than they are illustrated by ordinary teratological phenomena.

From this point of view perhaps the most interesting circumstance is the part that the sheath of the leaf plays.[553] In many cases of so-called metamorphosis, it is the sheath of the leaf that is represented and not the blade. In normal anatomy the sepals, petals, carpels, and even the stamens, as a general rule, correspond to the sheath rather than to the blade of the leaf, as may be seen by the arrangement of the veins. The blade of the leaf seems to be set apart for special respiratory and absorbent offices, while the sheath is in structure, if not in office, more akin to the stem. It would not be easy apart from their position to distinguish between a tubular sheathing leaf and a hollow stem. The development of adventitious growths by chorisis or enation has been frequently alluded to in the foregoing pages, and many illustrations have been given of the power that leaves have of branching in more than one plane, owing to the projection of secondary growing-points from the primary organ. These new centres of development are closely connected with the fibro-vascular system of the leaf, so that no sooner does a new growing point originate, than vessels are formed to connect the new growth with the general fibrous cord, see pp. 355, 445. This leads M. Casimir De Candollo to consider the entire leaf as a composite structure. The morphological unit, says he, is the cellular protrusion or growing point (saillie) and its corresponding fibro-vascular bundle.[554]

The identity, in a morphological point of view, of the leaves and the lateral parts of the flower is so thoroughly recognised that little need be said on that score, save to repeat that the homology of the floral organs is usually not so much with the entire leaf as with its sheath.

The most singular instances of morphological identity are those relating to the sexual organs. We have seen the gradual transition of stamens to pistils, and of pistils to stamens, the development of ovules on the edges of the anther, the co-existence of pollen with ovules on an antheroid body, and, stranger still, the actual development of pollen within the tissues of the ovule itself! From such facts, in addition to what we know of the relative position, internal structure, and mode of development of the organs, it is impossible to avoid coming to the conclusion that, however distinctly these parts may, under ordinary circumstances, be set apart for the performance of distinct functions, morphologically they are homologous.

These ideas may be carried yet farther—the same sort of evidence, which is adduced in support of the morphological identity of leaves with the parts of the flower, may be advanced in confirmation of the opinion, that, morphologically, there is no distinction between axis and leaf. The leaf, according to this view, is a specialised portion of the axis set apart to do certain work, just as the petals, stamens, &c., are leaves told off for distinct uses. It is unnecessary to refer to the intermediate productions linking the leaf-form to that of the axis, all that is requisite here is to point out the facts that teratology lends in support of these views. These may be summed up by the statement that almost all those attributes which morphologists recognise as peculiar to one or the other organ respectively, may be and are manifested by both. We have the stem acquiring the characters of the leaf, and the leaf those of the stem. Thus we have seen leaves, leaf-buds, branches, and flower-buds springing from leaves or leaf-organs;[555] see pp. 174, 177, 445, &c. The structure that we are apt to associate exclusively with one is found to pertain to the other. The arrangement of the vascular cords in the leaf-organ finds its counterpart in the axis, generally, it is true, modified to suit altered circumstances or diverse purposes. In some cases the disposition is absolutely indistinguishable in the two organs. It may then be said that the distinctions usually drawn between axis and leaf are not absolute, and that, however necessary such a separation may be for descriptive or physiological purposes, morphologically the two organs are identical. Again, it may be said that leaf and axis are two phases of the same organ,—an organ capable of existing in its undifferentiated state in the form of a thallus among Cryptogams, but which in the higher groups of plants becomes marked out into separate portions, each portion having its own distinct functions to fulfil for the common benefit of the whole organisation.[556]

Special morphology.—Under this heading brief reference may be made to some of the organs whose morphological nature has been, and still is, much contested. It is clear that for the due elucidation of these matters, development and the comparative investigation of similar structures in different plants must be studied. Teratological data by themselves can no more be trusted to give a correct solution of any particular question, than the evidence furnished by other departments of botanical science taken separately. With this statement by way of caution, allusion may be made to some of the organs whose morphological construction is illustrated by the facts recorded in the present volume.

Calyx-tube.—In descriptive botany it is the common practice to speak of a calyx-tube, by which is meant a tubular or sheathing portion at the base of the flower, below the sepals or calyx-lobes, and distinct or inseparable from the ovary. The question morphology has to solve is whether this tubular structure is to be considered as a portion of the axis, or whether it is to be regarded as composed of the confluent bases of the sepals.

Mr. Bentham, who has recently reviewed the evidence as to the nature of the calyx-tube in his paper on MyrtaceÆ,[557] still holds to the notion that the "calyx-tube" or "hypanthium" is formed from the concretion of the basal portions of the sepals. He founds his conclusions upon such facts as the following: the circumstance that the point of origin of the leaf is not always the same as the point of disarticulation or separation from the axis, inasmuch as the basal portion of the leaf is often adherent to the stem for some distance, though still recognisable as foliar not axial in its nature. In the same manner, the corolla and androecium may be concrete at the base, so that the stamens are for convenience' sake described as inserted into the tube of the corolla, though it is generally admitted that both stamens and petals are really hypogynous, and it is not usual to consider the corolla-tube up to the divergence of the stamens as part of the receptacle. A similar remark applies to the carpels and placentas. Mr. Bentham further considers that the gradual disconnection of the various whorls, that may be traced in many plants, is a further proof of concretion, rather than of expansion of the axis, but this argument may fairly be met by the consideration that the several whorls emerge at different heights.[558]

Organs originally free and distinct become ultimately combined at the base by the gradual protrusion from the receptacle of a ring or tube under them, as in the stamens of LeguminosÆ; yet, says Mr. Bentham, no one would propose to describe the staminal tube of monadelphous LeguminosÆ as part of the receptacle and not of the stamens. Perhaps not, for descriptive purposes, but morphologically it would not be easy to separate such a tube from the receptacle. The principal kinds of malformation which have a bearing on this subject are mentioned at pp. 77–81 and 247, from which it may be seen that the evidence furnished by teratology is conflicting. It would seem, indeed, that while in some families of plants there may be a real calyx-tube, in others the tubular portion is a sheath-like prolongation of the axis. In Primula or Pedicularis, where the venation is clearly laminar, the tubular portion is distinctly calycine. In other cases the so-called calyx-tube seems as certainly to be an expansion of the receptacle, as in RosaceÆ, MyrtaceÆ, MelastomaceÆ, Passiflora,[559] &c.

Where the petals and stamens are described as being inserted into the throat of the calyx, or are perigynous, it may be assumed as a general rule, subject to but few exceptions, that the so-called calyx-tube is really a portion of the receptacle.[560] After all, this is very much a question of words, and for the following reasons,—very often the base of the calyx does evidently form a tube, and no one can say where the calyx ends and the receptacle begins. Again, many leaves are known to originate in the form of a ring-like protrusion from the axis, and from this primary ring originate secondary developments. Thus the asserted difference between a leaf, with such a history of development, and an axial structure becomes obliterated. From this point of view, peltate leaves like those of TropÆolum or Nelumbium become very significant. In both the leaf-stalk is cylindrical and traversed, as in the case of all cylindrical leaf-stalks, by a circle of fibro-vascular cords, as in a branch, and which radiate in all directions in the blade of the leaf. Now, if (as often happens to a slight extent) the central portion of the leaf were much depressed, owing to the disproportionate growth of the peripheral, as contrasted with the central portions, we should have a funnel-like or tubular formation, precisely similar to many of the so-called calyx-tubes. And, if we further suppose new growths to originate from the sides of this funnel or tube, by chorisis or enation, we should have the homologue of a tubular calyx, to the inner surface of which are attached petals, stamens, &c. From the consideration of circumstances such as these just detailed, together with that of the arrangement of the vascular cords, M. Casimir De Candolle arrives at the conclusion that the calyx-tube is a ring-like projection from an axis whose further direct development is arrested. The secondary projections or growing-points correspond to the several fibro-vascular cords of the primary ring, and are ultimately developed into sepals, petals, stamens and ovaries (see pp. 394, 509).

Androecium.—The main points of morphological interest relating to the androecium, referred to in this volume, are those concerning the structure of the anther (see p. 292), the compound nature of the stamens in some orders (see pp. 294, 345), and the nature of the androecium in orchids (see p. 380).

Inferior ovary.—Is the pistil always foliar in its morphological nature, or is it, in some cases, as Schleiden taught, formed from the axis alone? To a great extent the reply to this question is dependent on the conclusions that may be arrived at as to the true nature of the calyx-tube. Considered from a teratological point of view, there is no reason for considering the inferior ovary to be purely axial. On the contrary, the evidence derived from this source supports the ordinary opinion that the carpels are invaginated within the expanded top of the flower-stalk and more or less adherent to it. Some of the gourds afford good illustrations of this, the upper part of the carpels in these fruits projecting beyond the axial portion. But this matter loses much of its importance if the morphological identity of axis and leaf-organ be conceded. The carpels in inferior ovaries seldom or never correspond to the lamina of the leaf, and between the vaginal portion of the carpellary leaf, and the axis who shall draw the distinction?

Placentation.—Some botanists have considered the placentas to be portions of the carpel, and have compared the production of ovules on them to the formation of buds on the leaf of Bryophyllum. Others have been led to see in each placenta, even when it is, to all outward appearance, a portion of the carpellary leaf, a direct prolongation from the axis, adherent to the leaf. Teratology shows that ovules may be formed indifferently on leaf-organs or on stem-organs. Sutural, parietal, axile, free-central placentation, and, if there be more forms, all may be met with even in the same ovary (see pp. 96, 508). Now, if there were such special tendencies in the axis, as contrasted with the leaf, to produce ovules, it is hardly likely that such anomalous arrangements as those just mentioned would be as frequent as they are. But as leaves produce other leaves, from their edges or their surfaces, and as they form buds in the same situations, just as axial organs do,[561] there is surely little ground for considering the placentas, or ovuliferous portions of the plant, to be of necessity axial. Here again, much of the difficulty vanishes if the morphological identity of the leaf-form and of the stem-form be admitted.

Structure of the ovule.—The nature of the ovule and of its coverings has been a fertile source of controversy. The teratological data bearing on this subject have been given at pp. 262–272. These data strongly support the notion of the foliar nature of the coatings, and of the axial nature of the nucleus, taking leaf and axis either in the ordinary sense, or as modifications one of the other. It has been shown that the ovular coats may themselves become carpels, and that ovules may be developed upon ovules, p. 268. Whether the intra-carpellary siliques of Cheiranthus, not uncommonly met with (p. 182), are instances of ovular transmutation may be open to doubt.

The axial nature of the nucleus has been inferred from its position, mode of growth, and from its occasionally lengthening into a leafy or even a floriferous shoot. Probably it may occasionally be invested by sheathing coats, more analogous to tubular processes from the receptacle, than to foliar organs, as is the case in Welwitschia. The discussion of this matter, however, pertains rather to normal morphology than to teratology.

Morphology of conifers.—The nature of the pseudo-leaves of Sciadopitys, and probably of other Conifers, is illustrated by teratology, as also is the true constitution of the scale of the cone (see pp. 192, 245, 352), though it must be admitted that little or no light is thrown on that much-contested point—the true nature of the ovule of Gymnosperms.

Relative position of organs.—When organs are considered, not separately, but in their relations to each other, the appearances presented are referable to similar causes. Thus, the separation of parts usually united has been shown to depend on an excess of development, the persistent union of parts, usually separate in the adult state, has been traced to an arrest of the process of development, by no means necessarily coexistent with diminished growth. The diminished or increased number of parts is, in like manner, attributable to analogous causes, as also are the variations in arrangement and form, spoken of under the heads of Displacement, Peloria, Substitution, &c.

In the instance of displacements, it has been shown how slight a change is required to transform the so-called inferior ovary into a superior one. A defective development of the top of the flower-stalk in some cases, in others a lack of union between the tube of the receptacle or of the calyx (comprising in those terms not only the apex of the receptacle, but the base of the sepals) and the carpels, suffice to bring about this change in a character which for systematic purposes is of great value.

Law of alternation.—The circumstances that interfere with the law of alternation may be briefly alluded to. The deviations from the customary arrangement have been very generally attributed to suppression, or to chorisis. It is unquestionable that either of these affords an efficient explanation of the arrangement in question, as also does that modification of chorisis, as it may be considered, which has been treated of under the head of Enation. Spiral torsion of the axis would likewise bring about analogous results. Still, it is quite conceivable that opposition or superposition of organs may occur without the intervention of any such operations. This will be the more readily conceded when it is remembered that the phyllotaxis of leaves not unfrequently varies on different branches of the same individual tree, and that a similar variation in the flower would at once disturb the customary alternate arrangement. Coalescence of the vascular bundles in an unusual manner, and an irregular disposition of these cords have also been considered to bring about deviations from the rule of alternation, but in general the formation of the cords is subsequent to that of the growing points or mamelons.

Adhesions, accompanied by displacements, occasionally produce similar deviations, the nature of which is usually easily detected.

Co-relation.—The importance of this subject first prominently brought into notice by Geoffroy St. Hilaire gains in force daily. Rarely is a malformation an isolated phenomenon, almost always it is associated, from the operations of cause or effect, with some others. Instances of this co-relation have been cited in the preceding pages, and many more might have been mentioned, had the consideration of the relationship between form and function formed part of the plan of this volume. A change in itself slight, often acquires importance from its association with other alterations. This is particularly well seen in the case of the receptacle. Let an ordinarily concave thalamus remain, from defective development, flat, and how great the change in the appearance of the flower. Let the usually contracted receptacle be lengthened, and the whole aspect of the flowers so affected is altered to such an extent that, were their history not known, botanists would have no hesitation in assigning them to widely separate groups in their schemes of classification. Peloria, too, of either form, affords excellent illustrations of the co-existence of one changed condition with another. Not only is the form of one set of organs altered, but the number, the relative proportion, and the direction of the other organs of the flower are altered likewise.[562] Not only is the whole symmetry changed, but the physiological operations carried on in the flower undergo corresponding alterations.

There are certain co-relations which do not appear to have hitherto attracted the attention they merit; such, for instance, is that which exists between the particular period at which an organ is developed and its position and form. In normal morphology this has, to some extent, been worked out, as in the case of definite and indefinite, centrifugal and centripetal inflorescences, and in the definite or indefinite formation of shoots, &c.

Other instances may be cited in the frequent co-existence of regular flowers and definite inflorescence, the terminal position of many peloriated flowers, the relationship between indefinite inflorescence and prolongation of the axis, &c.

Again, the simultaneous evolution of the parts of the flower and their consequent verticillate arrangement, are often associated with the production of different forms from those characteristic of organs developed in succession, and, in consequence, arranged spirally. In the case of simultaneous development we meet with a repetition of whorls, as in what are termed hose-in-hose flowers (flores duplicati, triplicati, &c.), and also with cases of peloria. In instances where the organs are formed successively in spiral order, we meet with such changes as median prolification, petalody, and phyllody. All these are alterations which we might anticipate from the activity of the growing point being checked at a certain stage in the one case, while it is continuous in the other. This relationship between the definite and indefinite modes of growth and the form of the several organs of the flower, is more constant in reality than it may appear to be from a perusal of the lists of genera in the foregoing pages, in which it was not possible to show sufficiently well the comparative frequency of any given changes in individual plants. Had it been possible to give statistics setting forth the frequency of certain deviations in plants or groups having a particular organisation, as compared with the rarity of their occurrence in other plants of a different conformation, these co-relationships would have been rendered much more evident. A hundred different plants, for instance, may be named in any particular list, of which fifty shall be of one type of structure, and the remainder of another. And the co-relative changes in each fifty may appear to be evenly balanced, but so far is this from being the case, that the frequency of the occurrence of a particular change, in one species in the list, may be so great as far to exceed the instances of its manifestation in all the rest put together. This difficulty is only very partially obviated by the addition of the * to signify especial frequency of occurrence of any given malformation in the plants to whose names it is affixed.

Compensation.—But little further need be said on this head. An atrophied condition of one part is generally associated with an hypertrophied condition of another, and scarcely a change takes place in one direction, but it is associated with an inverse alteration in some other. This principle is not universal, and its application must not be unduly strained. It requires specially to be considered in reference to differences in the degree or kind of functional activity exercised by the organs implicated—points beyond the scope of the present volume.

Teratology and classification.—Lastly, there remain to be mentioned the bearings of teratology on systematic botany. There are those who would entirely exclude teratology from such matters. It may be expedient to do so when the object sought is one of convenience and facility of determination only, but when broader considerations are concerned, teratology must no more be banished than variation. In most instances the one differs but in degree from the other. If variation affords aid in our speculations as to the affinities and genealogical descent of species and other groups, so does teratology, and in a far higher degree.

Take the characters of exogens as distinct from endogens; even under ordinary circumstances, no absolute distinction can be drawn between them. There are plants normally of an intermediate character, while, to take exceptional instances, there are exogens with the leaves and flowers of endogens, and endogens whose outward organisation, at any rate, assimilates them to exogens. Diclinous or monochlamydeous plants owe their imperfect conformation to suppression, and may become structurally complete by a species of peloria. Structurally hermaphrodite flowers become unisexual by suppression, or are rendered incomplete by the non-development of one or more of their floral whorls. Hypogynous flowers become perigynous by adhesion, or by lack of separation; perigynous ones become hypogynous by an early detachment from the receptacle that bears them, or by the arrested development of an ordinarily cup-like receptacle.

How the relative position of the carpels and the calyx may be altered has already been alluded to, as has also the circumstance that while it is common to find an habitually inferior or adherent ovary becoming superior or free, it is much more rare to find the superior ovary adherent to the receptacle or to the calyx.[563] Regular and irregular peloria, too, serve to show how slight are the boundaries, not only between different genera, but also between different families.

While, therefore, teratology may be an unsafe guide in strictly artificial schemes, it is obvious that its teachings should have great weight in all philosophical systems of classification.

The questions will constantly arise, does such and such a form represent the ancestral condition of certain plants? Is it a reversion to that form? or is it, on the other hand, the starting point of new forms?

Such questions cannot receive at present any satisfactory answer, but the evidence we have seems to indicate that pre-existing forms were simpler, and less specialised in structure than those now existing, and hence if we meet with malformations of a simple kind, we may consider them as possible reversions; while, if they present features of increased complexity, and more sharply defined differentiation, we may assume them to be evidences of a progressive rather than of a retrogressive tendency.

That monstrosities so called may become the starting points of new forms is proved by circumstance that, in many cases, the peculiarities are inherited so that a new "race" is produced and perpetuated: and if a new race, why not a new species? The difference is one of degree only.