FOOTNOTES:

[379] 'Phil Bot.,' § 274.

[380] 'El. Ter. Veget.,' p. 392.

[381] 'Bull. Acad. Belg.,' xvii, part ii, p. 38.

[382] 'Bull. Soc. Bot. Fr.,' vol. iv, 1857. p. 760.

[383] Ibid., vol. viii, 1861, p. 159.

[384] See 'Gard. Chron.,' July, 1866, p. 656, and Clusius, 'Plant. Rar.,' lib. 2, p. 143, Tulipa serotina πολυκλαδης, minor, &c. Hort. Eysttett. Plant. Vern.,' fol. 12.

[385] 'Jaeger de Pini sylvestris monstrositate,' Stuttgardt, 1828.

[386] Cited in 'Bull. Soc. Bot. Fr.,' xiv, p. 265.

[387] Duval Jouve, 'Hist. Nat. Equiset. Fr.,' tab. 8, also Milde, 'Nov. Act. Acad. Nat. Cur.,' t. xxvi, part 2. For branched inflorescence of orchids, see 'Reichenbach Proc. Lond. Bot. Congress,' 1866, p. 121.

[388] 'Bull. Soc. Bot. Fr.,' vi, 266, vii, 457. Irmisch, 'Knollen und Zwiebelgew.,' tab. 7, figs. 10, 11.

[389] 'Flora,' 1831, p. 5, tab. i; see also Hanstein, 'Flora,' 1857, p. 513. Schlechtendal, 'Bot. Zeit.,' xviii, p. 381.

[390] 'Bull. Soc. Bot. Fr.,' ix, p. 8.

[391] It will be seen, from what has been just said, that in some of the cases where the axile organs, branches, &c., appear to be multiplied, the increased number is due to subdivision rather than to renewed formation (see Fission). Of this last description is an instance which came under the writer's notice after the section relating to that subject was in print, and which may therefore here be alluded to. The instance is that of the subdivision of the leaf-like organs of Sciadopitys verticillata. In one instance the pseudo leaf divided, and from the division proceeded a little axis, bearing at its summit a verticil of pseudo leaves. This division and formation of new axes and verticils affords ample confirmation of the opinion thrown out by Professor Alexander Dickson, that the apparent leaves of this plant were really branches: see 'Revue Horticole,' 1867, and 'Report. Bot. Congress,' London, 1866, p. 124.

CHAPTER II.
MULTIPLICATION OF FOLIAR ORGANS.

The cases referrible to this head may be ranged under two sections according as the increase is due to plurality of ordinarily single organs, or to an increase in the number of verticils or whorls.

When, in place of a single leaf organ two or more are really or in appearance present the occurrence may be due to one of several causes; among them may be mentioned an actual formation of parts in unwonted number, hypertrophy or enation, chorisis or fission, disjunction, adhesion of one leaf to another or to the stem, as in some of the leaves called "geminate," wherein the two leaves, though apparently in juxtaposition, yet originate from different parts of the stem, but by coalescence or lack of separation produce the impression as if they sprang from the same node. In the adult state it is not always possible to ascertain with certainty to which of these causes the increase in the number of leaves is due, though a clue to the real state of things may be gained from attention to the distribution of the veins, to the arrangement or phyllotaxy of the leaves, the size and position of the supernumerary organs, &c.

The term "phyllomania," as ordinarily used, is applied to an unwonted development of leafy tissue, as in some begonias where the scales or ramenta are replaced by small leaflets, or as in some cabbage leaves, from the surface of which project, at right angles to the primary plane, other secondary leafy plates; but these are, strictly speaking, cases of hypertrophy (see Hypertrophy).

Those instances in which the actual number of leaves is increased, so that in place of one there are more leaflets, may be included under the term "pleiophylly," which may serve to designate both the appearance of two or more leaves in the place usually occupied by a single one, and also those normally compound leaves in which the number of leaflets is greater than usual.

The increased number of leaves in a whorl may well be designated as "polyphylly," using the word in the same sense as in ordinary descriptive botany, while "pleiotaxy" may be applied to those cases in which the number of whorls is increased.

Fig. 183.—Supernumerary leaflet, Ulmus campestris.

Pleiophylly.—As above stated, this term is proposed to designate those cases in which there is an absolute increase in the number of leaves starting from one particular point, as well as those in which the number of leaflets in a compound leaf is preternaturally increased. The simplest cases are such as are figured in the adjacent cuts, wherein, in place of a single leaf, two are produced in the elm. In the one case the new leaflet springs from the apex of the petiole and partially fills the space consequent on the obliquity of the base of the leaf. In the other it would seem as if two distinct leaves emerged from the stem in juxtaposition. This is probably due to a lateral chorisis or subdivision of the primitive tubercle or growing point, followed by a like subdivision of the vascular bundle supplying it. There are certain varieties of elm that very generally present this anomaly on their rank, coarse, growing shoots. In these cases the new growths have the same direction as the primary one, but in other cases the supplementary production is exactly reversed in direction. Thus, in the common hazel (Corylus) a second smaller leaf proceeding from the end of the leaf-stalk at the base of the primary one may frequently be seen. M. Germain de Saint Pierre records an instance in a mulberry leaf, from the base of which proceeded a large leafy expansion divided into two tubular, horn-like projections, and in the centre a thread-like process representing the midrib and terminated by a small two-lipped limb.[392] Dr. Ferdinand Müller speaks of a leaf of Pomaderris elliptica as bearing a secondary leaf on its under surface.[393]

Fig. 184.—Supernumerary leaf, Ulmus montana.

Fig. 185.—Supernumerary leaf of hazel.

The leaves of Heterocentron macrodon have likewise been observed occasionally to produce leaflets from their upper surface.

To this production of leaves from leaves the late Professor Morren applied the term "autophyllogeny."[394] The Belgian botanist figures a small perfect leaf springing from the nerves of the upper surface of the primary leaf in a species of Miconia. As in the hazel, the direction of the adventitious leaf is inversely that of the primary one, the upper surface of the supernumerary leaflet being turned towards the corresponding surface of the normal leaf. A similar occurrence took place in Gesnera zebrina, but the new growth in this case sprang from the lower face of the leaf. Morren explains the appearances in question by supposing that the supplementary leaf is one of a pair belonging to a bud borne on a slender stalk. This stalk and one of the bud-leaves are supposed to be inseparably united with the primary leaf. But there is no reason at all for supposing the existence of adhesion in these cases; no trace of any such union is to be seen. A much more natural explanation is that, from some cause or another, development at the apex of the petiole or on the surface of the nerves, instead of taking place in one plane only, as usual, takes place in more than one, thus showing the close relationship, if not the intrinsic identity, between the leaf-stalk and its continuation, the midrib, with the branch and its subdivisions. The form of the leaf-stalk and the arrangement of the vascular bundles in a circle in the case of the hazel, before alluded to, bear out this notion. Such cases are significant in reference to the notion propounded by M. Casimir de Candolle, that the leaf is the equivalent of a branch in which the upper portion of the vascular circle is abortive.[395]

Compound leaves, as has been stated, occasionally produce an extra number of leaflets; one of the most familiar illustrations of this is in the case of the four-leaved shamrock (Trifolium repens), which was gathered at night-time during the full moon by sorceresses, who mixed it with vervain and other ingredients, while young girls in search of a token of perfect happiness made quest of the plant by day. Linné, who in this matter, at any rate, had less than his usual feeling for romance, says of the four-leaved trefoil that it differs no more from the ordinary trefoil than a man with six fingers differs from one provided with the ordinary number. It should be stated that five and six adventitious leaflets are found almost as frequently as four.

Walpers describes a case where the leaf of T. repens bore seven leaflets. Schlechtendal alludes to a similar increase in number in Cytisus Laburnum, and many other instances might be cited.

For figures or descriptions of four-leaved shamrocks the reader is referred to Lobel, 'Stirp. Advers.,' Nov., p. 382. Tabernæmontanus 'Krauterbuch,' S. 222. Schlechtendal, 'Bot. Zeit.,' ix, p. 583, xiv, p. 71. Maugin, 'Bull. Soc Bot. Fr.,' 1866, t. xiii, p. 279. See also Cramer, 'Bildungsabweich,' p. 92. Walpers, 'Linnæa,' 1840, p. 362 (7-leaved). Schlechtendal, 'Bot. Zeit.,' 1844, p. 457, Cytisus. Wigand, 'Flora,' 1856, p. 706.

Frondiferous leaves have much the appearance of branches provided with leaves, and they may be compared with those instances in which an adventitious bud is placed on the surface or edges of the leaves, as in Gesnera, Cardamine, &c. In truth, the two conditions merge one into the other, as in some begonias, where the ramenta often become leaf-like and bear small bulbils in the axil.

When frondiferous leaves die the appendages die also, but when a true bud has been formed on a leaf it does not of necessity die with the leaf that bears it, but separates from it and continues to grow independently.

Increased number of stipules, spathes, &c.—Seringe relates the occasional presence of two or three additional stipules upon the leaf-stalks of Salix fragilis, and even makes a variety (Salix pendula, var. multistipulata).

An increase in the number of the spathes has been often noticed in Arads[396]. Prof. Alex. Braun has studied this subject in some detail[397]. In Calla palustris the shoot which continues the growth of the plant proceeds from the axil of the last leaf but one; the very last leaf producing no bud, but if accidentally a shoot is developed in this latter situation it produces flowers at once. No leaves are formed, but, on the contrary, two or three spathes surround the spadix, so that the presence of an increased number of spathes in this plant is associated with the development of a side shoot from the axil of the last leaf, the situation whence, under natural circumstances, no shoot at all issues. The supernumerary spathes are not always on the same level, but may be separated by a considerable interval. They vary very much in size, and sometimes assume the form and appearance of leaves. Similar anomalies occur in other Arads as Arum maculatum, Richardia æthiopica, and Anthurium Scherzerianum, frequently combined with a leaf-like appearance of the spathes and sometimes with a subdivision of the spadix into two or three branches.

Engelmann relates the occurrence of an increased number of glumes in Bromus velutinus associated with suppression of the flowers.

Polyphylly.—As previously explained, this term is here applied to those cases in which the members of any particular whorl are increased in number, the whorls themselves not necessarily being augmented.

The simplest cases of this kind are those in which we meet with an unusual number of leaves in a whorl.

Increased number of leaves in a whorl.—This may arise from actual multiplication, or from lateral chorisis, or fission. The true nature of the case may usually be ascertained by an examination of the distribution of the veins of the leaves, or of the fibrous cords of the stem, by the relative position of the supernumerary organs, &c.

Among plants with normally opposite leaves the following occasionally produce them in whorls of three:—Lonicera brachypoda, L. Xylosteum, Weigela rosea, Cornus mas, Vinca minor, &c.

Paris quadrifolia may frequently be met with five leaves in its whorl, or even six.[398]

Increased number of bracts.—This is not of infrequent occurrence; one of the most curious instances is that recorded by Mr. Edwards[399] in Cerastium glomeratum, where, in place of the usual pair of bracts at the base of the head of flowers, there was a whorl of six or eight, forming an involucre. The flowers in this case were apetalous and imperfect.

Polyphylly of the calyx.—This may occur without any other perceptible change, while at other times the number of the other parts of the flower is proportionately increased. In a flower of a plum six sepals in place of five sometimes exist; a precisely similar occurrence in the flowers of the elder (Sambucus), the Fuchsia, and of Œnanthe crocata, may occasionally be met with. In the latter case, indeed, there are sometimes as many as ten segments to the calyx, and this without the other parts of the flower being correspondingly augmented. Among monocotyledons a similar increase is not uncommon, as in Tulipa, Allium, Iris, Narcissus, &c.

In some plants there seems to exist normally much variation in the number of parts; thus in some species of Lacistema in adjacent flowers the calyx may be found with four, five, or six segments.

Most of these cases of polyphylly affecting the calyx may be explained by lateral chorisis or fission.

Polyphylly of the corolla.—This may happen in connection with similar alterations in the calyx and stamens, or sometimes as an isolated occurrence. In the latter case it may be due to lateral chorisis, to substitution, or to the development of organs usually suppressed; thus, when in aconites we meet with four or five horn-like nectaries (petals) instead of two only, as usual, the supernumerary ones are accounted for by the inordinate development of parts which ordinarily are in an abortive or rudimentary state only. This is borne out by what happens in Balsamineæ. In the common garden balsam the fifth petal is occasionally present, while in Hydrocera triflora this petal is always present.

In a flower of a Cyclamen recently examined there were ten petals in one series, the additional five being evidently due to the subdivision of the five primary ones; the natural circular plan of the flower was here replaced by an elliptical one. A similar occurrence takes place in the flowers of maples (Acer), which sometimes show an increased number of parts in their floral whorls and an elliptical outline. Whether the additional organs in this last case are the result of complete lateral chorisis or of multiplication proper I do not know.

Orchids are very subject to an increase in the number of their labella. As illustrations may be cited an instance recorded by Mr. J. T. Moggridge in a flower of Ophrys insectifera, and in which there were two labella without any other visible deviation from the ordinary conformation.[400]

I am indebted to Mr. Hemsley for the communication of a similar specimen in O. apifera, in which there were two divergent lips, each with the same peculiar markings. One of the sepals in this flower was adherent to one of the lateral petals. This augmentation of the labella depends sometimes on the separation, one from the other, of the elements of which the lip is composed, at other times on the development, in the guise of lips, of stamens which are usually suppressed ([see p. 380]).

The following enumeration will suffice to show the genera in which an increased number of petals or perianth-segments in any given whorl most frequently occurs.

For other illustrations see multiplication of whorls, petalody; see also Moquin, loc. cit., p. 350. Engelmann, loc. cit., p. 20, § 18. Cramer, loc. cit., p. 25.

Polyphylly of the andrœcium.—An increased number of stamens frequently accompanies the corresponding alterations in other whorls, and seems, if anything, to be more frequent among monocotyledonous plants than among dicotyledonous ones; thus, we occasionally find tetramerous flowers in Crocus, Hyacinthus, Tulipa, Iris, Tigridia, &c., and more rarely in Yucca (Y. flexilis[401]).

The increased number of stamens in a single whorl may result from a development of organs usually suppressed, and constitute a form of regular peloria as in Linaria, wherein a fifth stamen is occasionally met with. Among normally didynamous plants such numerical restitution, so to speak, is not unusual; thus, in Veronica four and five stamens occur. Fresenius has seen five stamens in Lamium, Mentha, Chelone;[402] Bentham in Melittis, and other instances are cited under the head of peloria. Chorisis may also serve to account for some of these cases; thus, Eichler[403] figures a flower of Matthiola annua with five long stamens instead of four; one of the long pairs of stamens has here undergone a greater degree of repetition than usual. De Candolle[404] cites and figures a curious form of Capsella Bursa-pastoris sent him by Jacquin, and which was to some extent reproduced by seed. In the flowers of this variety there were no petals, but ten stamens; hence De Candolle inferred that the petals were here replaced by stamens, but Moquin[405] objects, and with justice, to this view, as the ten stamens are all on the same line; he considers the additional stamens to be the result of chorisis. Buchenau[406] mentions the presence of seven stamens in another Crucifer, Ionopsidium acaule. Here the supernumerary organ was placed between two of the long stamens. The effect of chorisis in producing an augmentation of parts is well seen in some plants that have some of their flowers provided with staminodes or abortive stamens, and others with clusters or phalanges of perfect stamens. Thus, in the female flowers of Liquidambar there are five small staminodes without anthers, whereas in the male flower the stamens are numerous and grouped together in phalanges, so that the relation of simple to compound stamens is in this case readily seen, as also in many Malvaceæ, Sterculiaceæ, Byttneriaceæ, Tiliaceæ, and Myrtaceæ. It is probably the idea of splitting or dilamination involved in the word chorisis that has led many English botanists to hesitate about accepting the notion. Had they looked upon the process as identical with that by which a branched inflorescence replaces an unbranched one, or a compound leaf takes the place of a simple one, the objections would not have been raised with such force. The process consists, in most cases, not so much in actual cleavage of a pre-existing organ as in the development of new-growing points from the old ones.

An illustration given by Moquin from Dunal[407] goes far to support the notion here adopted. The majority of the stamens of laurels (Laurus) have, says M. Dunal, on each side of the base of their filaments a small glandular bifid appendage; these excrescences are liable to be changed into small stamens. The male flowers have a four-leaved calyx, and sometimes eight stamens, each with two glands, four in one row, opposite to the sepals, four in a second series alternating with the first. More generally two of the stamens are destitute of glands, but have in their place a perfectly developed stamen, so that in these latter flowers there are twelve stamens.

M. Clos[408] mentions a flower of rue (Ruta) wherein there were two stamens joined together below and placed in front of a petal, as in Peganum.

Buchenau[409] mentions a flower of Lotus uliginosus in which there were eleven stamens, namely, two free and nine monadelphous; and Hildebrand describes an analogous increase in a flower of Sarothamnus scoparius in which, in conjunction with a seven-toothed calyx, there were two carinas and fourteen stamens. It would seem probable in this case that there was a coalescence of two flowers at an early date and consequent suppression of some of the parts of the flower. Whether this was the case or not in this particular illustration, it is nevertheless certain that many of the recorded instances of increased number in the organs of a flower are really the results of a fusion of two or more flowers, though frequently in the adult state but few traces of the coalescence are to be seen.

Polyphylly of the gynœcium.—Moquin[410] remarks that, as the pistils are, generally speaking, more or less subject to pressure, owing to their central position, and it may be added owing to their later development, than the other parts of the flower, they are more subject to suppression than to multiplication; nevertheless, augmentation in the number of carpels does occasionally take place, especially when the other parts of the flower are also augmented in number. Sometimes this increase in the number of carpels is due to pure multiplication, without any other change. At other times the increase is due to a substitution of stamens or other organs for carpels (see Substitutions). In other cases the augmentation seems to be due to the development of parts usually suppressed; for instance, in Antirrhinum, where there are usually only two carpels present, but where, under peculiar circumstances, five may be found—thus rendering the symmetry complete.[411] In Papilionaceæ, wherein usually only one carpel is developed, we occasionally find two, or even more, as in Wistaria, Gleditschia, Trifolium, &c. In Prunus and Amygdalus from two to five carpels are occasionally to be found,[412] in Mimosa five, in Umbelliferæ three to five; in some composites, e.g. Spilanthes, five carpels have also been noticed; in Cruciferæ three and four, in grasses three.[413] The double cocoa-nut affords an illustration of the development of two carpels out of three, one only generally arriving at perfection. Triple nuts (Corylus) also owe their peculiarity to the equal development of all three carpels which exist in the original flower, but of which, under ordinary circumstances, two become abortive. It is necessary, however, to distinguish these cases from those in which two embryos are developed in one seed.

The following list may serve to show in what genera this change has been most frequently noticed, and it may be said in general terms that Cruciferæ, Umbelliferæ, and Liliaceæ, are the orders most frequently affected. Cases of peloria are not included in the subjoined list.

A few additional references may here be given to papers where an increased number of carpels is described:—Engelmann, 'De Antholys,' § 17, p. 19. Bernhardi, 'Flora,' 1838, p. 129. Schkuhr., 'Bot. Handb.,' t. 179. Godron, 'Ann. Sc. Nat.,' ser. 5, vol. ii, p. 280, tab. xviii, pluricarpellary Crucifers. Weber, 'Verhandl. Nat. Hist. Vereins. Rhein. Pruss.,' &c., 1860, Cerasus, &c., &c. Baillon, 'Adansonia,' iv, p. 71, Trifolium. Schlechtendal, 'Bot. Zeit.,' xv, p. 67, Datura, three-celled fruit; 'Bot. Zeit.,' xiii, p. 823, Phaseolus, double pistil—a common case. Cramer, 'Bildungsabweich,' p. 99, reference to several leguminous plants with polycarpellary pistils. Munro, Gen., 'Linn. Trans.,' vol. xxvi, p. 26, Bambuseæ. Alph. de Candolle, 'Neue Denkschrift,' Cheiranthus. Schimper, 'Flora,' 1829, ii, p. 433. Wigand, 'Bot. Untersuch.' Fleischer, 'Missbild. Cultur Pfl.' Cramer, 'Bildungsabweich,' p. 65, Umbelliferæ.

Polyphylly of the flower in general.—Although, for the sake of convenience, multiplication has here been treated of as it affects the members of individual whorls of the flower, yet it must be remembered that, in general, the augmentation is not confined to one whorl, but affects several; thus, if the sepals are increased, the petals are likely to be so likewise, and so forth. One of the most curious illustrations of this is that recorded by Mr. Berkeley[414] in a plum, wherein there was an increased number of sepals, a corresponding augmentation in the petals, while the pistil was composed of two and sometimes three carpels distinct from the calyx and from each other. In the flowers there did not appear to be any definite relation in the position of the parts either with reference one to another or to the axis.

Fig. 186.—Plum. Increased number of parts in the calycine, corolline, and carpellary whorls respectively.

In Primulaceæ this general augmentation has been frequently noticed.[415]

Among Orchideæ the instance related by Dr. Seubert is worth alluding to here. This botanist observed and figured a flower of Orchis palustris with tetramerous arrangement of parts, that is to say there were four outer segments to the perianth, four petals, of which two were lip-like, four stamens, three of which were rudimentary, and an ovary with four parietal placentæ.[416]

The following list will serve to show in what plants this general augmentation of parts has been observed most frequently:

Increased number of ovules or seeds.—This appears not to be of very frequent occurrence, at least in those plants where the number of these organs is normally small; where, as in Primula, the ovules and seeds are produced in large quantities, it is not practicable to ascertain whether the number be augmented or not in any particular case. Very probably, the attachment or source of origin of the ovules determines, in some measure, their number. Thus, in the case of marginal placentation the number must be limited by the narrow space from which they proceed, whereas in parietal and free central placentation the ovules are generally numerous. In the latter case, however, it will be remembered that solitary ovules are not rare. An increased number of ovules is generally remarked in conjunction with some other change, such as a foliaceous condition of the carpel, in which the margins are disunited. In such cases the ovules may occupy the margin or may be placed a short distance within it, as in the case of some open carpels of Ranunculus Ficaria,[417] and in which two ovules were borne in shallow depressions on the upper or inner surface of the open carpel and supplied with vascular cords from the central bundle or midrib. The outer coating of the ovule here contained barred or spiral fusiform vessels derived from the source just indicated.

In the very common cases where the pistil of Trifolium repens becomes foliaceous (see Frondescence), the outer ovules are generally two or more instead of being solitary. So, also, in the Rose with polliniferous ovules ([see p. 274]). Among Umbelliferæ affected with frondescence of the pistil a similar increase in the number of ovules takes place. It will be borne in mind that in most, if not all, these cases the structure of the ovule is itself imperfect.[418]

What are called in popular parlance double almonds or double nuts (Corylus) are cases where two seeds are developed in place of one.

In the 'Revue Horticole,' 1867, p. 382, mention is made of a bush which produces these double nuts each year—in fact, it never produces any single-seeded fruit. The plant was a chance seedling, perhaps itself the offspring of a double-seeded parent. It would be interesting to observe if the character be retained by the original plant, and whether it can be perpetuated by seed or by grafting.

It is necessary to distinguish in the case of the nut between additional seeds or ovules, as just described, and the double, triple, or fourfold nuts that are occasionally met with, and which are the result either of actual multiplication of the carpels or of the continued development of some of the carpels which, under ordinary circumstances cease to grow ([see ante, p. 364]). In the case of a ripe nut with two seeds it might be impossible to tell whether the adventitious seed were the product of multiplication, or whether it belonged, in the first instance, to the same carpel as that producing the fellow-seed, or to a different and now obliterated ovary. In all probability, however, the second seed would be accounted for by the development of two seeds in one carpellary cavity.

There is still another condition occasionally met with in the almond, and which must be discriminated from the more common multiplication of the seed, and which is the multiplication of the embryos within the seed, and which furnishes the subject of the succeeding paragraph.

Increased number of embryos.—A ripe seed usually contains but a single embryo, although in the ovular state preparation is commonly made for more; and, indeed, in certain natural orders plurality of embryos in the same seed does occur, as in Cycadeæ and Coniferæ. In the seeds of the orange (Citrus), in those of some Euphorbiaceæ, &c., there are frequently two or more additional embryos. A similar occurrence has been recorded in the mango, for a specimen of which I am indebted to the Rev. Mr. Parish, of Moulmein.[419]

Plurality of embryos has also been observed in—

See Schauer's translation of Moquin-Tandon, 'El. Terat. Veget.,' p. 245, adnot., and 'Al. Braun Polyembryonie.'

Increased number of the cotyledons.—Although the presence of one or of two cotyledons in the embryo is generally accepted as a valuable means of separating flowering plants into two primary groups, yet, like all other means of discrimination, it occasionally fails, and, indeed, almost always requires to be taken in conjunction with some other character. There are cases among flowering plants where the embryo is homogeneous in its structure, there are others in which the number of the cotyledons is more than two. Thus, in some seeds of Cola acuminata the cotyledons vary in number from two to five. I have not been able to ascertain precisely whether this multiplication of the cotyledons is characteristic of all the seeds of particular trees, or whether some only are thus affected. Some fruits that I examined bore out the latter view, as in the same pod were seeds with two, three, and four cotyledons respectively.

I have also seen three cotyledons present in embryo-plants of Correa, Cratægus Oxyacantha, Dianthus sinensis, Daucus Carota, Cerasus Lauro-cerasus. De Candolle alludes to a case of the kind in the bean, and figures a species of Solanum with three cotyledons.[420] Jaeger alludes to a similar instance in Apium Petroselinum;[421] Ehrenberg to one in the marigold (Calendula);[422] Reinsch to an analogous appearance in the beech (Fagus), associated with a union of the margins of two out of the three cotyledons, and of those of two out of the three leaves next adjacent.[423] This fusion seems frequently to accompany increase in the number of cotyledons. It was so in the Correa, and in the Cratægus previously mentioned. Some of these cases may be accounted for by chorisis or by a cleavage of the original cotyledons, as happens, according to Duchartre,[424] in some Coniferæ, which he considers to be improperly termed polycotyledonous. Whether this holds good in the Loranths, where (Nuytsia, Psittacanthus) an appearance of polycotyledony exists, is not stated. In the case of the rue (Ruta) figured by M. A. de Jussieu[425] this splitting of one cotyledon into two is sufficiently evident, as is also the case in the sycamore (Acer pseudo-platanus), seedlings of which may often be met with divided cotyledons.

In other instances a fusion of two embryo plants may give rise to a similar appearance, as in the Euphorbia and Sinapis found by M. Alph. de Candolle ([see ante, p. 56]).

Pleiotaxy or multiplication of whorls.—In the preceding section notice has been taken of the increased number of parts in a single whorl, but an augmentation of the number of distinct whorls is still more frequently met with. Many of the so-called double flowers owe their peculiarity to this condition. The distinction between the two modes in which the parts of the flower are increased in number has been pointed out by Engelmann, Moquin, and others, and the two seem to require distinctive epithets; hence the application of the terms polyphylly and pleiotaxy, as here proposed.

Pleiotaxy in the bracts.—An increase in the number of bracts has been met with very constantly in a species of Mæsa, and in a peculiar variety of carnation, called the wheat-ear carnation.[426] In some of these cases the increase in the number of bracts is attended by a corresponding suppression in the other parts of the flower. Such a condition has been frequently met with in Gentiana Amarella, where the bracts are increased in number, coloured purple, and destitute of any true floral organs. A similar condition exists in some varieties of Plantago major (var. paniculata), as has been previously stated, p. 109.

Fig. 187.—Wheat-ear carnation. The appearance is due to the multiplication of the bracts and the suppression of the other parts of the flower.

It has been noticed also in the common pea, Pisum sativum, and M. Lortet[427] records a case of the kind in Erica multiflora, the flowers of which, under ordinary circumstances, are arranged in clusters, but in this case the pedicels were more closely crowded than usual, and were covered for their whole length with small rose-coloured bracts arranged in irregular whorls, the upper ones sometimes enclosing imperfect flowers. In the 'Gardeners' Chronicle,' 1865, p. 769, is figured a corresponding instance of Delphinium Consolida, in which the bracts were greatly increased in number, petaloid, and, at the same time, the central organs of the flower were wholly wanting.

Fig. 188.—Delphinium Consolida. Multiplication of bracts at the expense of the other parts of the flower.

Fig. 189.—Multiplication of bracts, &c., Pelargonium.

In flowers of Pelargonium may occasionally be seen a repetition of the whorls of bracts, in conjunction with suppression and diminished size of some of the other portions of the flower (fig. 189).

The common foxglove (Digitalis purpurea) has likewise occasionally been observed subject to a similar malformation.

Cornus mas and C. suecica sometimes show a triple involucre.[428] Irmish[429] records an analogous case in Anemone Hepatica, wherein the involucre was doubled. Similar augmentation occurs in cultivated Anemone. In addition to the plants already mentioned, Engelmann[430] mentions as having produced bracts in unwonted numbers, Lythrum Salicaria, Plantago major, Veronica spicata, Echium vulgare, Melilotus arvensis, and Rubus fruticosus.

It must here be remarked that this great number of the bracts occurs naturally in such plants as Godoya, in which the bracts, or, as some consider them, the segments of the calyx, are very numerous, and arranged in several overlapping segments.

In some of the cultivated double varieties of Nigella the finely divided involucral bracts are repeated over and over again, but on a diminished scale, to the exclusion of all the other parts of the flower.

Pleiotaxy or repetition of the calyx.—The true calyx is very seldom affected in this manner, unless such organs as the epicalyx of mallows, Potentilla, &c., be considered as really parts of the calyx.

In Linaria vulgaris Rœper observed a calyx consisting of a double series, each of five sepals, in conjunction with other changes.[431] It is also common in double columbines, delphiniums, nigellas, &c.

In the 'Revue Horticole,' 1867, p. 71, fig. 9, is described and figured by M. B. Verlot a curious variety of vine grown for years in the Botanic Garden at Grenoble, under the name of the double-flowered vine. The place of the flower is occupied by a large number of successive whorls of sepals disposed in regular order, and without any trace of the other portions of the flower. It is, in fact, more like a leaf-bud than a flower. The outermost whorls of this flower open at the time when the ordinary flowers of vines do; the second series are gradually produced, and expand about the time when the ovaries of the normal flowers begin to swell; a third series then gradually forms, and so on, until frost puts a stop to the growth. This malformation, it appears, is produced annually in certain varieties of vine, and may be perpetuated by cuttings.

The flower of the St. Valèry apple, already alluded to under the head of sepalody, might equally well be placed here. It is not very material whether the second whorl of organs be regarded as a repetition of the calyx or as a row of petals in the guise of sepals.

Engelmann[432] cites the following plants as occasionally presenting a repetition of the calyx, in most cases with a suppression of the other floral whorls:—Stachys lanata, Myosotis palustris, Veronica media, Aquilegia vulgaris, Nigella damascena, Campanula rapunculoides.

Pleiotaxy in the perianth.—Increase in the number of whorls in the perianth is common in lilies, narcissus, hyacinths, &c. It may be also met with occasionally among orchids. The lily of the valley (Convallaria maialis) seems also to be particularly subject to an increase in the number of parts of which its perianth consists, the augmentation being due partly to repetition or pleiotaxy, partly to the substitution of petaloid segments for stamens and pistils.[433]

In this place may also be mentioned the curious deviation from the ordinary structure occasionally met with in Lilium candidum, and known in English gardens as the double white lily. In this case there are no true flowers, but a large number of petal-like segments disposed in an irregular spiral manner at the extremity of the stem, some of the uppermost being occasionally verticillate.[434]

Fig. 190.—Double white lily. Multiplication of perianth-segments and other changes.

Pleiotaxy of the corolla.—With reference to double flowers, it was remarked by Linné that polypetalous flowers were, as he said, multiplied, while monopetalous flowers were duplicated, or triplicated, as the case may be,[435] a statement that is true in the main, though it requires modification. In the case of polypetalous, or rather dialypetalous flowers, the petals may be very largely increased by multiplication, as in roses, anemones, pinks, &c. In the last-named genus the number is often so much increased that the calyx splits from the tension exercised on it by the increasing mass within. This multiplication may happen without any metamorphy or substitution of petals for stamens, though, in the majority of cases, it is associated with such a change. It is curious to observe in some of these flowers that the total number of parts is not greatly increased; thus, in some of the double-flowered Leguminosæ, such as Ulex europæus and Lotus corniculatus, the petals are repeated once or twice, the stamens are petalodic, but reduced in number, while the carpels are usually entirely wanting. Thus, owing to the diminished number of parts in the inner whorls of the flower, these very double-looking blooms do not contain any greatly increased number of parts.[436]

Flowers that, under ordinary circumstances, are gamopetalous, become, in some instances, multiplied by the formation of additional segments, just as in the case of polypetalous corollas; but in these cases the corollas become polypetalous, their petals do not cohere one with another. Among double flowers of this character may be mentioned Campanula rotundifolia, Gardenia sp., Nerium Oleander, Serissa sp., Arbutus Unedo, &c. The change is associated with petalody of the stamens and pistils.

A more frequent change among the monopetalous orders is the duplication or triplication of the corolla, in consequence of which there appear to be a series of corollas enclosed one within the other, the lobes of which generally alternate with one another, but which sometimes are superposed. This happens occasionally in the primrose (Primula acaulis), and constitutes the variety called by the gardeners "hose in hose."

The same condition occurs frequently in some species of Datura and Campanula.

Fig. 191.—Campanula rotundifolia. "Double flowers" resulting from dialysis and multiplication of the petals.

In Antirrhinum majus double flowers of this character sometimes occur; the outermost corolla is normal, the succeeding ones usually have their petals separate one from the other; the stamens are sometimes present, sometimes absent, and at other times petalodic. Similar occurrences may be met with in labiates and jasmines, and in Erica hyemalis.

Mr. W. B. Hemsley has kindly furnished me with flowers of a similar kind occurring in wild specimens of Epacris impressa,[437] and there are analogous phenomena in the common honeysuckle (Lonicera Periclymenum), in which three corollas and no stamens often occur.

This duplication may either be accounted for on the theory of chorisis above alluded to, or by supposing that the extra corolline whorl is due to a series of confluent petalodic stamens; that the latter is the true explanation, in certain cases at least, is shown by some flowers of Datura fastuosa, in which the second corolla was partially staminal in its appearance, and bore nearly perfect anthers, in addition to the five ordinary stamens, which were unaltered either in form or position. Some partially virescent honeysuckle flowers have a similar structure.

There are other cases of apparent multiplication or duplication, due, probably, rather to the formation of outgrowths from the petals than to actual augmentation of their number. These excrescences occur sometimes on the inner surface of the petals, or of the corolla; at other times on the outer surface, as in some gloxinias, &c. This matter will be more fully treated of under the head of hypertrophy and enation.

Pleiotaxy of the andrœcium.—An increase in the number of whorls in the stamens is very common, especially in cases where the number of circles of stamens is naturally large. The augmentation of the number of stamens is still more frequent where these organs are arranged, not in verticils, but in one continuous spiral line.

In Cruciferæ there is always an indication of two whorls of stamens, and this indication is rendered even more apparent in some varieties accidentally met with. So in Saponaria, in Dianthus, and other Caryophylleæ, three and four verticils of stamens have been met with. In Lonicera Periclymenum a second whorl of stamens more or less petalodic sometimes occurs.

Moquin mentions a variety of Rubus fruticosus in which nearly 900 petaloid organs existed in the place of the twenty-five or thirty stamens natural to the plant, the other organs of the flower being in their ordinary condition, with the exception of the pistil, which did not attain its full size. Baillon records the occasional existence of two rows of stamens in Ditaxis lancifolia.

Increased number of stamens in orchids, &c.—Various deviations from the ordinary type of orchid structure have been already alluded to under the head of displacement, fusion, peloria, substitution, &c., but the alterations presented by the andrœcium in this family are so important in reference to what is considered its natural conformation, that it seems desirable, in this place, to enter upon the teratological appearances presented by the andrœcium in this order, in somewhat greater detail than usual. The ordinary structure of the flower with its three sepals, two petals, labellum, column; and inferior ovary, is well known. Such a conformation would be wholly anomalous and inexplicable were it not that the real number and arrangement of parts have been revealed by various workers labouring to the same end in different fields. Thus, Robert Brown, Link, Bauer, Darwin, and others, paid special attention to the minute anatomy and mode of distribution of the vessels; Irmisch, Crueger, Payer, and others, to the evolution of the flower; Lindley, St. Hilaire, and Reichenbach, to the comparison of the completed structures in the various genera and species; while the teratological observers have been numerous, as will be seen from the selected references cited at the end of this paragraph and in other places. The result of this manifold study has been a pretty general agreement that the structure of the order (omitting minor details) is as follows:—A six-parted perianth in two rows, the outer three (sepals) generally regular and equal in shape; of the inner three (petals or tepals) two are regular, and one, the labellum very irregular, consisting not only of a petal, but of two abortive stamens incorporated with it. The column is considered to be made up of one perfect and three abortive stamens, in inseparable connection with three styles. By some, however, it is supposed that all the stamens are confluent with the column and none with the lip.

Fig. 192.—Diagram showing the arrangement of parts in an orchid flower. According to Crüger, the stamens A 2, A 3, should be distinct from the lip. The uppermost figure 2 should have been 1. (See text.)

In either case it is admitted that there are six stamens in two rows. The first row consists of one posterior stamen, which is generally perfect, and two abortive stamens incorporated with the labellum. The second row also consists of three stamens, all of which are usually abortive and inseparable from the column. Traces of them may occasionally be met with in the form of tubercles or wing-like processes from the column. In Cypripedium, while the ordinary stamen of the outer row is deficient, two of the inner series are present. The diagram, fig. 192, will serve to show the arrangement of the parts as above described. + represents the situation of the stem or axis; on the opposite side is the bract; between these are placed the sepals, one posterior or next the axis (incorrectly numbered 2 in the plan), two lateral 1, 1; next in order follow the petals, 2, 2, 2, two lateral and somewhat posterior, one larger (the lip), anterior; the outer series of stamens are represented by A 1, A 2, A 3, the two latter being fused with the labellum; a 1, a 2, a 3 represent the position of the inner verticil of stamens, while s, s, s denote the three carpels. It is foreign to the purpose of this book to detail the varied evidence in support of this explanation of the homologies of orchid flowers.[438] All that can be done in these pages is to set forth the evidence furnished by teratology as to this matter—evidence for the most part accumulated and recorded without any special reference to any theory of orchid structure.

The following details all refer to flowers in which the number of stamens in orchidaceous plants was increased beyond what is necessary. They are arranged with reference to the number of adventitious organs, beginning with those in which the number was smallest, and proceeding thence to those in which it was greatest. In some cases it has not been possible to ascertain whether the adventitious organs were really restorations of the numerical symmetry, substitutions of one part for another, stamen for petal, &c., or wholly adventitious productions. Unless otherwise stated, the interpretation put upon the facts thus recorded is that of the present writer, and not necessarily that of the original observer.

Mr. J. T. Moggridge has described and figured a flower of Ophrys insectifera in which there was a vestige of a second stamen present, probably one of the inner series fig. 192 (a2).[439] The same observer also records the presence of a second anther between the lobes of the normal one. This can hardly be referred to either of the typical stamens, but would seem to be a perverted development of the rostellum.[440]

Rœper is stated by Cramer[441] to have seen a specimen of Orchis morio with two stamens.

In a flower of Habenaria chlorantha, described by the late Professor Henslow,[442] the outer three stamens are suppressed, while two of the inner group are present, as happens normally in Cypripedium.

A flower of Cattleya violacea afforded a similar illustration; but in this case only one of the inner stamens was developed, and this in the form of a small petal, partly adherent to the column.

In Dendrobium normale, Falconer, not only is the perianth regular, but the column is triandrous,[443] the three stamens (according to the diagram of its structure given by Lindley) pertaining to the outer row.

In a specimen of Dendrobium hœmoglossum kindly forwarded from Ceylon by Mr. Thwaites there were three stamens present, of which one posterior belonged to the outer series A 1, and two lateral to the inner a 1, a 2, fig. 192.

M. His observed, several years in succession, some flowers of a species of Ophrys with three sepals, no lateral petals, one lip, and three perfect stamens. In this case probably the two supernumerary stamens were petals which had assumed an anther-like character.

Wydler describes a flower of Ophrys aranifera in which one outer and two inner stamens were present.[444] I have myself met with three such flowers in the same species. The stamens present were A 1, a 1, a 2.

Dr. J. E. Gray exhibited at the Botanical Society of London, in August, 1843, a specimen of Ophrys apifera with a triandrous column, the supernumerary anthers belonging, apparently, to the inner whorl.

In his 'Catalogue of the Plants of South Kent,' p. 56, tab. iv, f. 16, the Rev. G. E. Smith describes and figures a flower of O. aranifera with a triandrous column, seemingly of the same kind as that spoken of by Dr. Gray.

Mr. Moggridge met with a triandrous flower in the same species, and refers the appearance to "a fusion of two flowers, accompanied by suppression and modification."[445] As, however, no details are given in support of this opinion, it may be conjectured that the two additional stamens were members of the inner whorl a 1, a 2, and thus the conformation would be the same as in the flowers just mentioned. The figures given by Mr. Moggridge bear out this latter view, while they lend no support to the hypothesis advanced by him. Nevertheless, no decided opinion can be pronounced by those who have not had the opportunity of examining the flowers in question.

Alphonse de Candolle[446] figures a flower of Maxillaria in exactly the same condition, so far as the stamens are concerned, as in the Ophrys flowers just mentioned. It is curious to observe that in many of these cases the two lateral petals are suppressed.

Von Martius mentions the occurrence of three anthers (naturaliter conformatæ) in Orchis morio.[447] Richard, as cited by Moquin-Tandon, Lindley, and others, describes and figures a peloria of Orchis latifolia with regular triandrous flowers.[448]

The writer has examined, in the Royal Gardens at Kew, a flower of Cattleya crispa in which were three stamens, the central one normal; the two lateral ones, belonging probably to the inner whorl, were in appearance like the lateral petals, and one of them was adherent to the central perfect column. Duchartre[449] mentions a flower of Cattleya Forbesii in which there were two labella in addition to the ordinary one, the column being in its normal condition. From the analogy of other cases it would appear as if the additional labella in this instance were the representatives of two stamens of the outer whorl. Beer likewise has put on record the existence of a triandrous Cattleya.[450]

A specimen of Catasetum eburneum forwarded by Mr. Wilson Saunders was normal so far as the sepals and two lateral petals were concerned, but the anterior petal or labellum was flat and in form quite like the two lateral ones; the column was normal and in the situation of the two anterior stamens of the outer series A 2, A 3, were two labella of the usual form (fig. 156, p. 291). Perhaps the Oncidium represented at p. 68, fig. 29, may also be explained on the supposition that the two lateral lobes of the labellum in this flower were the representatives of stamens.

In Fig. 193 is shown the arrangement of parts in a flower of Ophrys aranifera. Here there were three sepals, two lateral petals, one of which was adherent to the side of the column; the central labellum was seemingly deficient, but there were two pseudo-labella placed laterally in the position of the two antero-lateral stamens of the outer series (A 2, A 3). Within these was another perfect stamen occupying the position of the anterior stamen of the inner series (a 3). In another flower of the same species, gathered at the same time (fig. 194), there were three sepals not at all different from those of the normal flower. The three petals next in succession were also, in form and position, in their ordinary state. In colour, however, the two upper lateral petals differed from what is customary, in having the same purplish-brown tint which characterises the lip. Within these petals, at the upper part of the flower, there was the ordinary column, and at the opposite side, alternating with the petals before mentioned, two additional lip-like petals, one provided with a half-anther containing a single perfectly formed pollen-mass (A 2, A 3). It is, perhaps, worthy of notice that the arrangement of the coloured spots on the true labellum, and that on the adventitious lips, replacing the two lower of the outer stamens, were not of a similar character. The supernumerary lips had the π-shaped marking which is so common in this species, while the true lip was, as to its spots, much more like O. apifera. Alternating with this last whorl were three columns, all apparently perfectly formed and differing only from the ordinary one in their smaller size and corresponding to a 1, a 2, a 3. The ovary in this flower was two-celled, with four parietal placentas, thus giving an appearance as though there had been a fusion of two or more flowers associated with suppression and other changes. The position of the supernumerary organs and the absence of any positive sign of fusion in the bracts or other part of the flower, seemed, however, to negative the idea of fusion.[451]

Fig. 193.—Diagram showing the arrangement of parts in a malformed flower of Ophrys aranifera ([see p. 384]).

Fig. 194.—Malformed flower of Ophrys aranifera with two supernumerary lips and three additional stamens.

A similar illustration, for a knowledge of which the writer is indebted to the kindness of Professor Asa Gray and Mr. Darwin, occurred in some specimens of Pogonia ophioglossoides collected by Dr. J. H. Paine in a bog near Utica, New York. It will be seen from the following description that these flowers presented an almost precisely similar condition to those of the Ophrys aranifera just mentioned. "The peculiarities of these flowers," writes Professor Gray, "are that they have three labella, and that the column is resolved into small petaloid organs. The blossom is normal as to the proper perianth, except that the labellum is unusually papillose, bearded almost to the base. The points of interest are, first, that the two accessory labella are just in the position of the two suppressed stamens of the outer series, viz. of a2 and a3, as represented in the diagram, fig. 192; and there is a small petaloid body on the other side of the flower, answering to the other stamen, a1. Secondly, in one of the blossoms, and less distinctly in another, two lateral stamens of the inner series (a1 and a2) are represented each by a slender naked filament. There are remaining petaloid bodies enough to answer for the third stamen of the inner series and for the stigmas, but their order is not well to be made out in the dried specimens." It may here be mentioned that Isochilus is normally triandrous.

A tetrandrous flower of Cypripedium has also been recorded.

In Isochilus, according to Cruger, there are often five stamens, and there are several, besides those already mentioned, in which six more or less perfect stamens have been seen—of these the following may be taken as illustrations. A hexandrous flower of Orchis militaris has been recorded by Kirschleger,[452] and in the accompanying diagram (fig. 195), from Cramer,[453] of a monstrous flower of Orchis mascula, there is one perfect stamen of the outer row and two lip-like stamens of the same series, while the inner verticil comprises one perfect and two abortive stamens.

Fig. 195.—Diagram of flower of Orchis mascula with two additional lips, two perfect and two imperfect stamens (after Cramer).

Morren[454] describes some flowers of Orchis morio in which there were three sepals, three petals, and within the latter two other ternary series of petals; this would seem to be a case of petalody of all six stamens. Morren, however, seems to have considered the additional segments as repetitions of the corolline whorl, though he describes a central mass as the column bearing a "souvenir of the anther." Nevertheless, there is no decisive evidence either in his figure or his description in support of his opinion as to the nature of the central mass, which might be a distorted condition of the styles, or, as is more probable, a rudimentary and irregular flower. Morren also describes another flower of the same plant in which there were three sepals, two lateral petals partially lip-like in aspect, a third labellum normal, two additional labella representing the two anterior stamens of the outer whorl, while more or less developed rudiments of the remaining four stamens also exist.

While, in most cases, the supernumerary stamens can, by reason of their relative position, their complete or partial antheriferous nature, be safely referred to one or other of the six stamens, making up a typical orchid flower, there are other specimens in which the additional stamens are altogether adventitious, and do not admit of reference to the homologue. Thus it was in a specimen of Odontoglossum Alexandræ examined by the writer, and in which, within a normally constructed perianth, there were six columns, all polliniferous, but arranged in so confused and complicated a manner that it was impossible to make out any definite relation in their position. There was nothing to indicate a fusion of flowers, but rather an extension of the centre of the flower, and consequent displacement of the stamens, &c. Again, the existence of adventitious stamens does not necessarily imply the development of organs usually suppressed, inasmuch as they may result from the assumption by the lateral petals of staminal characteristics.

Nevertheless, as far as teratology is concerned, specimens may be found in which some or all of the usually suppressed stamens of Orchidaceæ may be found. These stamens may be all perfect (polliniferous), or, as is more frequently the case, more or less petal-like. Moreover, when the stamens are petalodic, the form assumed is usually that of the labellum.

The presence of stamens in undue numbers in orchids is very generally, but not always, attended by some coincident malformation, of which the most frequent is cohesion of two or more sepals, and consequent displacement or adhesion of one petal to the side of the column. Petalody of the styles and median prolification are also sometimes found in association with an augmented number of stamens.

Fig. 196.—Increased number of carpels, tulip.

Fig. 197.—Fruit of St. Valery apple cut lengthwise.

Pleiotaxy of the gynœcium.—An increase in the number of whorls of which the pistil consists is not of very frequent occurrence. Generally after the formation of the whorl of carpels, the energy of the growing point ceases, or if by chance it be continued, the result is more generally the production of a new flower-bud (median prolification) than the repetition of the carpellary series. It is necessary also to distinguish between the veritable augmentation of the pistil and the semblance of it, brought about by the substitution of carpels for some other organs, as pistillody of the stamens, and even of the segments of the perianth, is not very unfrequent, as has already been stated under the head of substitution. Again, the increased number of carpels which is sometimes met with in such flowers, as Magnolia or Delphinium, where the ovaries are arranged in spiral series, is not strictly referable to the present category.

The orange is one of the plants most frequently subject to an augmentation in the number of carpellary whorls; sometimes this is due to the stamens assuming the guise of carpels, but at other times the increase occurs without any alteration in the stamens or other organs. If the adventitious carpels be exposed, they are covered with yellow rind, while those portions that are covered by the primary carpels are destitute of rind. Some varieties of the double tulip are very subject to a similar change, but, in this case, the petals and the stamens very frequently become more or less carpellary in their nature. Fig. 196 represents an increased number of whorls of carpels in the variety called "rex rubrorum," the segments of the perianth having been removed.

In the St. Valery apple, already referred to, there is a second whorl of carpels above the first, a fact which has been made use of to explain the similar structure of the pomegranate.

The tomato (Lycopersicum esculentum) is another plant in which an adventitious series is frequently produced, and generally in combination with the primary series.

In the Chinese primrose (Primula sinensis) a supernumerary whorl is frequently met with, generally associated with other changes in the construction and arrangement of the parts of the flower.

M. de Candolle[455] mentions a flower of Gentiana purpurea with four carpels in one series, and five others in the circle immediately above them. Wigand[456] alludes to an instance wherein there was a second pair of carpels above the first in Vinca herbacea. Dr. Sankey has forwarded flowers of a Pelargonium having a double series of carpels, eight in the outer row, five in the inner, and this condition is stated to exist in the flowers of the same plant for two years consecutively. In Aquilegia I have met with a similar increase in the whorls of carpels.[457] Meissner records a similar augmentation in Polygonum orientale.[458]

Wigand[459] describes and figures a flower of Vinca minor, in which there were two carpels intervening between the ordinary pair, and a similar illustration has been observed by the writer in Allamanda cathartica. Eichler[460] has put on record a similar case in a capparid.

Marchand[461] mentions a polycarpellary berberid (Epimedium Musschianum). The supernumerary carpels in this flower were placed on a short axis, which originated in the axils of the stamens, and as these latter organs were present in their usual number and position, the adventitious carpels could not be considered as resulting from a transformation, or substitution of carpels for stamens.

Lastly, the instance cited by Dr. Allman[462] in Saxifraga Geum may be alluded to. Here there was a row of adventitious carpels between the stamens and pistils, the backs of the carpels being turned towards the axis of the flowers. Dr. Allman explains the presence of the supernumerary parts by the supposed production of a whorl of secondary axes between the stamens and the centre of the flower. These axes are further supposed to bear imperfect flowers, of which the additional carpels are the only traces, but this explanation seems forced.

In addition to the references already cited the following may be given:

Duchartre, 'Ann. Sc. Nat.,' 4 ser., vii, p. 23 (Tulip).

Ferrari, 'Hesperides,' pp. 271, 395, 405. Duchartre, 'Ann. Sc. Nat.,' 4 ser., 1844, vol. i, p. 294. Maout, 'Leçons Elément.,' vol. ii, pp. 488–9. Clos, 'Ann. Sc. Nat.,' 1865, p. 317 (Citrus Aurantium).

Clos, 'Bull. Soc. Bot. Fr.,' vol. xiii; 'Rev. Bibl.,' p. 75. Pasquale, 'Reddicont Accad. Sc. Fis. e Math. Napoli.' Octr. 1866 (Solanum Lycopersicum).

On the general subject of multiplication, in addition to previous citations, the reader may be referred to A. P. de Candolle, 'Théorie Elément. Bot.,' ed. 3, p. 89.

Increased number of flowers in an inflorescence.—This happens generally as a result of over luxuriant growth, and scarcely demands notice here, being rather referable to variation than to malformation. The increased number of florets in the spikelets of some grasses has already been alluded to (p. 351). Thus spikelets of wheat occasionally produce more than the three florets which are proper to them.[463] It will be remembered that in this as in many other grasses there are rudimentary florets, and it is no matter for surprise that these florets should occasionally be fully developed.