266. There are all gradations between labiate and regular corollas. In those of Gerardia, of some species of Pentstemon, and of Catalpa (Fig. 263-265), the labiate character is slight, but is manifest on close inspection. In almost all such flowers the plan of five, which is obvious or ascertainable in the calyx and corolla, is obscured in the stamens by the abortion or suppression of one or three of their number.

267. Ligulate Corolla. The ligulate or Strap-shaped corolla mainly belongs to the family of Compositæ, in which numerous small flowers are gathered into a head, within an involucre that imitates a calyx. It is best exemplified in the Dandelion and in Chiccory (Fig. 266). Each one of these straps or Ligules, looking like so many petals, is the corolla of a distinct flower: the base is a short tube, which opens out into the ligule: the five minute teeth at the end indicate the number of constituent petals. So this is a kind of gamopetalous corolla, which is open along one side nearly to the base, and outspread. The nature of such a corolla (and of the stamens also, to be explained in the next section) is illustrated by the flower of a Lobelia, Fig. 285.

268. In Asters, Daisies, Sunflower, Coreopsis (Fig. 268), and the like, only the marginal (or Ray) corollas are ligulate; the rest (those of the Disk) are regularly gamopetalous, tubular, and five-lobed at summit; but they are small and individually inconspicuous, only the ray-flowers making a show. In fact, those of Coreopsis and of Sunflower are simply for show, these ray-flowers being not only sterile, but neutral, that is, having neither stamens nor pistil. But in Asters, Daisies, Golden-rods, and the like, these ray-flowers are pistillate and fertile, serving therefore for seed-bearing as well as for show. Let it not be supposed that the show is useless. See Section XIII.

269. Adnation, or Consolidation, is the union of the members of parts belonging to different circles of the flower (256). It is of course understood that in this (as likewise in coalescence) the parts are not formed and then conjoined, but are produced in union. They are born united, as the term adnate implies. To illustrate this kind of union, take the accompanying series of flowers (Fig. 270-274), shown in vertical section. In the first, Fig. 270, Flax-flower, there is no adnation; sepals, petals, and stamens, are free as well as distinct, being separately borne on the receptacle, one circle within or above the next; only the five pistils have their ovaries coalescent. In Fig. 271, a Cherry-flower, the petals and stamens are borne on the throat of the calyx-tube; that is, the sepals are coalescent into a cup, and the petals and stamens are adnate to the inner face of this; in other words, the sepals, petals, and stamens are all consolidated up to a certain height. In Fig. 272, a Purslane-flower, the same parts are adnate to or consolidated with the ovary up to its middle. In Fig. 273, a Hawthorn-flower, the consolidation has extended over the whole ovary; and petals and stamens are adnate to the calyx still further. In Fig. 274, a Cranberry-blossom, it is the same except that all the parts are free at the same height; all seem to arise from the top of the ovary.

270. In botanical description, to express tersely such differences in the relation of these organs to the pistil, they are said to be

Hypogynous (i. e. under the pistil) when they are all free, that is, not adnate to pistil nor connate with each other, as in Fig. 270.

Perigynous (around the pistil) when connate with each other, that is, when petals and stamens are inserted or borne on the calyx, whether as in Cherry-flowers (Fig. 271) they are free from the pistil, or as in Purslane and Hawthorn (Fig. 272, 273) they are also adnate below to the ovary.

Epigynous (on the ovary) when so adnate that all these parts appear to arise from the very summit of the ovary, as in Fig. 274. The last two terms are not very definitely distinguished.

271. Another and a simpler form of expression is to describe parts of the flower as being

Free, when not united with or inserted upon other parts.

Distinct, when parts of the same kind are not united. This term is the counterpart of coalescent, as free is the counterpart of adnate. Many writers use the term "free" indiscriminately for both; but it is better to distinguish them.

Connate is a term common for either not free or not distinct, that is, for parts united congenitally, whether of same or of different kinds.

Adnate, as properly used, relates to the union of dissimilar parts.

272. In still another form of expression, the terms superior and inferior have been much used in the sense of above and below.

Superior is said of the ovary of Flax-flower, Cherry, etc., because above the other parts; it is equivalent to "ovary free." Or it is said of the calyx, etc., when above the ovary, as in Fig. 273-275.

Inferior, when applied to the ovary, means the same as "calyx adnate;" when applied to the floral envelopes, it means that they are free.

273. Position of Flower or of its Parts. The terms superior and inferior, or upper and lower, are also used to indicate the relative position of the parts of a flower in reference to the axis of inflorescence. An axillary flower stands between the bract or leaf which subtends it and the axis or stem which bears this bract or leaf. This is represented in sectional diagrams (as in Fig. 275, 276) by a transverse line for the bract, and a small circle for the axis of inflorescence. Now the side of the blossom which faces the bract is the

Anterior, or Inferior, or Lower side; while the side next the axis is the

Posterior, or Superior, or Upper side of the flower.

274. So, in the labiate corolla (Fig. 256-258), the lip which is composed of three of the five petals is the anterior, or inferior, or lower lip; the other is the posterior, or superior, or upper lip.

275. In Violets (Fig. 238, 276), the odd sepal is posterior (next the axis); the odd petal is therefore anterior, or next the subtending leaf. In the papilionaceous flower (Fig. 261, and diagram, Fig. 275), the odd sepal is anterior, and so two sepals are posterior; consequently, by the alternation, the odd petal (the standard) is posterior or upper, and the two petals forming the keel are anterior or lower.

§ 5. ARRANGEMENT OF PARTS IN THE BUD.

276. Æstivation was the fanciful name given by Linnæus to denote the disposition of the parts, especially the leaves of the flower, before Anthesis, i. e. before the blossom opens. Præfloration, a better term, is sometimes used. This is of importance in distinguishing different families or genera of plants, being generally uniform in each. The æstivation is best seen by making a slice across the flower-bud; and it may be expressed in diagrams, as in the accompanying figures.

277. The pieces of the calyx or the corolla either overlap each other in the bud, or they do not. When they do not overlap, the æstivation is

Valvate, when the pieces meet each other by their abrupt edges, without any infolding or overlapping; as the calyx of the Linden or Basswood (Fig. 277).

Induplicate, which is valvate with the margins of each piece projecting inwards, as in the calyx of a common Virgin's-bower, Fig. 278, or

Involute, which is the same but the margins rolled inward, as in most of the large-flowered species of Clematis, Fig. 279.

Reduplicate, a rarer modification of valvate, is similar but with margins projecting outward.

Open, the parts not touching in the bud, as the calyx of Mignonette.

278. When the pieces overlap in the bud, it is in one of two ways; either every piece has one edge in and one edge out, or some pieces are wholly outside and others wholly inside. In the first case the æstivation is

Convolute, also named Contorted or Twisted, as in Fig. 280, a cross-section of a corolla very strongly thus convolute or rolled up together, and in the corolla of a Flax-flower (Fig. 281), where the petals only moderately overlap in this way. Here one edge of every petal covers the next before it, while its other edge is covered by the next behind it. The other mode is the

Imbricate or Imbricated, in which the outer parts cover or overlap the inner so as to "break joints," like tiles or shingles on a roof; whence the name. When the parts are three, the first or outermost is wholly external, the third wholly internal, the second has one margin covered by the first while the other overlaps the third or innermost piece: this is the arrangement of alternate three-ranked leaves (187). When there are five pieces, as in the corolla of Fig. 225, and calyx of Fig. 281, as also of Fig. 241, 276, two are external, two are internal, and one (the third in the spiral) has one edge covered by the outermost, while its other edge covers the innermost; which is just the five-ranked arrangement of alternate leaves (188). When the pieces are four, two are outer and two are inner; which answers to the arrangement of opposite leaves.

279. The imbricate and the convolute modes sometimes vary one into the other, especially in the corolla.

280. In a gamopetalous corolla or gamosepalous calyx, the shape of the tube in the bud may sometimes be noticeable. It may be

Plicate or Plaited, that is, folded lengthwise; and the plaits may either be turned outwards, forming projecting ridges, as in the corolla of Campanula; or turned inwards, as in that of Gentian Belladonna; or

Supervolute, when the plaits are convolutely wrapped round each other, as in the corolla of Morning Glory and of Stramonium, Fig. 282.

Section IX. STAMENS IN PARTICULAR.

281. Andrœcium is a technical name for the staminate system of a flower (that is, for the stamens taken together), which it is sometimes convenient to use. The preceding section has dealt with modifications of the flower pertaining mainly to calyx and corolla. Those relating to the stamens are now to be indicated. First as to

282. Insertion, or place of attachment. The stamens usually go with the petals. Not rarely they are at base

Epipetalous, that is, inserted on (or adnate to) the corolla, as in Fig. 283. When free from the corolla, they may be

Epigynous, borne apparently on the top of the ovary; all which is explained in Fig. 270-274.

Gynandrous is another term relating to insertion of rarer occurrence, that is, where the stamens are inserted on (in other words, adnate to) the style, as in Lady's Slipper (Fig. 284), and in the Orchis family generally.

283. In Relation to each Other, stamens are more commonly

Distinct, that is, without any union with each other. But when united, the following technical terms of long use indicate their modes of mutual connection:—

Monadelphous (from two Greek words, meaning "in one brotherhood"), when united by their filaments into one set, usually into a ring or cup below, or into a tube, as in the Mallow Family (Fig. 286), the Passion-flower (Fig. 260), the Lupine (Fig. 287), and in Lobelia (Fig. 285).

Diadelphous (meaning in two brotherhoods), when united by the filaments into two sets, as in the Pea and most of its near relatives (Fig. 288), usually nine in one set, and one in the other.

Triadelphous (three brotherhoods), when the filaments are united in three sets or clusters, as in most species of Hypericum.

Pentadelphous (five brotherhoods), when in five sets, as in some species of Hypericum and in American Linden (Fig. 277, 289).

Polyadelphous (many or several brotherhoods) is the term generally employed when these sets are several, or even more than two, and the particular number is left unspecified. These terms all relate to the filaments.

Syngenesious is the term to denote that stamens have their anthers united, coalescent into a ring or tube; as in Lobelia (Fig. 285), in Violets, and in all of the great family of Compositæ.

284. Their Number in a flower is commonly expressed directly, but sometimes adjectively, by a series of terms which were the name of classes in the Linnæan artificial system, of which the following names, as also the preceding, are a survival:—

Monandrous, i. e. solitary-stamened, when the flower has only one stamen,

Diandrous, when it has two stamens only,

Triandrous, when it has three stamens,

Tetrandrous, when it has four stamens,

Pentandrous, when it has five stamens,

Hexandrous, when with six stamens, and so on to

Polyandrous, when it has many stamens, or more than a dozen.

285. For which terms, see the Glossary. They are all Greek numerals prefixed to -andria (from the Greek), which Linnæus used for andrœcium, and are made into an English adjective, -androus. Two other terms, of same origin, designate particular cases of number (four or six) in connection with unequal length. Namely, the stamens are

Didynamous, when, being only four, they form two pairs, one pair longer than the other, as in the Trumpet Creeper, in Gerardia (Fig. 263), etc.

Tetradynamous, when, being only six, four of them surpass the other two, as in the Mustard-flower and all the Cruciferous family, Fig. 235.

286. The Filament is a kind of stalk to the anther, commonly slender or thread-like: it is to the anther nearly what the petiole is to the blade of a leaf. Therefore it is not an essential part. As a leaf may be without a stalk, so the anther may be Sessile, or without a filament.

287. The Anther is the essential part of the stamen. It is a sort of case, filled with a fine powder, the Pollen, which serves to fertilize the pistil, so that it may perfect seeds. The anther is said to be

Innate (as in Fig. 292), when it is attached by its base to the very apex of the filament, turning neither inward nor outward;

Adnate (as in Fig. 293), when attached as it were by one face, usually for its whole length, to the side of a continuation of the filament; and

Versatile (as in Fig. 294), when fixed by or near its middle only to the very point of the filament, so as to swing loosely, as in the Lily, in Grasses, etc. Versatile or adnate anthers are

Introrse, or Incumbent, when facing inward, that is, toward the centre of the flower, as in Magnolia, Water-Lily, etc.

Extrorse, when facing outwardly, as in the Tulip-tree.

288. Rarely does a stamen bear any resemblance to a leaf, or even to a petal or flower-leaf. Nevertheless, the botanist's idea of a stamen is that it answers to a leaf developed in a peculiar form and for a special purpose. In the filament he sees the stalk of the leaf; in the anther, the blade. The blade of a leaf consists of two similar sides; so the anther consists of two Lobes or Cells, one answering to the left, the other to the right, side of the blade. The two lobes are often connected by a prolongation of the filament, which answers to the midrib of a leaf; this is called the Connective. This is conspicuous in Fig. 292, where the connective is so broad that it separates the two cells of the anther to some distance.

289. A simple conception of the morphological relation of an anther to a leaf is given in Fig. 295, an ideal figure, the lower part representing a stamen with the top of its anther cut away; the upper, the corresponding upper part of a leaf.

290. So anthers are generally two-celled. But as the pollen begins to form in two parts of each cell (the anterior and the posterior), sometimes these two strata are not confluent, and the anther even at maturity may be four-celled, as in Moonseed (Fig. 296); or rather, in that case (the word cell being used for each lateral half of the organ), it is two-celled, but the cells bilocellate.

292. By confluence, when the two cells run together into one, as they nearly do in most species of Pentstemon (Fig. 297), more so in Monarda (Fig. 300), and completely in the Mallow (Fig. 298) and all the Mallow family.

293. By suppression in certain cases the anther may be reduced to one cell or halved. In Globe Amaranth (Fig. 299) there is a single cell without vestige of any other. Different species of Sage and of the White Sages of California show various grades of abortion of one of the anther-cells, along with a singular lengthening of the connective (Fig. 302-305).

294. The splitting open of an anther for the discharge of its pollen is termed its Dehiscence.

295. As the figures show, this is commonly by a line along the whole length of each cell, either lateral or, when the anthers are extrorse, often along the outer face, and when introrse, along the inner face of each cell. Sometimes the opening is only by a chink, hole, or pore at the top, as in the Azalea, Pyrola (Fig. 307), etc.; sometimes a part of the face separates as a sort of trap-door (or valve), hinged at the top, and opening to allow the escape of the pollen, as in the Sassafras, Spice-bush, and Barberry (Fig. 308).

296. Pollen. This is the powdery matter, commonly of a yellow color, which fills the cells of the anther, and is discharged during blossoming, after which the stamens generally fall or wither away. Under the microscope it is found to consist of grains, usually round or oval, and all alike in the same species, but very different in different plants. So that the plant may sometimes be recognized from the pollen alone. Several forms are shown in the accompanying figures.

297. An ordinary pollen-grain has two coats; the outer coat thickish, but weak, and frequently adorned with lines or bands, or studded with points; the inner coat is extremely thin and delicate, but extensible, and its cavity when fresh contains a thickish protoplasmic fluid, often rendered turbid by an immense number of minute particles that float in it. As the pollen matures this fluid usually dries up, but the protoplasm does not lose its vitality. When the grain is wetted it absorbs water, swells up, and is apt to burst, discharging the contents. But when weak syrup is used it absorbs this slowly, and the tough inner coat will sometimes break through the outer and begin a kind of growth, like that which takes place when the pollen is placed upon the stigma.

298. Some pollen-grains are, as it were, lobed (as in Fig. 315, 316), or formed of four grains united (as in the Heath family, Fig. 317): that of Pine (Fig. 318) has a large rounded and empty bladder-like expansion upon each side. This renders such pollen very buoyant, and capable of being transported to a great distance by the wind.

299. In species of Acacia simple grains lightly cohere into globular pellets. In Milkweeds and in most Orchids all the pollen of an anther-cell is compacted or coherent into one mass, called a Pollen-mass, or Pollinium, plural Pollinia. (Fig. 319-322.)

Section X. PISTILS IN PARTICULAR.

§ 1. ANGIOSPERMOUS OR ORDINARY GYNŒCIUM.

300. Gynœcium is the technical name for the pistil or pistils of a flower taken collectively, or for whatever stands in place of these. The various modifications of the gynœcium and the terms which relate to them require particular attention.

301. The Pistil, when only one, occupies the centre of the flower; when there are two pistils, they stand facing each other in the centre of the flower; when several, they commonly form a ring or circle; and when very numerous, they are generally crowded in rows or spirals on the surface of a more or less enlarged or elongated receptacle. Their number gives rise to certain terms, the counterpart of those used for stamens (284), which are survivals of the names of orders in the Linnæan artificial system. The names were coined by prefixing Greek numerals to -gynia used for gynœcium, and changed into adjectives in the form of -gynous. That is, a flower is

Monogynous, when it has a single pistil, whether that be simple or compound;

Digynous, when it has only two pistils; Trigynous, when with three; Tetragynous, with four; Pentagynous, with five; Hexagynous, with six; and so on to Polygynous, with many pistils.

302. The Parts of a Complete Pistil, as already twice explained (16, 236), are the Ovary, the Style, and the Stigma. The ovary is one essential part: it contains the rudiments of seeds, called Ovules. The stigma at the summit is also essential: it receives the pollen, which fertilizes the ovules in order that they may become seeds. But the style, commonly a tapering or slender column borne on the summit of the ovary, and bearing the stigma on its apex or its side, is no more necessary to a pistil than the filament is to the stamen. Accordingly, there is no style in many pistils: in these the stigma is sessile, that is, rests directly on the ovary (as in Fig. 326). The stigma is very various in shape and appearance, being sometimes a little knob (as in the Cherry, Fig. 271), sometimes a point or small surface of bare tissue (as in Fig. 327-330), and sometimes a longitudinal crest or line (as in Fig. 324, 341-343), or it may occupy the whole length of the style, as in Fig. 331.

303. The word Pistil (Latin, Pistillum) means a pestle. It came into use in the first place for such flowers as those of Crown Imperial, or Lily, in which the pistil in the centre was likened to the pestle, and the perianth around it to the mortar, of the apothecary.

304. A pistil is either simple or compound. It is simple when it answers to a single flower-leaf, compound when it answers to two or three, or a fuller circle of such leaves conjoined.

305. Carpels. It is convenient to have a name for each flower-leaf of the gynœcium; so it is called a Carpel, in Latin Carpellum or Carpidium. A simple pistil is a carpel. Each component flower-leaf of a compound pistil is likewise a carpel. When a flower has two or more pistils, these of course are simple pistils, that is, separate carpels or pistil-leaves. There may be only a single simple pistil to the flower, as in a Pea or Cherry blossom (Fig. 271); there may be two such, as in many Saxifrages; or many, as in the Strawberry. More commonly the single pistil in the centre of a blossom is a compound one. Then there is seldom much difficulty in ascertaining the number of carpels or pistil-leaves that compose it.

306. The Simple Pistil, viewed morphologically, answers to a leaf-blade with margins incurved and united where they meet, so forming a closed case or pod (the ovary), and bearing ovules at the suture or junction of these margins: a tapering upper portion with margins similarly inrolled, is supposed to form the style; and these same margins, exposed at the tip or for a portion of the length, become the stigma. Compare, under this view, the three accompanying figures.

307. So a simple pistil should have a one-celled ovary, only one line of attachment for the ovules, a single style, and a single stigma. Certain variations from this normal condition which sometimes occur do not invalidate this morphological conception. For instance, the stigma may become two-lobed or two-ridged, because it consists of two leaf-margins, as Fig. 324 shows; it may become 2-locellate by the turning or growing inward of one of the sutures, so as to divide the cavity.

308. There are two or three terms which primarily relate to the parts of a simple pistil or carpel, and are thence carried on to the compound pistil, viz.:—

Ventral Suture, the line which answers to the united margins of the carpel-leaf, therefore naturally called a suture or seam, and the ventral or inner one, because in the circle of carpel-leaves it looks inward or to the centre of the flower.

Dorsal Suture is the line down the back of the carpel, answering to the midrib of the leaf,—not a seam therefore; but at maturity many fruits, such as pea-pods, open by this dorsal as well as by the ventral line.

Placenta, a name given to the surface, whatever it be, which bears the ovules and seeds. The name may be needless when the ovules grow directly on the ventral suture, or from its top or bottom; but when there are many ovules there is usually some expansion of an ovule-bearing or seed-bearing surface; as is seen in our Mandrake or Podophyllum, Fig. 326.