The most remarkable arrangement of bracts takes place in Grasses, in which they occupy the place of calyx and corolla, and have received a variety of names from different systematic writers. In order to explain the application of these terms, it is necessary to describe with some minuteness the structure of a locusta or spikelet, as the partial inflorescence of Grasses is denominated. Take, for example, any common Bromus; each spikelet will be seen to have at its base two opposite empty bracts (fig. 67. b), one of which is attached to the rachis a little above the base of the other: these are the glumes of Linnæus and most botanists, the gluma exterior or calycinalis of some writers, the tegmen of Palisot de Beauvois, the lepicena of Richard, the catonium of Trinius, and, finally, the peristachyum of Panzer. Above the glumes are several florets sitting in denticulations of the rachis (fig. 67. c): each of these consists of one bract, with the midrib quitting the blade a little below the apex, and elongated into a bristle called the awn, beard, or arista, and of another bract facing the first, with its back to the rachis, bifid at the apex, with no dorsal vein, but with its edges inflexed, and a rib on each side at the line of inflexion (fig. 67. a). These bracts are the corolla of Linnæus, the calyx of Jussieu, the perianthium of Brown, the gluma interior or corollina and perigonium of some, the stragulum of Palisot de Beauvois, the gluma of Richard, the bale or glumella of De Candolle and Desvaux, the palea of others. When the arista proceeds from the very apex of the bracts, and not from below it, it is denominated in the writings of Palisot a seta. Within the last-mentioned bracts, and opposite to them, are situated two extremely minute, colourless fleshy scales (fig. 67. e), which are sometimes connate these are named corolla by Micheli and Dumortier, nectarium by Linnæus, squamulæ by Jussieu and Brown, glumella by Richard, glumellula by Desvaux and De Candolle, lodicula by Palisot de Beauvois, and periphyllia by Link. Amidst these conflicting terms it is not easy to determine which to adopt. I recommend the exterior empty bracts to be called glumes; those immediately surrounding the fertilising organs palea; and the minute hypogynous ones scales or squamulæ.

The pieces of which these three classes of bracts are composed are called valves or valvulæ by the greater part of botanists; but, as that term has been thought not to convey an accurate idea of their nature, Desvaux has proposed to substitute that of spathella, which is adopted by De Candolle. Palisot proposed to restrict the term glume to the pieces of the glume, and to call the pieces of the perianthium paleæ. Richard called the pieces of both glume and perianthium palea, and the scales paleola. It seems to me most convenient to use the term valvula, because it is more familiar to botanists than any other, and because I do not see the force of the objection which is taken to it.

In the genus Carex two bracts (fig. 67. i, h) become confluent at the edges, and enclose the pistil, leaving a passage for the stigmas at their apex. They thus form a single urceolate body, named urceolus or perigynium. De Candolle justly observes, in his Théorie, that some botanists call this nectarium, although it does not produce honey; others capsula, although it has nothing to do with the fruit; but he does not seem to me more correct than those he criticises in arranging the urceolus among his miscellaneous appendages of the floral organs, which are "ni organes génitaux ni tégumens." I believe I was the first who explained the true nature of the urceolus, in my translation of Richard's Analyse du Fruit, printed in 1819 (p. 13.).

At the base of the ovary of Cyperaceæ are often found little filiform appendages, called hypogynous setæ (fig. 67. d) by most botanists, and perigynium by Nees von Esenbeck. These are probably of the nature of the squamulæ of Grasses, and have been named perisporum by some French writers.

Bracts are generally distinct from each other, and imbricated or alternate. Nevertheless, there are some striking exceptions to this; as remarkable instances of which may be cited Althea and Lavatera among Malvaceæ, Euphorbia, all Dipsacea, and some Trifolia, particularly my Tr. cyathiferum, in all which the bracts are accurately verticillate, and their margins confluent, as in a true calyx.

The Flower is a terminal bud enclosing the organs of reproduction by seed. By the ancients the term flower was restricted to what is now called the corolla; but Linnæus wisely extended its application to the union of all the organs which contribute to the process of fecundation. The flower, therefore, as now understood, comprehends the calyx, the corolla, the stamens, and the pistil, of which the two last only are indispensable. The calyx and corolla may be wanting, and a flower will nevertheless exist; but, if neither stamens nor pistil nor their rudiments are to be found, no assemblage of leaves, whatever may be their form or colour, or how much soever they may resemble the calyx and corolla, can constitute a flower.

We usually consider the flower to consist of a certain number of whorls, or of parts originating round a common centre from the same plane. But Adolphe Brongniart has correctly pointed out the fact that what we call whorls in a flower are in many cases not so, strictly speaking, but only a series of parts in close approximation, and at different heights upon the short branch that forms the axis. This is particularly obvious in a Cistus, where, of the five sepals, two are lower and exterior, and three higher and within the first. The manner also in which the petals overlap each other evidently points to a similar cause, although the fact of those pieces

being inserted at different heights may not be apparent. (See Ann. des Sc. v. xxiii. p. 226.)

The flower, when in the state of a bud, is called the alabastrus (bouton of the French); a name used by Pliny for the rose-bud. Some writers say alabastrum, forgetting, as it would seem, that that term was used by the Romans for a scent-box, and not for the bud of a flower. Link calls the parts of a flower generally, whether united or connate, moria, whence a flower is bi-polymorious (Elem., 243.); but I know of no other writer who employs these terms, which indeed are superfluous.

The flowers of a capitulum, small, and somewhat different in structure from ordinary flowers, are called florets (flosculi ; elytriculi of Necker; fleurons of the French).

The period when a flower opens is called its anthesis; the manner in which its parts are arranged, with respect to each other, before the opening, is called the astivation. Estivation is the same to a flower-bud as vernation is to a leaf-bud: the terms expressive of its modifications are to be sought in Glossology. This term æstivation is applied separately to the parts of which a flower may consist; thus, we speak of the æstivation of the calyx, of the corolla, of the stamens, and of the pistil; but not of the aestivation of a flower collectively.

5. Of the Inflorescence.

Inflorescence is a term contrived to express generally the arrangement of flowers upon a branch or stem. The part which immediately bears the flowers is called the pedunculus or peduncle, and is to be distinguished from any portion of a branch by not producing perfect leaves; those which are tound upon it, called bracts, being much reduced in size and figure from what are borne by the rest of the plant.

The normal position of the inflorescence is axillary to a leaf, the necessary consequence of its being a kind of branching. But in some plants, especially of the natural order Solanaceæ, it grows apparently opposite the leaves. It is

believed that cases of such irregularity are caused by the peduncle, which is axillary to a leaf, contracting an adhesion with the internodium above it, and not separating till it is opposite the succeeding leaf. Flowers of this kind are called oppositifolii.

The term peduncle, although it may be understood to apply to all the parts of the inflorescence which bear the flowers, is practically only made use of to denote the immediate support of a single solitary flower or the whole mass of inflorescence, and is therefore confined to that part of the inflorescence which first proceeds from the stem. If it is divided, its principal divisions are called branches; and its ultimate ramifications, which bear the flowers, are named pedicels. There are also other names which are applied to its modifications. In plants which are destitute of stem, it often rises above the ground, supporting the flowers on its apex, as in the Cowslip. Such a peduncle is named a scape (hampe, Fr.). Some botanists distinguish from the scape the pedunculus radicalis, confining the former term to the peduncle which arises from the central bud of the plant, as in the Hyacinth; and applying the latter to a peduncle proceeding from a lateral bud, as in Plantago media. When a peduncle proceeds in a nearly right line from the base to the apex of the inflorescence, it is called the rachis, or the axis of the inflorescence. This latter term was used by Palisot de Beauvois to express the rachis of Grasses, and is perhaps the better term of the two, especially as the term rachis is applied by Willdenow and others to the petiole and midrib of Ferns. In the spikelets of Grasses the rachis has an unusual, toothed, flexuose appearance, and has received the name of scobina from Dumortier; if it is reduced to a mere bristle, as in some of the single-flowered spikelets, the same writer then distinguishes it by the name of acicula.

When the part which bears the flowers is repressed in its developement, so that, instead of being lengthened into a rachis, it forms a flattened area on which the flowers are arranged, it becomes what is called a receptacle; or, in the language of some botanists, the receptacle of the flower (fig. 72.)