Part 21

When bracts become united, and overlie each other in several rows, it often happens that the outer ones do not produce flowers, that is, are empty or sterile. In the artichoke the outer imbricated scales or bracts are in this condition, and it is from the membranous white scales or bracts (_paleae_) forming the choke attached to the edible receptacle that the flowers are produced. The sterile bracts of the daisy occasionally produce capitula, and give rise to the hen-and-chickens daisy. In place of developing flower-buds, bracts may, in certain circumstances, as in proliferous or viviparous plants, produce leaf-buds.

A sheathing bract enclosing one or several flowers is called a _spathe_. It is common among Monocotyledons, as _Narcissus_ (fig. 4), snow-flake, _Arum_ and palms. In some palms it is 20 ft. long, and encloses 200,000 flowers. It is often associated with that form of inflorescence termed the _spadix_, and may be coloured, as in _Anthurium_, or white, as in arum lily (_Richardia aethiopica_). When the spadix is compound or branching, as in palms, there are smaller spathes, surrounding separate parts of the inflorescence. The spathe protects the flowers in their young state, and often falls off after they are developed, or hangs down in a withered form, as in some palms, _Typha_ and _Pothos_. In grasses the outer scales or glumes of the spikelets are sterile bracts (fig. 5, gl); and in Cyperaceae bracts enclose the organs of reproduction. Bracts are frequently changed into complete leaves. This change is called _phyllody_ of bracts, and is seen in species of _Plantago_, especially in the variety of _Plantago media_, called the rose-plantain in gardens, where the bracts become leafy and form a rosette round the flowering axis. Similar changes occur in _Plantago major_, _P. lanceolata_, _Ajuga reptans_, dandelion, daisy, dahlia and in umbelliferous plants. The conversion of bracts into stamens (_staminody_ of bracts) has been observed in the case of _Abies excelsa_. A lengthening of the axis of the female strobilus of Coniferae is not of infrequent occurrence in _Cryptomeria japonica_, larch (_Larix europaea_), &c., and this is usually associated with a leaf-like condition of the bracts, and sometimes even with the development of leaf-bearing shoots in place of the scales.

The arrangement of the flowers on the axis, or the ramification of the floral axis, is called the _inflorescence_. The primary axis of the inflorescence is sometimes called the _rachis_; its branches, whether terminal or lateral, which form the stalks supporting flowers or clusters of flowers, are _peduncles_, and if small branches are given off by it, they are called _pedicels_. A flower having a stalk is called _pedunculate_ or _pedicellate_; one having no stalk is _sessile_. In describing a branching inflorescence, it is common to speak of the rachis as the _primary_ floral axis, its branches as the _secondary_ floral axes, their divisions as the _tertiary_ floral axes, and so on; thus avoiding any confusion that might arise from the use of the terms _rachis_, _peduncle_ and _pedicel_.

The _peduncle_ is simple, bearing a single flower, as in primrose; or branched, as in London-pride. It is sometimes succulent, as in the cashew, in which it forms the large coloured expansion supporting the nut; spiral, as in _Cyclamen_ and _Vallisneria_; or spiny, as in _Alyssum spinosum_. When the peduncle proceeds from radical leaves, that is, from an axis which is so shortened as to bring the leaves close together in the form of a cluster, as in the primrose, auricula or hyacinth, it is termed a _scape_. The floral axis may be shortened, assuming a flattened, convex or concave form, and bearing numerous flowers, as in the artichoke, daisy and fig (fig. 6). The floral axis sometimes appears as if formed by several peduncles united together, constituting a fasciated axis, as in the cockscomb, in which the flowers form a peculiar crest at the apex of the flattened peduncles. Adhesions occasionally take place between the peduncle and the bracts or leaves of the plant, as in the lime-tree (fig. 7). The adhesion of the peduncles to the stem accounts for the extra-axillary position of flowers, as in many Solanaceae. When this union extends for a considerable length along the stem, several leaves may be interposed between the part where the peduncle becomes free and the leaf whence it originated, and it may be difficult to trace the connexion. The peduncle occasionally becomes abortive, and in place of bearing a flower, is transformed into a tendril; at other times it is hollowed at the apex, so as apparently to form the lower part of the outer whorl of floral leaves as in _Eschscholtzia_. The termination of the peduncle, or the part on which the whorls of the flower are arranged, is called the _thalamus_, _torus_ or _receptacle_.

Inflorescence.

There are two distinct types of inflorescence--one in which the flowers arise as lateral shoots from a primary axis, which goes on elongating, and the lateral shoots never exceed in their development the length of the primary axis beyond their point of origin. The flowers are thus always _axillary_. Exceptions, such as in cruciferous plants, are due to the non-appearance of the bracts. In the other type the primary axis terminates in a single flower, but lateral axes are given off from the axils of the bracts, which again repeat the primary axis; the development of each lateral axis is stronger than that of the primary axis beyond its point of origin. The flowers produced in this inflorescence are thus _terminal_. The first kind of inflorescence is _indeterminate_, _indefinite_ or _axillary_. Here the axis is either elongated, producing flower-buds as it grows, the lower expanding first (fig. 8), or it is shortened and depressed, and the outer flowers expand first (fig. 9). The expansion of the flowers is thus _centripetal_, that is, from base to apex, or from circumference to centre.

The second kind of inflorescence is _determinate_, _definite_ or _terminal_. In this the axis is either elongated and ends in a solitary flower, which thus terminates the axis, and if other flowers are produced, they belong to secondary axes farther from the centre; or the axis is shortened and flattened, producing a number of separate floral axes, the central one expanding first, while the others are developed in succession farther from the centre. The expansion of the flowers is in this case _centrifugal_, that is, from apex to base, or from centre to circumference. It is illustrated in fig. 10, _Ranunculus bulbosus_; a' is the primary axis swollen at the base in a bulb-like manner b, and with roots proceeding from it. From the leaves which are radical proceeds the axis ending in a solitary terminal flower f'. About the middle of this axis there is a leaf or bract, from which a secondary floral axis a" is produced, ending in a single flower f", less advanced than the flower f'. This secondary axis bears a leaf also, from which a tertiary floral axis a"' is produced, bearing an unexpanded solitary flower f"'. From this tertiary axis a fourth is in progress of formation. Here f' is the termination of the primary axis, and this flower expands first, while the other flowers are developed centrifugally on separate axes.

A third series of inflorescences, termed _mixed_, may be recognized. In them the primary axis has an arrangement belonging to the opposite type from that of the branches, or vice versa. According to the mode and degree of development of the lateral shoots and also of the bracts, various forms of both inflorescences result.

Amongst indefinite forms the simplest occurs when a lateral shoot produced in the axil of a large single foliage leaf of the plant ends in a single flower, the axis of the plant elongating beyond, as in _Veronica hederifolia_, _Vinca minor_ and _Lysimachia nemorum_. The flower in this case is _solitary_, and the ordinary leaves become bracts by producing flower-buds in place of leaf-buds; their number, like that of the leaves of this main axis, is indefinite, varying with the vigour of the plant. Usually, however, the floral axis, arising from a more or less altered leaf or bract, instead of ending in a solitary flower, is prolonged, and bears numerous bracteoles, from which smaller peduncles are produced, and those again in their turn may be branched in a similar way. Thus the flowers are arranged in groups, and frequently very complicated forms of inflorescence result. When the primary peduncle or floral axis, as in fig. 8, is elongated, and gives off pedicels, ending in single flowers, a _raceme_ is produced, as in currant, hyacinth and barberry. If the secondary floral axes give rise to tertiary ones, the raceme is branching, and forms a _panicle_, as in _Yucca gloriosa_. If in a raceme the lower flower-stalks are developed more strongly than the upper, and thus all the flowers are nearly on a level, a _corymb_ is formed, which may be simple, as in fig. 11, where the primary axis a' gives off secondary axes a", a", which end in single flowers; or branching, where the secondary axes again subdivide. If the pedicels are very short or wanting, so that the flowers are sessile, a _spike_ is produced, as in _Plantago_ and vervain (_Verbena officinalis_) (fig. 12). If the spike bears unisexual flowers, as in willow or hazel (fig. 13), it is an _amentum_ or _catkin_, hence such trees are called _amentiferous_; at other times it becomes succulent, bearing numerous flowers, surrounded by a sheathing bract or spathe, and then it constitutes a _spadix_, which may be simple, as in _Arum maculatum_ (fig. 14), or branching as in palms. A spike bearing female flowers only, and covered with scales, is a _strobilus_, as in the hop. In grasses there are usually numerous sessile flowers arranged in small spikes, called _locustae_ or _spikelets_, which are either set closely along a central axis, or produced on secondary axes formed by the branching of the central one; to the latter form the term panicle is applied.

If the primary axis, in place of being elongated, is contracted, it gives rise to other forms of indefinite inflorescence. When the axis is so shortened that the secondary axes arise from a common point, and spread out as _radii_ of nearly equal length, each ending in a single flower or dividing again in a similar radiating manner, an _umbel_ is produced, as in fig. 15. From the primary floral axis a the secondary axes come off in a radiating or umbrella-like manner, and end in small umbels b, which are called _partial umbels_ or _umbellules_. This inflorescence is seen in hemlock and other allied plants, which are hence called umbelliferous. If there are numerous flowers on a flattened, convex or slightly concave receptacle, having either very short pedicels or none, a _capitulum_ (head) is formed, as in dandelion, daisy and other composite plants (fig. 2), also in scabious (fig. 9) and teazel. In the American button-bush the heads are globular, in some species of teazel elliptical, while in scabious and in composite plants, as sunflower, dandelion, thistle, centaury and marigold, they are somewhat hemispherical, with a flattened, slightly hollowed, or convex disk. If the margins of such a receptacle be developed upwards, the centre not developing, a concave receptacle is formed, which may partially or completely enclose a number of flowers that are generally unisexual. This gives rise to the peculiar inflorescence of _Dorstenia_, or to that of the fig (fig. 6), where the flowers are placed on the inner surface of the hollow receptacle, and are provided with bracteoles. This inflorescence has been called a _hypanthodium_.

Lastly, we have what are called _compound indefinite_ inflorescences. In these forms the lateral shoots, developed centripetally upon the primary axis, bear numerous bracteoles, from which floral shoots arise which may have a centripetal arrangement similar to that on the mother shoot, or it may be different. Thus we may have a group of racemes, arranged in a racemose manner on a common axis, forming a raceme of racemes or compound raceme, as in _Astilbe_. In the same way we may have compound umbels, as in hemlock and most Umbelliferae (fig. 15), a compound spike, as in rye-grass, a compound spadix, as in some palms, and a compound capitulum, as in the hen-and-chickens daisy. Again, there may be a raceme of capitula, that is, a group of capitula disposed in a racemose manner, as in _Petasites_, a raceme of umbels, as in ivy, and so on, all the forms of inflorescence being indefinite in disposition. In _Eryngium_ the shortening of the pedicels changes an umbel into a capitulum.

The simplest form of the definite type of the inflorescence is seen in _Anemone nemorosa_ and in gentianella (_Gentiana acaulis_), where the axis terminates in a single flower, no other flowers being produced upon the plant. This is a _solitary terminal_ inflorescence. If other flowers were produced, they would arise as lateral shoots from the bracts below the first-formed flower. The general name of _cyme_ is applied to the arrangement of a group of flowers in a definite inflorescence. A _cymose_ inflorescence is an inflorescence where the primary floral axis before terminating in a flower gives off one or more lateral unifloral axes which repeat the process--the development being only limited by the vigour of the plant. The floral axes are thus centrifugally developed. The cyme, according to its development, has been characterized as _biparous_ or _uniparous_. In fig. 16 the biparous cyme is represented in the flowering branch of _Cerastium_. Here the primary axis t ends in a flower, which has passed into the state of fruit. At its base two leaves are produced, in each of which arise secondary axes t' t', ending in single flowers, and at the base of these axes a pair of opposite leaves is produced, giving rise to tertiary axes t" t", ending in single flowers, and so on. The term _dichasium_ has also been applied to this form of cyme.

In the natural order Carophyllaceae (pink family) the dichasial form of inflorescence is very general. In some members of the order, as _Dianthus barbatus_, _D. carthusianorum_, &c., in which the peduncles are short, and the flowers closely approximated, with a centrifugal expansion, the inflorescence has the form of a contracted dichasium, and receives the name of _fascicle_. When the axes become very much shortened, the arrangement is more complicated in appearance, and the nature of the inflorescence can only be recognized by the order of opening of the flowers. In Labiate plants, as the dead-nettle (_Lamium_), the flowers are produced in the axil of each of the foliage leaves of the plant, and they appear as if arranged in a simple whorl of flowers. But on examination it is found that there is a central flower expanding first, and from its axis two secondary axes spring bearing solitary flowers; the expansion is thus centrifugal. The inflorescence is therefore a contracted dichasium, the flowers being sessile, or nearly so, and the clusters are called _verticillasters_ (fig. 17). Sometimes, especially towards the summit of a dichasium, owing to the exhaustion of the growing power of the plant, only one of the bracts gives origin to a new axis, the other remaining empty; thus the inflorescence becomes unilateral, and further development is arrested. In addition to the dichasial form there are others where more than two lateral axes are produced from the primary floral axis, each of which in turn produces numerous axes. To this form the terms _trichasial_ and _polychasial cyme_ have been applied; but these are now usually designated _cymose umbels_. They are well seen in some species of _Euphorbia_. Another term, _anthela_, has been used to distinguish such forms as occur in several species of _Luzula_ and _Juncus_, where numerous lateral axes arising from the primary axis grow very strongly and develop in an irregular manner.

In the uniparous cyme a number of floral axes are successively developed one from the other, but the axis of each successive generation, instead of producing a pair of bracts, produces only one. The basal portion of the consecutive axes may become much thickened and arranged more or less in a straight line, and thus collectively form an apparent or false axis or _sympodium_, and the inflorescence thus simulates a raceme. In the true raceme, however, we find only a single axis, producing in succession a series of bracts, from which the floral peduncles arise as lateral shoots, and thus each flower is on the same side of the floral axis as the bract in the axil of which it is developed; but in the uniparous cyme the flower of each of these axes, the basal portions of which unite to form the false axis, is situated on the opposite side of the axis to the bract from which it apparently arises (fig. 18). The bract is not, however, the one from which the axis terminating in the flower arises, but is a bract produced upon it, and gives origin in its axil to a new axis, the basal portion of which, constituting the next part of the false axis, occupies the angle between this bract and its parent axis--the bract from which the axis really does arise being situated lower down upon the same side of the axis with itself. The uniparous cyme presents two forms, the _scorpioid_ or _cicinal_ and the _helicoid_ or _bostrychoid_.

In the scorpioid cyme the flowers are arranged alternately in a double row along one side of the false axis (fig. 19), the bracts when developed forming a second double row on the opposite side; the whole inflorescence usually curves on itself like a scorpion's tail, hence its name. In fig. 20 is shown a diagrammatic sketch of this arrangement. The false axis, a b c d, is formed by successive generations of unifloral axes, the flowers being arranged along one side alternately and in a double row; had the bracts been developed they would have formed a similar double row on the opposite side of the false axis; the whole inflorescence is represented as curved on itself. The inflorescence in the family Boraginaceae are usually regarded as true scorpioid cymes.

In the helicoid cyme there is also a false axis formed by the basal portion of the separate axes, but the flowers are not placed in a double row, but in a single row, and form a spiral or helix round the false axis. In _Alstroemeria_, as represented in fig. 18, the axis a1 ends in a flower (cut off in the figure) and bears a leaf. From the axil of this leaf, that is, between it and the primary axis a1 arises a secondary axis a2, ending in a flower f2, and producing a leaf about the middle. From the axil of this leaf a tertiary floral axis a3, ending in a flower f3, takes origin. In this case the axes are not arranged in two rows along one side of the false axis, but are placed at regular intervals, so as to form an elongated spiral round it.

Compound definite inflorescences are by no means common, but in _Streptocarpus polyanthus_ and in several calceolarias we probably have examples. Here there are _scorpioid cymes of pairs of flowers_, each pair consisting of an older and a younger flower.

Mixed inflorescence.