CHAPTER XX

FLOWER-CLUSTERS

=Origin of the Flower-cluster.=--We have seen that branches arise from the axils of leaves. Sometimes the leaves may be reduced to bracts and yet branches are borne in their axils. Some of the branches grow into long limbs; others become short spurs; _others bear flowers_. In fact, a flower is itself a specialized branch.

Flowers are usually borne near the top of the plant. Often they are produced in great numbers. It results, therefore, that flower branches usually stand close together, forming a cluster. The shape and arrangement of the flower-cluster _differ with the kind of plant_, since each plant has its own mode of branching.

[Illustration: FIG. 211.--TERMINAL FLOWERS OF THE WHITEWEED (in some places called ox-eye daisy).]

Certain definite or well-marked types of flower-clusters have received names. Some of these names we shall discuss, but the flower-clusters that perfectly match the definitions are the exception rather than the rule. The determining of the kinds of flower-clusters is one of the most perplexing subjects in descriptive botany. We may classify the subject around three ideas: =solitary flowers=, =centrifugal or determinate clusters=, =centripetal or indeterminate clusters=.

=Solitary Flowers.=--In many cases flowers are borne singly; they are separated from other flowers by leaves. They are then said to be =solitary=. The solitary flower may be either at the end of the main shoot or axis (Fig. 211), when it is said to be =terminal=; or from the side of the shoot (Fig. 212), when it is said to be =lateral= or =axillary=.

[Illustration: FIG. 212.--LATERAL FLOWER OF AN ABUTILON. A greenhouse plant.]

=Centripetal Clusters.=--If the flower-bearing axils were rather close together, an open or leafy flower-cluster might result. If the plant continues to grow from the tip, the older flowers are left farther and farther behind. If the cluster were so short as to be flat or convex on top, the outermost flowers would be the older. A flower-cluster in which the lower or outer flowers open first is said to be a =centripetal cluster=. It is sometimes said to be an =indeterminate cluster=, since it is the result of a type of growth which may go on more or less continuously from the apex.

The simplest form of a definite centripetal cluster is a =raceme=, which is an open elongated cluster in which the _flowers are borne singly on very short branches_ and open from below (that is, from the older part of the shoot) upwards (Fig. 213). The raceme may be _terminal_ to the main branch; or it may be _lateral_ to it, as in Fig. 214. Racemes often bear the flowers on one side of the stem, thus forming a single row.

When a centripetal flower-cluster is long and dense and the flowers are sessile or nearly so, it is called a =spike= (Fig. 215). Common examples of spikes are plantain, mignonette, mullein.

[Illustration: FIG. 213.--RACEME OF CURRANT. Terminal or lateral?]

[Illustration: FIG. 214.--LATERAL RACEMES (in fruit) OF BARBERRY.]

[Illustration: FIG. 215.--SPIKE OF PLANTAIN.]

A very _short and dense spike_ is a =head=. Clover (Fig. 216) is a good example. The sunflower and related plants bear many small flowers in a very dense and often flat head. Note that in the sunflower (Fig. 189) the outside or exterior flowers open first. Another special form of spike is the =catkin=, which usually has scaly bracts, the whole cluster being deciduous after flowering or fruiting, and the flowers (in typical cases) having only stamens or pistils. Examples are the “pussies” of willows (Fig. 182) and flower-clusters of oak (Fig. 180), walnuts (Fig. 204), poplars.

[Illustration: FIG. 216.--HEAD OF CLOVER BLOSSOMS.]

[Illustration: FIG. 217.--CORYMB OF CANDY-TUFT.]

When a loose, elongated centripetal flower-cluster has some primary branches simple, and others irregularly branched, it is called a =panicle=. It is a branching raceme. Because of the earlier growth of the lower branches, the panicle is usually broadest at the base or conical in outline. True panicles are not very common.

When an indeterminate flower-cluster is short, so that the _top is convex or flat_, it is a =corymb= (Fig. 217). The outermost flowers open first. Centripetal flower-clusters are sometimes said to be corymbose in mode.

When the branches of an indeterminate cluster _arise from a common point_, like the frame of an umbrella, the cluster is an =umbel= (Fig. 218). Typical umbels occur in carrot, parsnip, caraway and other plants of the parsley family: the family is known as the Umbelliferæ, or umbel-bearing family. In the carrot and many other Umbelliferæ, there are small or secondary umbels, called =umbellets=, at the end of each of the main branches. (In the center of the wild carrot umbel one often finds a single, blackish, often aborted flower, comprising a 1-flowered umbellet.)

[Illustration: FIG. 218.--REMAINS OF A LAST YEAR’S UMBEL OF WILD CARROT.]

=Centrifugal or Determinate Clusters.=--When the terminal or central flower opens first, the cluster is said to be =centrifugal=. The growth of the shoot or cluster is =determinate=, since the length is definitely determined or stopped by the terminal flower. Fig. 219 shows a determinate or centrifugal mode of flower bearing.

Dense centrifugal clusters are usually flattish on top because of the cessation of growth in the main or central axis. These compact flower-clusters are known as =cymes=. Centrifugal clusters are sometimes said to be cymose in mode. Apples, pears (Fig. 220), and elders bear flowers in cymes. Some cyme-forms are like umbels in general appearance. A head-like cymose cluster is a =glomerule=; it blooms from the top downwards rather than from the base upwards.

=Mixed Clusters.=--Often the cluster is mixed, being determinate in one part and indeterminate in another part of the same cluster. The main cluster may be indeterminate, but the branches determinate. The cluster has the appearance of a panicle, and is usually so called, but it is really a =thyrse=. Lilac is a familiar example of a thyrse. In some cases the main cluster is determinate and the branches are indeterminate, as in hydrangea and elder.

[Illustration: FIG. 219.--DETERMINATE OR CYMOSE ARRANGEMENT.--Wild geranium.]

[Illustration: FIG. 220.--CYME OF PEAR. Often imperfect.]

=Inflorescence.=--The mode or method of flower arrangement is known as the =inflorescence=. That is, the inflorescence is cymose, corymbose, paniculate, spicate, solitary, determinate, indeterminate. By custom, however, the word “inflorescence” has come to be used for the _flower-cluster itself_ in works on descriptive botany. Thus a cyme or a panicle may be called an inflorescence. It will be seen that even solitary flowers follow either indeterminate or determinate methods of branching.

[Illustration: FIG. 221.--FORMS OF CENTRIPETAL FLOWER-CLUSTERS.

1, raceme; 2, spike; 3, umbel; 4, head or anthodium; 5, corymb.]

[Illustration: FIG. 222.--CENTRIPETAL INFLORESCENCE, _continued_.

6, spadix; 7, compound umbel; 8, catkin.]

[Illustration: FIG. 223.--CENTRIFUGAL INFLORESCENCE.

1, cyme; 2, scirpioid raceme (or half cyme).]

=The flower-stem.=--The stem of a solitary flower is known as a =peduncle=; also the general stem of a _flower-cluster_. The stem of the individual flower in a cluster is a =pedicel=. In the so-called stemless plants the peduncle may arise directly from the ground, or crown of the plant, as in dandelion, hyacinth, garden daisy; this kind of peduncle is called a =scape=. A scape may bear one or many flowers. It has no foliage leaves, but it may have bracts.

SUGGESTIONS.--=166.= Name six columns in your notebook as follows: spike, raceme, corymb, umbel, cyme, solitary. Write each of the following in its appropriate column: larkspur, grape, rose, wistaria, onion, bridal wreath, banana, hydrangea, phlox, China berry, lily-of-the-valley, Spanish dagger (or yucca), sorghum, tuberose, hyacinth, mustard, goldenrod, peach, hollyhock, mullein, crêpe myrtle, locust, narcissus, snapdragon, peppergrass, shepherd’s purse, coxcomb, wheat, hawthorn, geranium, carrot, elder, millet, dogwood, castor bean; substitute others for plants that do not grow in your region. =167.= In the study of flower-clusters, it is well to choose first those that are fairly typical of the various classes discussed in the preceding paragraphs. As soon as the main types are well fixed in the mind, random clusters should be examined, for the pupil must never receive the impression that all flower-clusters follow the definitions in books. Clusters of some of the commonest plants are very puzzling, but the pupil should at least be able to discover whether the inflorescence is determinate or indeterminate. Figures 221 to 223 (from the German) illustrate the theoretical modes of inflorescence. The numerals indicate the order of opening.