be more than one influence at work to determine what whorl shall follow each of those of the perianth.

The immediate cause is nutrition; but the deeper question, what directs the nutrition to one cord rather than another, can only be guessed at in most cases: but as the petaline stamens are generally absent from at least the gamopetala, it would seem that the enhancement of the corolla through the agency of insects has caused the whorl of stamens in front of it to be atrophied through compensation. Some special circumstance, however, we know not what, have interfered to retain that whorl in Primulacea, and some few other plants.

The reader must be reminded, however, that this method of branching in order to give rise to stamens and carpels from the cords of the perianth is not universal. When they are many, it is done by the fibro-vascular cylinder of the pedicel becoming much enlarged, and consisting of a great number of cords, all arising by lateral chorisis, it is true, but long before they enter the floral members; so that by the time the latter are about to emerge they each receive their own cords from the general axial cylinder. This is what happens e.g., in Ranunculacea and Cruciferæ.

CHAPTER V.

THE PRINCIPLE OF COHESION.

COHESION.-GENERAL OBSERVATIONS. This term signifies the union between parts of the same kind or whorl; and the prefix gamo- is used in conjunction with the terminations -sepalous, -petalous, and -phyllous,—to indicate that the parts of the calyx, corolla, and perianth respectively cohere. In the case of the stamens, they are said to be mon-, di-, tri-, or poly-adelphous, according as the filaments cohere into one, two, three, or more groups; while syngenesious is used for the coherence of anthers, and, lastly, syncarpous denotes that the carpels of a pistil cohere.

There are two kinds of cohesion, congenital and by contact.* Congenital cohesion I regard as an advance upon freedom, or a further state of differentiation; for, according to the principles of Evolution, freedom or separation of parts must precede their union; just as, for example, bones are free in the embryo which become "ankylosed" in the adult; or always free in a fish, while their homologues cohere in higher types of vertebrates.

Congenital cohesion applies to by far the greater number of cases of union amongst the parts of the different whorls

*We might appropriately distinguish these two kinds of union by the terms connate or "born together," and coherent or "sticking together."

of flowers, respectively. Cohesion by contact is the cause of the anthers being syngenesious in the Compositæ. It applies, sometimes at least, to the two margins of each carpel when in contact up the axis of an ovary, as of that of a Lily. The stigmas of Asclepias are at first free, but later in their development they become coherent by contact.

Congenital cohesion takes place almost from the very commencement of growth and development of the parts, so that when full-grown there may be no trace of the line of cohesion. Fibro-vascular cords, indeed, often occur in the very position of it, not unfrequently branching off in various ways, as, e.g., at the fork to nourish the adjacent free portions of the limb. This occurs in the calyx of Stachys and the corolla of Primula, etc. In Campanula rotundifolia the fibrovascular system of the corolla becomes completely altered, and instead of representing that of distinct leaves in contact by their edges, the veins ramify and anastomose all over the general space between the two adjacent dorsal ribs, completely obliterating all trace of the line of union between them. In the case of the Primrose, however, the calyx has the exact appearance of five pinnately nerved leaves being united by their thin and impoverished edges, where there is nothing but translucent tissue without any cords at all.

It is important to observe this more or less complete modification of the fibro-vascular system under congenital cohesion, as it shows how much more highly differentiated a condition has been acquired than when the parts are free. In the latter case they represent more closely the forms and venation of distinct foliar organs.

As a curious instance of cohesion of both kinds in the same organ, may be mentioned the corolla of Phyteuma; the basal portion of which consists of five petals congenitally united; but the five portions of the limb cohere by contact

E

at the apex, and so form a tube which collects the pollen shed into it by the five free anthers, which are included

within this corolla-tube (Fig. 9). They thus form the "cylinder" for the "piston" action of the pistil which continues to grow, and so sweeps out the pollen beyond the extremity of the tube, just as it does from the syngenesious anthers of the Composite and Lobelia. The five portions of the corolla thus cohering by contact subsequently become more or less free.

The rationale of Cohesion lies in its adaptation to insect agency, and Fig. 9.—Phyteuma (after Müller). implies a greater degree of specialization than when the parts of the whorls are free. Thus in Thalamiflora, of such an order as Ranunculaceae with regular flowers and with all the parts of the perianth whorls free, the flowers are usually visited by a much greater number and variety of insects than are those of orders of Corolliflora. For example, Müller records sixty-two species of insects as seen by him to visit Ranunculus acris; whereas the humble-bee alone enters the gamopetalous tube of the Foxglove. This adaptation of form to insect visitors will be better appreciated when we come to discuss that principle of Variation, which so powerfully affects floral structure.

It occasionally happens that parts normally united become free: the process is called "dialysis," and may be regarded as a reversion to an ancestral free condition. Fig. 10 represents a flower of Mimulus in this condition. The rationale of cohesion in the sepals, petals, and stamens, I regard as the immediate result of hypertrophy set up by insect agency,

aided by the close proximity of the parts; and as a resulting effect, is the ever-increasing adaptation to the requirements of insects, which are more and more specialized for them, so that, for example, Lepidoptera are almost solely adapted to long tubular flowers like the Honeysuckle.

Fig. 10.-Mimulus undergoing "Dialysis" (after Baillon).

An analogous process of congenital cohesion is well seen in the fasciation of stems which occurs particularly often in succulent shoots, as Asparagus, Cabbage, Lettuce, and the young shoots of the Ash tree. This is most reasonably referred to hypertrophy coupled with the close proximity of the buds which ought to have developed into independent shoots. Again, cohesion between the sepals or petals of Orchids is not uncommon abnormally under cultivation; and would also seem to be due to the stimulating conditions under which they are artificially cultivated.

Hypertrophy in an organ is due to a special flow of nutriment to it; and cohesion may result from the close proximity of the parts of the whorl to one another; but the influence which brings about the determination of sap to a particular point, I take to be the mechanical strains induced by the insect visitors when alighting upon the flower in search for nectar or pollen.

If this principle be correct, that the tubular structure of calyces and corollas, as we see them now, has arisen through the requirements of those organs to meet strains thrown upon them; I think it will furnish the solution to many a question that may arise as to the peculiar shapes of corollas, etc., besides explaining the very principle of cohesion itself. An