Corollas with six, eight, nine, ten, or twelve petals are relatively rarer; and when the petals become so numerous that we cannot count them, we have to deal with transformations of stamens into petals, with those monstrosities of cultivation which we call double flowers, flores pleni, where all the male organs have disappeared,—flowers wholly unfit for fructification.
The petals of the corolla are not always free. Like those of the calyx, they may be attached to one another by their edges, but this union invariably takes place, as in the former, from bottom to top. Therefore, we never see any petals united at the top, and disengaged at the bottom (see Fig. 47, a.) But the reader must take careful note that this invariable characteristic is not peculiar only to the floral envelopes; it is not met with in the stamens, for these may be united, either by their anthers, as in the whole family of the Synantheræ, or by their filaments, as in the Leguminosæ. And what we have said of the stamens applies also to the parts constituting the pistil. This radical difference between the perianth and the true reproductive organs ought, from the beginning, to have fixed the attention of botanists on the centripetal and centrifugal metamorphosis of the leaf, which we have spoken of in "The Circle of the Year."
Fig. 47, a.—Natural junction of Petal.
Fig. 47, b.—Unnatural (and impossible) junction.
In many plants we are permitted to follow step by step, as it were, the union of the petals, and their definitive transformation into what is called the monopetalous corolla. The term monopetaloid ought then to be rejected, if we are to believe that the monopetalous corolla is the result of the metamorphosis of a single leaf. The word gamopetalous, or, rather, gamophyllous, is preferable. As we have remarked in reference to the calyx, we shall here repeat that the expressions bilobed, trilobed, quadrilobed, or bipartite, tripartite, quadripartite corollas, are radically vicious, because they are the consequence of a false point of view, according to which the monopetalous corolla will be simply a single metamorphosed leaf, susceptible of being more or less deeply divided from top to bottom.
The polypetalous corolla, as well as the gamophyllous corolla, may be regular or irregular, according as the foliola are equally or unequally united. But here again we must be careful not to lay down too absolute rules. Examine, for instance, the gamophyllous corolla of the gentians; their two lobes are very unequal, and yet the corolla is regular: five very large lobes alternate with five very small, in such wise that each of the latter is situated between two larger lobes. Each division of the corolla of the periwinkle is irregular, and yet the aggregate of their divisions is regular: these are all of the same form and the same size.
In the gamophyllous corolla we are able to discern, through the different forms it assumes, the form of the parts which compose the union. Thus, for example, the tubulate corolla supposes the pre-existence of unguiculate petals.
Inequality of union produces the bilabiate corolla, which is invariably tubular. This is the case in the natural family of the Labiatæ. The upper lip is composed of two petals, and the lower of three. The parts of the upper are sometimes so closely joined that they appear to be but one, as in the Lamium; and the lower often becomes quadrilobed by the division of the middle petal, as in the Stæcha.