contained perfect stigmata; in it the style branched into two capitate arms, pubescent externally, and in all respects analogous to the stigmata of the former plant; the ovary was however in so very young a state, that I could not detect any concomitant character in the ovules; the anthers were decidedly abortive. In F. sedifolia, L., I have seen no other stigmata than two small uncinate fleshy bodies, concealed between the two upper valves of the anthers, parallel with them, and alternating with two small glands ? at the back of these organs. In form and situation they answer to the plumose stigmas of the two former, but they are smooth throughout. In another flower I find the apex of the style to be depressed and to appear minutely 4-lobed, with the lobes unequal and rounded: in both these cases the anthers were full of pollen, and the ovules in a rudimentary state. In Swartz's description of this plant he notices a crest of projecting hairs, arising from a fleshy septum, obscurely lobed under the microscope, which connects the two anthers and separates the two lobes of the true apex of the style or stigmas. Swartz distinctly alludes to the two small glands or stigmata as being protected by the upper valves of the anthers, and they are hence probably analogous to two of the four lobes into which, in the flower I examined, the apex of the style appeared to be divided. Swartz's supposition, that the septum and crista of fine hairs form a connectivum between the anthers, separating the stigmata, appears to me to indicate a most anomalous condition of those parts; and as it is, from its position and structure, analogous to the arms of the style and stigmata in the two former species, I conclude that that author examined fertile flowers of F. sedifolia. It is still more remarkable that so acute and very accurate an observer should have been unable to detect the glands at the base of the column, which in both my specimens are exceedingly large, and project upwards like two horns from the top of the ovarium for half the length and upwards of the tube of the corolla, and whose apices in the young state of the flower lie between the anthers. It is possible that they may be obscurely developed in fertile flowers of this species, which however is not the case in those of F. clavigera or of Phyllachne uliginosa.
In F. clavigera there are apparently two very different states of the corolla: in many of my specimens of this plant that organ is divided into 5–7 lobes, all of them concave and even, of the same thickness throughout; more rarely they are 4 or 9; but in other corollas taken from the same specimens the divisions are undulated, with the borders of the sinuses much thickened, and each of them furnished at the throat with two linear, elevated, divaricating ridges or glands, which branch off from the middle nerve in the upper part of the tube, and are abruptly clavate at the extremity, near the margin of the segment, with whose thickened margins they sometimes unite. In some respects they resemble the nectaries of Ranunculus pinguis (Tab. I.), being only occasionally present; they however contain no secretion. Though I could trace no connection between this, the common form of the corolla, and the fertile or abortive state of the ovarium, I may remark, that where the segments are smooth and even, the apex of the style is hardly prominent or visible between the anthers, and also that in the most divided corollas the segments were most undulated and thickened; in F. sedifolia they are also very distinct, though nowhere described that I am aware of; and they are also evident, but not so fully developed, in the few flowers of Phyllachne which I have examined. I have also described the corolla as somewhat two-lipped, a character not very evident in all instances, and depending upon the inequality and comparative size of the segments; one or two are almost invariably larger than the rest, and external in æstivation; when there are two large lobes they are placed near one another; and when the corolla has more than five segments, these two are subdivided into four by short sinuses; where only four segments exist, it is caused by the union of two of the small lobes.
All the species have the anthers spuriously 2–celled, by means of a thick fleshy ridge which runs at the base of the anther, between the valves, and projects half-way across the cavity. After the dehiscence of these organs, they together form a cross placed horizontally on the top of the column, from their unsymmetrical contraction; of these, the lower one on each side projects horizontally and forms a right angle with the axis of the column, its two lower lobes approximating below; the upper becomes erect, and its upper margin being revolute, meets that of the opposite anther; this appearance is represented at fig. 10. The ovary, which is generally
