rudimentary; the vesicles are in some species erect, in others decumbent, but in all cases more or less fibrous. (See Plate I. figs. 4. 13, 14, 15. 18, 19, 20.). For an elaborate treatise on the subject, see Joh. Ev. Purkinje de Cellulis Antherarum Fibrosis: Vratislaviæ, 1830, 4to; with eighteen plates. The stamen deviates in a greater degree than any other organ from the structure of the leaf, by a modification of which it is produced; and, at first sight, in many cases, it appears impossible to discover any analogy between the type and its modification; as, for instance, between the stamen and leaf of a Rose. Nevertheless, if we watch the transitions which take place between the several organs in certain species, what was before mysterious, or even inscrutable, becomes clear and intelligible. In Nymphæa alba the petals so gradually change into stamens, that the process may be distinctly seen to depend upon a contraction of the lower half of a petal into the filament, and by a developement of yellow matter within the substance of the upper end of the same petal on each side into pollen. A similar kind of passage from petals to stamens may be found in Calycanthus, Illicium, and many other plants. Now, as no one can doubt that a petal is a modified leaf, it will necessarily follow, from what has been stated, that a stamen is one also. But it is not from parts in their normal state that the best ideas of the real nature of the stamen may be formed; it is rather by parts in a monstrous state, when reverting to the form of that organ from which they were transformed, that we can most correctly judge of the exact nature of the modification. Take for example that wellknown double Rose, called by the French R. Eillet. In that very remarkable variety, the claw of the petals may at all times be found in every degree of gradation from its common state to that of a filament, and the limb sometimes almost of its usual degree of developement, - sometimes contracting into a lobe of the anther on one side, or perhaps on both sides, now having the part that assumes the character of the anther merely yellow, -now polliniferous, and finally acquiring, in many instances, all the characters of an un --- doubted though somewhat distorted stamen. Double Pæonies, Double Tulips, and many other monstrous flowers, particularly of an icosandrous or polyandrous structure, afford equally instructive specimens. It is for these reasons that it is stated in the Outlines of the first Principles of Botany, 307., that "the anther is a modification of the lamina, and the filament of the petiole." I ought, perhaps, to have put the explanation in a more extended form. A leaf consists of a midrib, on each side of which is a parenchymatous expansion, consisting of a double stratum of tissue, separated by vessels. In the anther the midrib assumes the form of the connective; the double stratum on each side of the midrib is, at the centre, developed in the form of pollen, and hence the primitive quadrilocular structure of the anther, as above described. The line of dehiscence in ordinary circumstances is the margin of the modified leaf. Schleiden makes this additional remark: · "The normal leaf, as is well known, exhibits upon its upper surface cellular tissue, different in structure from that on the under; to this we find that the pollen of the anterior and posterior cells of the anther corresponds. It may, perhaps, be possible, and certainly not uninteresting, to ascertain, by experiment, whether or not the pollen of one of these compartments only possesses the external characters✶ of pollen, and likewise different functions in the process of impregnation, or whether in diœcious plants one kind would produce male, the other female embryos." Agardh considers a stamen to be composed of two leaves in a state of adhesion; and that it is in fact a bud axillary to a sepal or petal. This is very nearly the opinion formerly entertained by Wolff. Endlicher adopts this view to a certain extent; and supposes the leaves to be rolled backwards, so that their under surface becomes the polliniferous part. But all this is mere hypothesis, unsupported by evidence, and in opposition to the direct observations of Mirbel and Schleiden. The latter well observes, that the stamens are evidently * It is so expressed in the translation in Taylor's Magazine; in the original it is: "ob vielleicht der pollen einer von beiden, nur der form nach pollen sei, und bei der befruchtung sich verschieden verhalte," u. s. w. modified simple leaves, for they constantly appear at a later period than the petals, although they afterwards develope themselves more rapidly; they stand at first higher up upon the axis than the preceding circle of corollary leaves, and they alternate invariably with them. Such is the structure of the stamens in their perfect state. It often, however, happens that, owing to causes with which we are unacquainted, some of the stamens are developed imperfectly, without the anther and pollen. In such cases they are called sterile stamens (parastemones Link), and are frequently only to be recognised by the position they bear with respect to the other parts of the flower. Botanists consider every appendage, or process, or organ, which forms part of the same series of organs as the true stamens, or which originates between them and the pistil, as stamens, or as belonging to what Röper calls the andrœceum, namely, to the male system; and every thing on the outside of the fertile stamens is in like manner often referred to modifications of petals, a remarkable instance of which is exhibited by Passiflora. There is however no certain rule by which it can be determined whether such bodies belong to the stamens or petals. The appearances assumed by these sterile stamens are often exceedingly curious, and generally very unlike those of the fertile stamens; thus in Canna they are exactly like the petals; in Hamamelis they are oblong fleshy bodies, alternating with the fertile stamens; in Pentapetes they are filiform, and placed between every three fertile stamens; in Zingiberacea they are minute gland-like corpuscles, a very common form (Plate IV. fig. 10. c); in Brodiæa they are bifid petaloid scales; and in Asclepiadacea they undergo yet more remarkable transformations. Dunal calls these sterile stamens lepals (lepala); a term which has not yet been adopted. 9. Of the Pollen. The pollen is the pulverulent substance which fills the cells of the anther. It consists of extremely minute grains, varying in size, and enclosing a fluid containing molecular The pollen-grains are often called granules. matter. To this important part of the organisation of perfect plants, attention has been directed with great care and skill by numerous observers of the first class, among whom are especially to be noticed R. Brown, Ad. Brongniart, Fritzsche, Griffith, Mirbel, Mohl, and Schleiden. From their enquiries, we have arrived at a knowledge of the history of the pollen, notwithstanding its minuteness, from its first secretion to its final destruction, after the important purpose for which it is provided has been attained. The origin of the pollen, according to the testimony of all observers, occurs in the cells of which the anther is composed, and appears to consist in a peculiar organisation of their granular interior. The grains are usually produced in fours, by the bisection of their generating cell in two opposite directions, but are occasionally formed in pairs or singly. In 1831, Brown speaks thus of the evolution of the pollen of Tradescantia virginica. "In the very early stage of the flower bud, while the antheræ are yet colourless, their loculi are filled with minute lenticular grains, having a transparent flat limb, with a slightly convex and minutely granular semiopake disk. This disk is the nucleus of the cell, which probably loses its membrane or limb, and, gradually enlarging, forms in the next stage a grain also lenticular, and which is marked either with only one transparent line, dividing it into two equal parts, or with two lines crossing at right angles, and dividing it into four equal parts. In each of the quadrants a small nucleus is visible: and even where one transparent line only is distinguishable, two nuclei may often be found in each semicircular division. These nuclei may be readily extracted from the containing grain by pressure, and, after separation, retain their original form. In the next stage examined, the greater number of grains consisted of the semicircular divisions already noticed, which had naturally separated, and now contained only one nucleus, which had greatly increased in size. In the succeeding state the grain apparently consisted of the nucleus of the former stage, considerably enlarged, having a regular oval form, a somewhat granular surface, and originally a small nucleus. This oval grain continuing to increase in size, and in the thickness and opacity of its membrane, acquires a pale yellow colour, and is now the perfect grain of pollen." (On Orchid. and Asclep. p. 21.) In 1832, Mirbel examined the developement of pollen in the anther of a Gourd. He states that "when the flower-bud of this plant is about a line in length, each lobe of the anther is entirely composed of cellular tissue, the bladders of which present in general a pentagonal or hexagonal figure more or less regular when cut across. In every cell, without excepting even those which constitute the superficial layer of the lobe, are certain loose particles, of such extreme minuteness that a magnifying power of 500 or 600 diameters is required to examine them satisfactorily. I cannot compare them to anything better than to little transparent bladders, nearly colourless, more or less rounded, and of an equal size. I examined the cells of the lobe of the anther one by one; and I affirm that, at this early period, there is no trace of either the cells of the anther or of the grains of pollen. The whole of the tissue is perfectly uniform. In a flower-bud, but little larger than the first, I remarked on each side of the medial line of the slice a group, consisting of a few bladders, which were rather larger than the others, but otherwise like them. These larger bladders I propose to call pollen-cells, seeing that it is in their inside that the pollen is organised. In flower-buds, from 1 to 2 lines in length, some remarkable changes were observable. The pollen-cells had become larger; their granules were so much multiplied that they were grouped and packed in opaque masses, and wholly filled the cells. These cells and granules together constituted a greyish body, joined to the rest of the tissue by the intervention of a cellular membrane, a sort of integument which, notwithstanding its organic continuity with the surrounding parts, was readily distinguishable; for while the bladders of the surrounding parts lengthened parallel to the plane of the surface, and to the plane of the base of the anther, those of the integument lengthened from the centre to the circumference. In anthers a little further advanced, the sides of the pollen-cells, instead of being thin and dry as they had previously been, acquired a notable thickness, and their substance, gorged with fluid, resembled a colourless jelly. The cellular integument continued to adhere by its |