live under water. Its usual character is that of a delicate membrane, but in some plants it is so hard as almost to resist the blade of a knife, as in the pseudo-bulbs of certain Orchidaceous plants. The most usual form of its reticulations is the hexagonal (Plate III. fig. 11.): sometimes they are exceedingly uncertain in figure; often prismatical; and not unfrequently bounded by sinuous lines, so irregular in their direction as to give the meshes no determinate figure (fig. 5.).

Botanists were formerly not agreed upon the exact nature of the epidermis; while some inclined to the opinion that it is an external layer of cellular tissue in a compressed state, others, among whom were included both Kieser and Amici, considered it a membrane of a peculiar nature, transversed by veins, or vasa lymphatica.

By the latter it was contended, that the sinuous direction of the lines in many kinds of epidermis is incompatible with the idea of any thing formed by the adhesion of cellular tissue; that when it is once removed, the subjacent tissue dies, and does not become epidermis in its turn, and that it may often be torn up readily without laceration.

On the other hand, it was replied, that the reticulations of the epidermis are mostly of some figure analogous to that of cellular tissue, and that the sinuous meshes themselves are not so different as to be incompatible with the idea of a membrane formed of adhering bladders. We are accustomed to see so much variety in the mere form of all parts of plants, that an anomalous configuration in cellular tissue should not surprise

us.

The lines, or supposed lymphatic vessels, are nothing more than the united sides of the bladders, and are altogether the same as are presented to the eye by any section of a mass of cellular substance. It is certain that the epidermis cannot be removed without lacerating the subjacent tissue, with however much facility it may be sometimes separable: on the under surface of the leaf of the Box, for instance, there has plainly been some tearing of the tissue, before the epidermis acquired the loose state in which it is finally found.

There is now no anatomist to be found who doubts the fact of epidermis being cellular tissue. In many plants the cel

E

lular state is distinctly visible upon a section (Plate I. fig. 2. a); it even consists occasionally of several layers of vesicles, as in the Oleander and many Orchidaceæ, and it varies in the density, form, and arrangement of its component cells in different plants, according to the peculiar conditions to which they are exposed.

External to the epidermis is a thin homogeneous membrane, formed of organic mucus (see page 1.) and overlying every part except the stomates and the stigmatic tissue. It was first observed by Adolphe Brongniart in the Cabbage-leaf, afterwards by Henslow in Digitalis, and by myself in Dionæa ; it has subsequently been the subject of more extended observations, and appears to be a universal coating, which is even drawn over the hairs, as if to protect the tender cell forming their interior, and the plexus of capillary Cinenchyma, which is stationed on the outside of the walls of that cell. I have found this cuticular membrane on the delicate petals of Hydrotania Meleagris, from which it may be easily removed after maceration for a few days in spirit of wine; and Ad. Brongniart succeeded in separating it from the leaves of Potamogoton lucens, after very long maceration in water. It is stated to be sometimes covered with a minute granular appearance, the nature of which is unknown, and which is not found at the lines indicating the place where the cuticle was pressed upon the united sides of cells. There are some good observations upon this subject by M. Adolphe Brongniart (Ann. des Sc. 2 ser. 1. 65.), who finds the cuticle by no means uncommon; and imagines that it overlies the stigma in Nymphæa and Mirabilis. It certainly does not cover the stomates, nor the glands found on the surface of the inside of the pitchers of Nepenthes.

In most plants the cuticle has certain openings of a very peculiar character, which appear connected with respiration, and which are called Stomates, Stomata, or Stomatia. (Plate III. passim.)

STOMATES are passages through the cuticle, having the appearance of an oval space, in the centre of which is a slit

opens or closes according to circumstances, and lies above a cavity in the subjacent tissue.

There is, perhaps, nothing in the structure of plants upon which more different opinions have been formed than these stomates. Malpighi and Grew, the latter of whom seems first to have figured them (t. 48., fig. 4.), call them openings or apertures, but had no exact idea of their structure. Mirbel also, for a long time, considered them pores, and figured them as such; admitting, however, that he suspected the openings to be an optical deception. De Candolle entertains no doubt of their being passages through the epidermis. He says their edge has the appearance of a kind of oval sphincter, capable of opening and shutting. The membrane that surrounds this sphincter is always continuous with that which constitutes the network of the epidermis: under the latter, and in the interval between the pore and the edge of the sphincter, are often found molecules of adhesive green matter (Organogr. i. 80.); and recently Adolphe Brongniart, in his beautiful figures of the anatomy of leaves, would seem to have settled

the question beyond all dispute. (Annales des Sciences, vol. xxi.) Nevertheless there are anatomists of high reputation who entertain a directly opposite opinion; denying the existence of passages, and considering the stomates rather in the light of glands. Nees von Esenbeck and Link have denied the existence of any perforation in the stomates, and considered that the supposed opening is a space more pellucid than the surrounding tissue, and that what seems a closed up slit is the thickened border of the space. Link further added in his Elementa (ed. 1. p. 225.), that the obscuration of the centre of the stomates is caused by a peculiar secretion of matter, as is plainly visible in Baryosma serratum. To the views of these writers is to be added the testimony of Brown (Suppl. prim. Prodr. p. 1.), who describes the stomates as glands which are really almost always imperforate, with a disk formed by a membrane of greater or less opaqueness, and even occasionally coloured; at the same time he speaks of the disk being, perhaps, sometimes perforated. Link, however, has now abandoned his first idea (Elementa, ed. 2., vol. ii., p. 6.), recognising them as openings; and most anatomists have come to the same conclusion.

In no plants are stomates larger than in some Monocotyledons; they are, therefore, the best subjects for examination for general purposes. In Crinum amabile they evidently consist of two kidney-shaped bodies filled with green matter, lying in an area of the cuticle smaller than those that surround it, and having their incurved sides next each other. In some, at the part where the kidney-shaped bodies come in contact, there is an elevated ridge, dark, as if filled with air, and having its principal diameter distinctly divided by a line. (Plate III. fig. 11.) In this state the stomate is at rest: but in others the kidney-shaped bodies are much more curved; their sides are more separated from each other; and there is no elevated ridge: at their former line of contact there is an opening so distinct and wide as to be equal to half the diameter of one of the kidney-shaped bodies.

This structure of the stomate in Crinum amabile may be taken as the type of all others; for, no doubt, they are all constructed upon a similar plan, though modified in different

species. That is to say they are composed of a pair of cells placed side by side, communicating freely with a hollow chamber in the parenchyma of the leaf built up of cells, arranged in various ways. (Fig. 14. c represents the appearance of the stomate in Acrostichum alcicorne when cut through perpendicularly; figs. d and g show it in the seed-coat of Cannas, and fig. f is the appearance of the same stomate seen from above; all these are copied from Dr. Schleiden's figures.) It is not, however, always two cells which by lying side by side form the stomate; occasionally a greater number is present; as in Marchantia, where, according to Mirbel, they are minute funnels in the epidermis, composed of four or five vesicles arranged circularly in several tiers; at the bottom of this funnel is a large square aperture, communicating with a subjacent chamber, and caused either by the destruction of a central vesicle, or by the separation of the sides of four or five vesicles at the angles next the centre of the funnel.

Several varieties are represented at Plate III.; besides which, stomates have been noticed by Link to be occasionally quadrangular, as in Yucca gloriosa (Plate III. fig. 10.), and Agave americana, and by Brown to be very rarely angular, of which, however, no instance is cited by that botanist. The former case is one in which the quadrangular figure is caused by the cellules of the opening being straight, and bounded by four other cells which appear to be inside the areolations of the cuticle. I have never been so fortunate as to discover the membrane which this great observer describes as generally overlying the apertures; nor do I know of any other botanist having confirmed that observation. It cannot be the cuticle already described, because it has been found that that part never overlies the stomates (see page 50.).

Nerium oleander, and some other plants have, in lieu of stomates, cavities in the cuticle, curiously filled up or protected by hairs. (See Annales des Sciences, xxi. 438.)

In Nepenthes there are stomates of two kinds, the one oblong, semi-transparent, and almost colourless, with numerous pellucid globules in the cavity of the cells; the other roundish, much more opaque, and coloured red. The latter do not communicate immediately with internal cavities in the paren