when they are not visible. Usually it is nearly as long as the secundine, but sometimes it is remarkably shorter, as in Euphorbia Lathyris when very young (Plate V. fig. 22.).

The outermost but one of the sacs (Plate V. fig. 23. b, 20. b, 25. d) is called the secundine; it immediately reposes upon the primine, and often contracts an adhesion with it, so that the two integuments become confounded. In order to ascertain its existence, it is, therefore, often necessary to examine the ovule at a very early period of its growth. Myrica, Alnus, Corylus, Quercus, and Juglans have been named by Mirbel as plants in which the secundine is not perceptible (Plate V. fig. 24.). Its point is usually protruded beyond the foramen of the primine.

The nucleus (Plate V. fig. 22. b, 18, 19, 20. a, 24. d, 25. e) is a pulpy conical mass, enclosed by the primine and secundine, and often covered by them; but frequently protruded beyond the latter, and afterwards, at a subsequent period of its growth, again covered by them. Sometimes its epidermis is said to separate in the form of a third coating called the tercine.

These three parts, the primine, the secundine, and the nucleus, have all an organic connection at some one point of their surface. That point is, in ovules whose parts do not undergo any alteration of direction in the course of their growth, at the base next the placenta; so that the nucleus is like a cone, growing from the base of a cup, the base of which is connected with the hilum through another cup like itself (Plate V. fig. 23.). The axis of such an ovule, which Mirbel calls orthotropous, is rectilinear, as in Myrica, Cistus, Urtica, &c.; and the foramen is at the end of the ovule most remote from the hilum.

But sometimes, while the base of the nucleus and that of the outer sacs continue contiguous to the hilum, the axis of the ovule, instead of remaining rectilinear, is curved down upon itself (Plate V. fig. 26, 27.); so that the foramen, instead of being at the extremity of the ovule most remote from the hilum, is brought almost into contact with it. Examples of this are found in caryophylleous plants, Mignionette, &c. Mirbel, who first distinguished these ovules, calls them campy

lotropous. In both these modifications, the base of the ovule and the base of the nucleus are the same.

In a third class the axis of the ovule remains rectilinear; but one of the sides grows rapidly, while the opposite side does not grow at all, so that the point of the ovule is gradually pushed round to the base; while the base of the nucleus is removed from the hilum to the opposite extremity (Plate V. fig. 16-21.): and when this process is completed the whole of the inside of the ovule is reversed; so that the apex of the nucleus, and consequently the foramen, correspond with the base of the ovule. Such ovules as these Mirbel terms anatropous; they are very common: examples may be found in the Almond, the Apple, the Ranunculus, the Cucumber, &c. When the base of the nucleus is thus removed from the base of the ovule, a communication between the two is always maintained by means of a vascular cord, called the raphe (Plate V. fig. 24. e, 25. f). This raphe, which originates in the placenta, runs up one side of the ovule, until it reaches the base of the nucleus: and there it expands into a sort of vascular disk, which is called the chalaza (Plate V. fig. 24. f. 25. g). As the chalaza is uniformly at the base of the nucleus, it will follow that, in orthotropous and campylotropous ovules, it is confounded with the hilum; while it is only distinguished in anatropous ones, in which alone it is distinctly to be recognised.

In addition to these there is the amphitropous ovule, whose foraminal and chalazal ends are transverse with respect to the hilum, which is connected with the latter by a short raphe; and the semianatropous, which is only different from the last, in the ovule being parallel with the funiculus instead of being at right angles with it.

The following figures give a comparative plan of these ovules.

a, Orthotropous, or atropous; b, campylotropous; c, anatropous; d, amphitropous; e, semianatropous. In these figures represents the chalaza, and the foramen,

It has been remarked that the raphe or vascular extension of the placenta always occupies the side next the ventral suture of the ovary; and that when, as in Euonymus, it is turned towards the dorsal suture, that circumstance arises from an alteration in the position of the ovule subsequent to its being fertilised.

It has also been stated that the passage through the primine and secundine is called the foramen; or the exostome, when speaking of that of the primine; and the endostome, in speaking of the secundine. Upon these Mirbel remarks:"These two orifices are at first very minute, but they gradually enlarge; and, when they have arrived at the maximum of dilatation they can attain, they contract and close up. This maximum of dilatation is so considerable in a great number of species, in proportion to the size of the ovule, that, to give an exact idea of it, I would compare it not to a hole, as those express themselves who have hitherto spoken of the exostome and endostome, but to the mouth of a goblet or of a cup. It may therefore be easily understood, that, to perceive either the secundine or the nucleus, it is not necessary to have recourse to anatomy. I have often seen, most distinctly, the primine and secundine forming two large cups, one of which encompassed the other without entirely covering it, and the nucleus extending itself in the form of an elongated cone beyond the secundine, to the bottom of which its base was fixed."

In practical botany the detection of the foramen is often a matter of great importance; for it enables an observer to judge from the ovule of the direction of the radicle of the future embryo: it having been ascertained by many observations that the radicle of the embryo is almost always pointed to the foramen. A partial exception to this law exists, however, in Euphorbiaceæ, in many of which Mirbel has noticed that, after fertilisation, the axis of the nucleus and the endostome are inclined five or six degrees, without the exostome changing its position; by this circumstance the foramen of the secundine and that of the primine cease to correspond, and the radicle, instead of pointing when formed to the exostome, is directed to a point a short distance on one side of it.

Besides the two external integuments, Mirbel has remarked the occasional presence of three others peculiar to the nucleus, which he calls the tercine, quartine, and quintine.

The former is the external coat of the nucleus, and is very generally, if not universally, present. As I am unacquainted with the distinctions between these supposed integuments, especially the quartine, I can add nothing to the following remarks of Mirbel upon the subject:-"The quartine and quintine are productions slower to show themselves than the preceding. The quartine is not very rare, although no one has previously indicated it; as to the quintine, which is the vesicula amnios of Malpighi, the additional membrane of Brown, and the sac of the embryo of Adolphe Brongniart, I am far from thinking that it only exists in a small number of species, as Brown seems to suppose. If no one has noticed the quartine, it is, no doubt, because it has been confounded with the tercine; nevertheless these two envelopes differ essentially in their origin and mode of growth. I have only discovered the quartine in ovules of which the tercine is incorporated at an early period with the secundine; and I think that it is only in such cases that it exists. At its first appearance it forms a cellular plate, which lines all the internal surface of the wall of the cavity of the ovule; at a later period it separates from the wall, and only adheres to the summit of the cavity: at this period it is a sac, or rather a perfectly close vesicle. Sometimes it rests finally in this state, as in Statice; in other cases it fills with cellular tissue, and become a pulpy mass; under this aspect it is seen in Tulipa Gesneriana. All this is the reverse of what takes place in the tercine; for this third envelope always begins by being a mass of cellular tissue, (and at that time it has the name, as we have seen, of nucleus,) and generally finishes by becoming a vesicle.

"I have remarked the fifth envelope, or quintine, in many species: its general characters are such as to prevent its being mistaken. Its complete developement takes place only in a nucleus which remains full of cellular tissue, or in a quartine that has filled with the same. At the centre of the tissue is organised, as in a womb, the first rudiment of the quintine;

it is a sort of delicate intestine, which holds by one end to the summit of the nucleus, and by the other end to the chalaza. The quintine swells from top to bottom; it forces back on all sides the tissue that surrounds it, and it often even invades the place occupied by the quartine or the nucleus. A very delicate thread, the suspensor (hypostasis of Dutrochet), descends from the summit of the ovule into the quintine, and bears at its extremity a globule which is the nascent embryo."

It is apparently this suspensor that Brown describes, in the ovule of Orchidaceæ, as a thread consisting of a simple series of short cells, the lowermost joint or cell of which is probably the original state of what afterwards, from enlargement and deposit of granular matter, becomes the opaque speck, or rudiment of the future embryo. (Observ. on the Organs, &c., of Orch. and Asclepiad. pp. 18, 19.) For further information concerning the suspensor, see Mr. Griffith's observations in the chapter on FERTILISATION in Book II.

"The existence," continues Mirbel, "of a cavity in the quartine, or, indeed, the destruction of the internal tissue of the nucleus, at the period when the quintine developes, becomes the cause of some modifications in the manner of existence of this latter integument. The quintine is never seen, in certain Cucurbitaceæ, adhering to the chalaza: it is nevertheless evident that the adhesion has existed. The quintine, distended at its upper part, and suspended like a lustre from the top of the cavity, still presents at its lower end a portion of a rudimentary intestine become distinct; the separation having occurred very early, in consequence of the tearing of the tissue of the nucleus.

"The quintine of Statice is reduced to a sort of cellular placenta, to the lower surface of which the embryo is attached. This abortion of the quintine arises from the quartine having a large internal cavity, which prevents the young quintine from placing itself in communication with the chalaza, and taking that developement which it acquires in a multitude of other species."

I have continued in this edition to quote the preceding statements of Mirbel; but I have great doubt of their accuracy in some respects, and Schleiden asserts that Mirbel has not