circumstance with dicotyledons, and forms a distinct approach to the structure of monocotyledons: it occurs in Trapa and Sorocea, in which they are extremely disproportionate. In Cycas they are also rather unequal; but the structure of that plant is essentially dicotyledonous.

The embryo of MONOCOTYLEDONS (Plate VI. fig. 1. B. &c.) is usually a solid, cylindrical, undivided, homogeneous body, slightly conical at each extremity, with no obvious distinction of radicle, plumula, or cotyledons. In germination the upper end swells and remains within the testa (fig. 10. C. b, &c.) ; the lower lengthens, opens, and emits from within one or more radicles; and a thread-like green body is protruded from the upper part of the portion, which is lengthened beyond the testa. Here the portion remaining within the testa is a single cotyledon; that which lengthens, producing radicles from within its point, is the cauliculus and radicle; and the thread-like protruded green body is the plumula. If this is compared with the germination of dicotyledons, an obvious difference will be at once perceived in the manner in which the radicles are produced: in monocotyledons they are emitted from within the substance of the radicular extremity, and are actually sheathed at the base by the lips of the passage through which they protrude; while in dicotyledons they appear at once from the very surface of the radicular extremity, and consequently have no sheath at their base. Upon this difference in economy, Richard proposed to substitute the term Endorhizæ for monocotyledons, and Exorhizæ for dicotyledons. Some consider the former less perfect than the latter; endorhiza being involute, or imperfectly developed, exorhiza evolute, or fully developed. Dumortier adds to these names endophyllous and exophyllous; because the young leaves of monocotyledons are evolved from within a sheath (coleophyllum or coleoptilum), while those of dicotyledons are always naked. The sheath at the base of the radicle of monocotyledons is called the coleorhiza by Mirbel. Another form of monocotyledonous embryo is that of Aroidea and their allies, in which the plumula is not so intimately combined with the embryo as to be undistinguishable, but is indicated externally by a little slit above the base (Plate VI.

fig. 6. B. e), within which it lies until called into developement by germination.

The exceptions to what has been now described ought, like those of dicotyledons, rather to be called remarkable modifications. Much stress has been laid upon them by several writers, who have thought it requisite to give particular names to their parts. To me, however, it appears far more advisable to explain their analogies without the unnecessary creation of new and bad names. In Graminea (Plate VI. fig. 4.) the embryo consists of a lenticular body lying on the outside of the base of the albumen on one side, and covered on its inner face by that body, and on its outer face by the testa: if viewed on the face next the testa, a slit will be observed of the same nature as that in the side of the embryo of Aroidea; opening this cleft a small conical projection is discovered, pointing towards the apex of the seed. If the embryo be then divided vertically through the conical projection, it will be seen that the latter (c) is a sheath including other little scales resembling the rudiments of leaves; that that part of the embryo which lies next the albumen (d), and above the conical body, is solid; and that the lower extremity of the embryo (e) contains within it the indication of an internal radicle, as in other monocotyledons. In this embryo it is to be understood that the conical projection is the plumula; that part of the embryo lying between it and the albumen, a single scutelliform cotyledon; and the lower point of the embryo, the radicle. In wheat there is a second small cotyledon on the outside of the embryo, inserted a little lower down than the scutelliform cotyledon. This last is called scutellum by Gartner, who considered it of the nature of vitellus. The late M. Richard considered the scutelliform cotyledon a particular modification of the radicle, which he called hypoblastus; the plumula a form of cotyledon, called blastus; the anterior occasional cotyledon a peculiar appendage, named epiblastus; and the radicle a protuberance of the cauliculus, called radiculoda. He further, in reference to this peculiar opinion, termed embryos of this description macropodous. In these ideas, however, Richard was manifestly wrong, as is now well known.

From what has been stated, it is apparent that dicoty

ledons are not absolutely characterised by having two cotyledons, nor monocotyledons by having only one. The real distinction between them consists in their endorhizal or exorhizal germination, and in the cotyledons of dicotyledons being opposite or verticillate, while they are in monocotyledons solitary or alternate. Some botanists have, therefore, recommended the substitution of other terms in lieu of those in common use. M. Cassini suggests isodynamous or isobrious for dicotyledons, because their force of developement is equal on both sides; and anisodynamous or anisobrious for monocotyledons, because their force of developement is greater on one side than on the other. Another writer, M. Lestiboudois, would call dicotyledons exoptiles, because their plumula is naked; and monocotyledons endoptiles, because their plumula is enclosed within the cotyledons; but there seems little use in these proposed changes, which are moreover as open to objections as the terms in common use.

The ACOTYLEDONOUS embryo is not exactly, as its name seems to indicate, an embryo without cotyledons; for, in that case, Cuscuta would be acotyledonous. On the contrary, it is an embryo, which does not germinate from two fixed invariable points, namely, the plumula and the radicle, but indifferently from any point of the surface; as in some of the Arum tribe, and in all flowerless plants.

For further illustrations of the embryo, consult Plate VI. and the explanation of its figures.

The direction of the embryo is either absolute or relative. Its absolute direction is that which it has independently of the parts that surround it. In this respect it varies much in different genera; it is either straight (Plate VI. fig. 5.), arcuate (fig. 9.), or falcate, uncinate, or coiled up (fig. 8.) (cyclical), folded up, spiral (fig. 19.), or bent at right angles (Plate V. fig. 28.) (gromonical, Link), serpentine, or in figure like the letter S (sigmoid).

Its relative position is determined by the relation it bears to the chalaza and micropyle of the seed; or, in other words, upon the relation that the integuments, the raphe, chalaza, hilum, micropyle, and radicle bear to each other. If the sacs of the ovulum are in no degree inverted, but have their com

mon point of origin at the hilum, there being (necessarily) neither raphe nor chalaza visible, the radicle will in that case be at the extremity of the seed most remote from the hilum, and the embryo inverted with respect to the seed, as in Cistus, Urtica, and others, where it is said to be antitropal. But if the ovulum undergoes that remarkable extension of one side already described in speaking of that organ, in which the sacs are so inverted that their orifice is next the hilum, and their base at the apex of the ovulum, then there will be a raphe and chalaza distinctly present; and the radicle will, in the seed, be at the end next the hilum, and the embryo will be erect with respect to the seed, or orthotropal, as in the apple, plum, &c. On the other hand, supposing that the sacs of the embryo suffer only a partial degree of inversion, so that their foramen is neither at the one extremity nor the other, there will be a chalaza and a short raphe; and the radicle will point neither to the apex nor to the base of the seed, but the embryo will lie, as it were, across it, or be heterotropal, as is the case in the primrose. When an embryo is so curved as to have both apex and radicle presented to the hilum, as in Reseda, it is amphitropal.

In the words of Gærtner, an embryo is ascending when its apex is pointed to the apex of the fruit; descending, if to the base of the fruit; centripetal, if turned towards the axis of the fruit; and centrifugal, if towards the sides of the fruit; those embryos are called wandering, or vagi, which have no evident direction.

The cotyledons are generally straight, and placed face to face; but there are numberless exceptions to this. Some are separated by the intervention of albumen (Plate VI. fig. 11.); others are naturally distant from each other without any intervening substance. Some are straight, some waved, others arcuate or spiral. When they are folded with their back upon the radicle, they are called incumbent; if their edges are presented to the same part, they are accumbent ; terms chiefly used in speaking of Cruciferæ.

15. Of Naked Seeds.

By naked seeds has been understood, by the school of Linnæus, small seed-like fruit, like that of Labiatæ, Boragineæ, grasses, and Cyperaceæ. But as these are distinctly covered by pericarpia, as has been shown above, the expression in the sense of Linnæus is obviously incorrect, and is now abandoned. Hence it has been inferred that there is no such thing in existence as a naked seed; that is to say, a seed which bears on its own integuments the organ of impregnation. To this proposition botanists had assented till the year 1825, when Mr. Brown demonstrated the existence of seeds strictly naked; that is to say, from their youngest state destitute of pericarpium, and receiving impregnation through their integuments without the intervention of style or stigma, or any stigmatic apparatus. That learned botanist has demonstrated that seeds of this description are uniform in Coniferæ and Cycadeæ, in which no pericarpial covering exists. But we have no knowledge at present of such an economy obtaining in other plants, as a constant character. It does however happen, as the same observer has pointed out, that in particular species the ovarium is ruptured at an early period by the ovula, which thus, when ripe, become truly naked seeds; remarkable instances of which occur in Ophiopogon spicatus, Leontice thalictroides, and Peliosanthes Teta.