diate consequences of this theory of abortion, we see in it, first, an explanation of a multitude of anomalies in the number of the parts of plants; secondly, of many, perhaps all, the inequalities of proportion in similar parts; thirdly, of the changes of form, and consequently of use, so frequent in organization, and incomprehensible without this theory.

The next source of error to be considered, is the adhesion or engrafting of organs. Every body knows that a bud or shoot, placed upon another tree under certain conditions, is united to it in such a manner as to form a part of it and grow as if it were on its own stem. Every body knows, too, that in forests we find trees of the same or analogous species, which having been accidentally approximated, are united together so as to form but one trunk, and many have observed that certain organs of plants, that have been brought near one another, are united in a most intimate manner; that two neighboring flowers may be so united as to form but one, having a double number of parts, and that two leaves may also adhere together, so as to form but one of a singular shape. So long as these adhesions take place rarely, they are considered, and justly too, as simple accidents, and no importance is attached to them in classification. But let us suppose that two ovaries, for instance, stand very close to each other from their origin, as in the case of the Pigeon-berry, (Mitchella Repens); it is clear, that by reason of this approximation, the opportunity of coalescing is so great, that union will always take place and we shall never see them separate. Now, this adhesion is nothing more than an accident, but it is one which is determined by causes belonging to organization, and as constant as the organ itself, insomuch that we have what may be called a constant accident, and though these terms seem contradictory, this kind of phenomenon is still very common in

nature.

Not only may similar organs be primarily disposed in such a manner as not to be able to grow without adhering together, but the same thing takes place in different organs; and it is remarkable, that while this phenomenon has been recognised under certain circumstances, it has, in analogous cases, been entirely overlooked or denied. Any organ, a calyx or corolla, for instance, may be described in two ways; either analytically, by considering it as an unique whole divided into parts more or less distinct, or synthetically, as an aggregate of parts essentially distinct, but more or less approx

imated or united. In the first method, we are bound to render an explanation of the causes and laws of the separation of the parts; in the second, to give a similar explanation as to their approximation or union. Both methods involve some hypothetical considerations, and yet, we must follow one or the other. If we are describing a Hollyhock, we must either regard the corolla as an unique whole, divided into several portions called petals, or the petals as distinct organs, which by their union form the corolla. Each of these modes of reasoning may possibly have some good foundation, but certainly it cannot be right to adopt one in the case of the Hollyhock, and the other, when treating of a different flower. We must be consistent, and a method being once admitted, it must be adhered to in all analogous cases. The phenomena of crystallization, to borrow an illustration from a neighboring science, were explained by Rome de l'Isle, by considering crystals as integral bodies, which, in consequence of different truncations, assume all the secondary forms. The Abbé Hauy, on the contrary, explained the same facts, by supposing primitive molecules, which, aggregating after particular laws, determine all the secondary forms. Either theory may be adopted, though the former is now abandoned; but what would be thought of a mineralogist, who should describe one crystal after Rome de l'Isle's method, and another, after Hauy's? And yet, such is the state of botany, that this is constantly done in regard to that science. It becomes, therefore, a matter of serious inquiry, which of these two methods best expresses the whole of the facts, and whether there be cases where they may be blended together. When we speak of the perfoliated leaves of the Honeysuckle, the idea meant to be conveyed is, that an unique or orbicular leaf is traversed or enfiladed by the stem that bears it, yet no one at the present day hesitates to consider this pretended perfoliated leaf as composed of two opposite leaves united at their base. In precisely similar cases we use the term connate leaves, which expresses nearly enough the idea of union; we follow all its degrees from the slightest to the most intimate kind, and when we perceive an interval towards the point of junction, we still consider it as two leaves imperfectly united, not as an unique leaf deeply gashed. The reason is, that at the base of the plant the two opposite leaves are separate and distinct, and that as we approach the summit, they tend more and more to be united; that we find

each half of the perfoliate leaf unique in appearance, and possessing all the organization of one of the inferior leaves. Thus, though the phenomenon is constant, no one hesitates to consider it as a kind of accident, determined by the organization itself.

The law here recognised is applicable to every case of connate leaves, and we must admit the general conclusion, that as leaves may adhere together accidentally, there are cases in which this phenomenon occurs constantly, in consequence of their nature and position. All that has been said of leaves must be readily admitted of stipules, which resemble them so closely; so that when we see all the Leguminosae having a stipule on each side of the petiole, we may conceive that, if these two stipules should be so large as to touch on the side farthest from the petiole, they might be united, and consequently assume the appearance of an unique stipule opposite the leaf. The involucres, too, are subject to the same law of adhesion, as might readily be supposed from analogy, since these organs are now universally regarded as only assemblages of floral leaves. In the Umbelliferae, the involucre generally consists of a certain number of whorled and separate leaflets, but in some species of this order, there is found instead of this whorl, a leafy disk, presenting as many teeth and furrows, as there are leaflets in the neighboring species. We are therefore constrained to regard this disk as formed by the natural union, more or less complete, of many leaflets, and not as a single-leaved involucre. If then the leaves and involucres be so readily regarded as subject to this law of adhesion, of the union of several distinct parts into one,-why should not the fact of its operation be admitted in regard to the calyx ? This organ resembles the involucre in every respect; the anatomy of the sepals shows that they are entirely leafy organs; they are green and decompose carbonic acid like the leaves; they are almost always furnished with the same hairs, glands, and sacks as the true leaves, and finally, in a multitude of cases, accidental or habitual, we see them developed into true leaves. If then the calyx is of a leafy nature and so very analogous to the involucre, why describe it on a diametrically opposite plan? Why consider it as a unique organ, more or less divided, instead of saying, as in the preceding cases, that it is formed of pieces more or less united together? Besides, the latter method involves no more hypothe

sis than the former; since, in a very considerable number of plants, the sepals are completely distinct from one another, and even attached separately to the peduncles. It is best supported too by their anatomy, for all the nerves of the calyx are directed from the base to the summit, as in leaves, though constantly described as if they proceeded from the summit to the base, and since all modern botanists admit the union of the calyx to the ovary, it would be strangely inconsistent to imagine, that the sepals could not be united as easily to one another as to a foreign organ. Instead of saying of a calyx, that it is deeply cleft, the most proper language obviously is, that the sepals are united only at the base; instead of describing it as lobed and toothed, the sepals should be considered as united half or more of their length; instead of distinguishing calices into polysepalous and monose palous, we are bound to use the distinctions of polysepalous, or free sepals, and gamasepalous, or sepals more or less united, and reserve the term monosepalous for the rare cases, where there really exists but one lateral sepal.

The same reasoning, the same analogies are applicable with perhaps still greater force, to the operation of the same law upon the corolla. This is not an unique whole, more or less divided, any more than the calyx, but an assemblage or whorl of petals, sometimes perfectly free and sometimes more or less united. In many cases this union is in a manner manifest to the eye, while in others, it is indicated by the disposition of the vessels; where it is not thus visible, and the tubes are continuous, it may be conjectured by analogy, and by the insensible gradations to be observed between corollas with petals entirely free, and those with petals united. The corolla of the clover is formed of but one piece, instead of four separate and distinct petals, as in all the rest of the Leguminosae; yet who, on that account, would deny its analogy to that order, and that it differs only in the natural adhesion of its petals? Adopting the ordinary way of distinguishing corollas into monopetalous and polypetalous, we must suppose an organization entirely different, for what analogy is there between a flat petal associated with several others in a whorl, each attached to a single point, and a circular tubular petal, with many points of attachment and a sinuated margin? Such a fact can be considered as hardly possible, when we recollect how many families there are, in which we see plants with monopetalous

and polypetalous corollas, indiscriminately mingled together. And what are we to make of those corollas, whose pieces, as in the vine, are separate at their base, but united at the summit? This reasoning becomes still more striking, when we consider the light in which stamens have been viewed. These parts possess an extraordinary analogy to petals; their point of attachment is constantly the same; their number and position are generally symmetrical; the anatomy and physiology of the filament of the stamen is perfectly similar to that of the claws of the petals, and in some flowers, they pass into each other by such insensible gradations, that it is impossible to say where one begins and the other ends. This being the case, we ought certainly to expect that the same mode of reasoning in regard to the adhesion of one should be equally applicable to that of the other. Now, however much the stamens may be united together, they never are considered in the light of an unique organ, divided more or less deeply into several parts, but always as separate and distinct organs, united according to the law of adhesion. But is this union of the filaments any more apparent than that of the petals? Are not the two phenomena equally constant in the same species? Are any more evident traces of it left in one than in the other? These two organs are of the same nature, and we must either consider the whorl of stamens as an unique whole, deeply cleft, or the whorl of petals as formed of many pieces more or less united. What would be thought of a zoologist, who should describe the feet of the web-footed birds as orbicular disks, divided to a greater or less extent? All naturalists regard them as distinct digits, united by a membrane, and this manner of considering organs as compound bodies, is the only one that represents the natural state of things,-the only one that admits of clear expressions and exact comparisons.

The truth of this theory becomes still more manifest, when we attend to the manner in which petals adhere at their base. In a polypetalous flower we see that generally each petal is fixed at its base by a fibre which carries its nourishment, and that if its base be very large, the rest adheres only by cellular tissue. Every family has thus a certain disposition in the vessels of the petals, and it is always the same, whether they be united or not. This analogy is equally striking, when considered in another point of view. Petals are composed generally of a claw and limb, as stamens are of filament and anther, and