who has investigated the nature of spiral vessels with singular skill (De vera Vasorum Plantarum spiralium Structurâ et Functione Commentatio, 1829) that it is solid, if it did not appear to have been ascertained by Hedwig that, when coloured fluids rise in spiral vessels, they follow the direction of the spires. This fact may, however, be explained upon the supposition that they rise in the channels formed by the approximation of cylindrical fibres, and not in the fibres themselves; in which case there could be little doubt that the fibres are really solid.

The nature of the termination of spiral vessels is now placed beyond all doubt, by the preparations of Mr. Valentine, above alluded to, and by some observations of my own. It is stated by Professor Nees von Esenbeck, in his Handbuch der Botanik, published in 1820, that they terminate in a conical manner; and in 1824 M. Dutrochet asserts, that they end in conical spires, the point of which becomes very acute; but one would not suppose, judging from the figure given by the latter writer, that he had seen the terminations very clearly. It is, however, certain that the statement of Nees von Esenbeck is correct, and that the spiral vessel generally terminates in a cone. If the point of such a vessel in the Hyacinth (Plate II. fig. 9.) be examined, it will be seen that the end of the spiral fibre lies just within the acute point of the vessel, and that the spires become gradually more and more relaxed as they approach the extremity, as if their power of extension gradually diminished, and the membrane acquired its pointed figure by the diminution of elasticity and extensibility in the fibre. It is not, however, always in a distinct membrane that the spiral vessel ends. In Nepenthes the fibres terminate in a blunt cone, in which no membrane is discoverable. (Plate II. fig. 11.) *

* A singular change occurs in the appearance of the spiral vessels of Nepenthes, after long maceration in dilute nitric acid, or caustic potash ; the extremities cease to be conical and spirally fibrous, but become little transparent oblong sacs, in which the spires of the fibres gradually lose themselves. This alteration, which is a very likely cause of deception, is perhaps owing to the extremities of the vessels being more soluble than the other part, the sac being the confluent dissolved fibres. This is in some measure confirmed by the subsequent disappearance of all trace of fibres in any part of the vessels, under the influence of those powerful solvents.

A spiral vessel is formed by the convolutions either of a single spire, or of many. In the former case it is called simple, in the latter compound. The simple is the most common. (Plate II. fig. 9.) Kieser finds from two to nine fibres in the Banana; M. de la Chesnaye as many as twenty-two in the same plant. There are four in Nepenthes. (Plate II. fig. 11.) In general, compound spiral vessels are thought to be almost confined to Monocotyledonous plants, where they are very common in certain families, especially Marantaceæ, Scitamineæ, and Musacea; but their existence in Nepenthes, and, according to Rudolphi, in Heracleum speciosum, renders it probable that future observations will show them to be not uncommon among Dicotyledons also.

In Coniferæ the spiral vessels have in some cases their spires very remote, and even have glands upon their membrane between the spires. (Plate II. fig. 6.)

In size, spiral vessels, like other kinds of tissue, are variable; they are generally very small in the petals and filaments. Mirbel states them to be sometimes as much as the 288th of an inch in diameter; Hedwig finds them, in some cases, not exceeding the 3000th; a very common size is the 1000th.

An irritability of a curious kind has been noticed by Malpighi in the fibre of a spiral vessel. He says (Anat. p. 3.) that in herbaceous plants and some trees, especially in the winter, a beautiful sight may be observed, by tearing gently asunder a portion of a branch or stem still green, so as to separate the coils of the spires. The fibre will be found to have a peristaltic motion which lasts for a considerable time. An appearance of the same nature has been described by Mr. Don in the bark of Urtica nivea. These observations are, however, not conformable to the experience of others. M. De Candolle is of opinion that the motion seen by Malpighi is due to a hygroscopic quality combined with elasticity; and as spiral vessels do not exist in the bark of Urtica nivea, it seems that there is some inaccuracy in Mr. Don's remark.

The situation of spiral vessels is in that part of the axis of the stem surrounding the pith, and called the medullary sheath, and also in every part the tissue of which originates from it; such as the veins of leaves, and petals; and of all other

modifications of leaves. It has been supposed that they are never found either in the bark, the wood, or the root; and this appears to be generally true. But there are exceptions to this: Mirbel and Amici have noticed their existence in roots; and Mr. Valentine and Mr. Griffiths have both extracted them from the root of the Hyacinth; they do not, however, appear to have been hitherto seen in the roots of Dicotyledonous plants. I know of no instance of their existence in bark, except in Nepenthes, where they are found in prodigious quantities, not only between the alburnum and the liber, embedded in cellular tissue, as was first pointed out to me by Mr. Valentine, but also sparingly both in the bark and wood. They have been described by myself as forming part of the testa of the seed of Collomia, and Mr. Brown has described them as existing abundantly in that of Casuarina. In the former case, the tissue was rather the fibro-cellular, as has been already explained (p. 11.); in the latter, they are apparently of an intermediate nature between the cellular-fibrous and the vascular; agreeing with the former in size, situation, and general appearance, but differing in being capable of unrolling. In the stem of Monocotyledonous plants, spiral vessels occur in the bundles of woody tissue that lie among its cellular substance; in the leaves of some plants of this description they are found in such abundance, that, according to M. de la Chesnaye, as quoted by De Candolle, they are collected in handfuls in some islands of the West Indies for Amadou. The same author informs us, that about a drachm and a half is yielded by every plantain, and that the fibres may be employed either in the manufacture of a sort of down, or may be spun into thread. In Coniferous plants they are few and very small, and in Flowerless plants they are for the most part altogether absent; the only exceptions being in Ferns and Lycopodiaceæ, orders occupying a sort of middle place between flowering and flowerless plants: in these they no doubt exist. My friend Mr. Griffiths has succeeded in unrolling them in the young shoots of Lycopodium denticulatum.

Some have thought that the spiral vessels terminate in those little openings of the cuticle called stomata; but there does not seem to be any foundation for this opinion.

DUCTS (fig. 8, 9, 10, 11, 12.) (Fausses trachées, Fr.; Saftröhren, Germ; Tubes corpusculifères of Dutrochet, Lymphaducts, or Sap-vessels of Grew and others; Vaisseaux lymphatiques of De Candolle, Vaisseaux pneumatiques of others ;) are membranous tubes, with conical or rounded extremities; their sides being marked with transverse lines, or rings, or bars, or dots arranged spirally, and being incapable of unrolling.

In some states these approach so nearly to the spiral vessel, that it is impossible to doubt their being a mere modification of it, as is the case in the annular duct (Plate II. fig. 13.); but in other states, as in the dotted duct, it is impossible to trace the transition from the one form to the other. Some writers confound all the forms under the common name of spiral vessels, but it is more convenient to consider them as distinct, not only because of their peculiar appearances, but because they occupy a station in plants in which true spiral vessels are not found; and it is therefore probable that their functions are different.

All the forms of the duct seem reducible to the following varieties:

1. The Annular (fig.11., and Plate II. fig. 13.). These are well described by Bischoff as being formed of fibrous rings, placed at uncertain intervals; or, to speak more accurately, they, like spiral vessels, are formed of a spiral thread, but it is broken at every coil, so as to separate into a number of distinct rings. These rings are included within a membranous tube, by which they are held together. When the rings are distant from each other (Plate II. fig. 1. b), the duct has a very peculiar appearance; when the rings are packed together, so as to touch each other (Plate II. fig. 18.), the external appearance is exactly that of a spiral vessel, from which they are known by being incapable of unrolling. Both these forms are common in the soft parts of plants, particularly in the root, and also in Ferns and Lycopodiaceae among flowerless plants.

2. The Reticulated (fig. 10. 12., and Plate II. fig. 13. a.). In these the spiral fibre, instead of separating into a number of distinct rings, is continuous in some places, anastomoses in others, so as to form a sort of netted appearance, or even

breaks into short lengths, which, adhering to the sides of the membrane, give the vessel the appearance of having transverse bars. It is these appearances that have given rise to the notion of cracked, or pierced ducts (fig. 10.) existing in plants; the membrane between the spires, or bars, having been mistaken for pores; hence the term vaisseau fendu, used by Mirbel and others. Vessels of this kind are found in the stem of some herbaceous plants; as, for example, the Impatiens Balsamina, in which they may be found in a great variety of states.

3. The Dotted (fig. 9.). Ducts of this kind are tubes having their sides marked with numerous dots, arranged in a more or less spiral manner, and being divided internally by transverse partitions. Usually, in addition to the dots, there is distinctly visible an oblique or annular transparent line upon the walls of the vessel. (Plate II. fig. 15. 17.) Hence Kieser considered them as spiral vessels, the spires of which, when old, elongate, and become connected by a dotted membrane. Bischoff, on the contrary, considers the dots to be caused by the separation of a spiral fibre into extremely minute portions; and he gives a figure (Plate II. fig 16.) of the manner in which he considers this change to occur.

It is certain, however, that the dotted duct is really an entirely distinct kind of vessel, or at least a modification of cellular rather than of vascular tissue, as has been asserted by Du Petit Thouars (Ann. des Sciences, vol. xxi. p. 224.); for the following reasons: If it were such a modification of the spiral vessel as Kieser supposes, it would have none of those internal septa by which it is particularly known. The same remark applies to the theory of Bischoff, which is also imperfect, in not accounting for the nature of the transverse transparent lines that mark the sides of dotted ducts. Besides, the dotted ducts always terminate abruptly, not in acute cones, as has been seen by myself, and well represented by Mr. Griffiths, in his excellent illustrations of the anatomy of Phytocrene (Plate II. fig. 19. 20.), and they readily separate at the septa; none of which properties are those of a spiral vessel. That the partitions above alluded to really exist, as has been correctly stated by Dr. Dutrochet, there can be no doubt, notwithstanding the denial of the fact by Link and