OUTLINES OF SELANTHOLOGIA.

(4891.) Certain splendid Oriental parasites, only within the last few years discovered, and several others, natives both of Europe and America, which, although longer found, can scarcely be said to have been longer truly known, so strangely were they at one time misunderstood, form, like the Cycases and the Pines, a small but very natural class, and one in which, by a rapid series. of degradations, the most highly developed plants, the most elaborate examples of vegetable structure, are affianced, and become connected with some of the earliest and simplest grades.

(4892.) The gorgeous appearance and gigantic size of some of these extraordinary parasites, such as the Rafflesia, may well vindicate their collective appellation, SELANTHI, should that word seem preferable to CYTINARES, a derivative of Cytinus, the name of one of the longest known, but more humble and less showy genera. To these terms, however, as the whole of the plants are foreigners, and the chief of extra-European growth, no familiar English equivalents can be expected to be found; and hence Selworts, formed on the model of some of the older names of plants, such as Sel-ago, ground-sel, &c. may be taken as a translation or a synonyme.

(4893.) The Selanthi are cellular or almost evascular leafless flowering plants; thus, in their organs of vegetation being connected with the leafless, flowerless fungi, and by their organs of fructification with the tubivascular mono- and di-cotyledons; and hence has arisen the systematic paradoxes of at one time beholding these so-called cellular plants arranged amongst the tubivascular Exogene and Endogenæ, and at another of finding some of them,

as the Rafflesiæ, which consist exclusively of flower, located amongst the flowerless acotyledons in the class Cryptogamia.

Several anticipations of a relapse towards the earliest and simplest forms of vegetable structure were observable in the preceding class; for some of the parasitic Rosares, such especially as the Orobanchacea, Monotropa of the Pyrolide, and even Cuscuta of the Convolvulacea, agree in their parasitic habit with these plants; and the former are not only destitute of leaves, but also are furnished with scales in their stead, while the latter not only has an acotyledonous embryo, but establishes, through its associates, the Convolvulida, where the seed-lobes are shrivelled, a connexion between the embryonate and inembryonate vegetables: indeed, it is a question still undecided, whether Lathraa does or does not produce perfect seeds.

(4894.) CYTINUS, the Hypocist of the ancients, and CYNOMORIUM [§ 114], the old Fungus Melitensis, or mushroom of Malta, have been the longest known, and at various times, as already hinted, they have been very variously placed in systematic arrangements. The latter, as its old name, Maltese Champignon, hints, was at one time supposed to be a fungus: by Jussieu it was left unarranged; but by Richard it was subsequently associated with Cytinus, and placed amongst the dicotyledonous Exogene by others, however, both it and several of its West Indian allies have been affirmed to be mono- rather than dicotyledonous plants, and their nearest affinities declared to be with the Aroidea (Callaceæ) and Hydrocharideæ. Agardh, on the contrary, includes them among the Urticidæ, considering them indeed only a subdivision of his Urticea; while Brown, who associates Rafflesia with Cytinus, and sanctions Brongniart's opinion of its relationship with Nepenthes, also corroborates Jussieu's approximation of Cytinus to the Asarina, or Aristolochiacea; but he considers the other genera as essentially distinct. Blume, in his Flora Java, has given to Cytinus, the Rafflesiæ, and their immediate allies, the common name Rhizanthea; and Sprengel makes the Rhizanthea, some of which consist of flower alone, the first order of his Cryptogamia, or flowerless class. Bartling associates the whole with the peppers, the taccas, and the water-lilies, under the collective term Chlamydoblasta, a subdivision of his Vegetabilia dicotyledonea; and numerous other speculations as to their systematic disposition have been adventured, one of the most plausible of which is their association with the parasitic Orobanches; but these it would be tedious further to dwell on, as the foregoing examples will sufficiently prove the uncertainty as to their chief affinities which has so long prevailed, and may excuse, if not justify, the present scheme.

(4895.) As already stated in the general Outline [§ 109, 110, 113], these vegetables differ from other flowering plants, whether of the Exogenous or Endogenous series, not only by their habit and port, both of which are peculiar, but also in their internal structure; for anatomical investigations have shown that, like the fungi, which they so curiously simulate in their destitution of leaves and parasitic habits, they chiefly consist of cellular or subcellular structure, which is

a further resemblance of great importance. It has indeed been affirmed, and generally believed, that they are altogether destitute of tubular vessels, whether

spiral or simple. Blume, who had admirable opportunities of pursuing anatomical investigations in the countries where the Rafflesiacea are indigenous, figures them as wholly cellular, and expressly states that they are destitute of tubular vessels; the only ones found in or about them being, according to him,

Anatomy of the Brugmansia Zippeli.

(i) Section through a plant while in the state of a bud, shewing its enveloping scales, and attachment to the roots of the cissus. (a) Longitudinal section through a plant previous to the opening of the flower. (bc) Internal views of the perianth. (e) The central column, to which the anthers are attached. (f) Portion of the column, with several anthers attached. (g) Transverse section of an anther. (h) Longitudinal ditto. (d) Section of the walls of the fruit, or pseudocarp, with the seeds (spores) attached.

those of the cissi, on the roots of which they grow, and which sometimes are confounded with the bases of the parasites. This doctrine was also held by Brown, who, when he published his admirable monograph on Rafflesia Arnoldi, maintained their cellular structure. This opinion of their utter destitution of

tubular vessels he has, however, since found reason to modify; for, as might almost have been anticipated, from the subfoliaceous appearance and strong ribs of the bud-scales of the Rafflesia, he has detected tubular vessels in them, both spiral and others, but they are very few in number, and not only spare, but also oftentimes imperfect. This is an important discovery, as it renders the transition from the Rosares, and the other tubivascular flowering plants, less sudden and abrupt than if they had been wholly evascular. Their fruit likewise he describes as being not quite so simple as had been previously supposed; for, although the seeds consist of a soft and nearly homogeneous mass, albumen and embryo are distinguishable, and hence, although spore-like, they are not veritable spores.

(4896.) These late observations of Dr. Brown were communicated to the Linnean Society while these pages have been passing through the press; and

Anatomy of the Brugmansia Zippeli. (Blume.)

(d) Transverse section of the root of the cissus, on which the plants grow parasitically, with a view of the confused substance at the base of the Brugmansia. (a) Section of the intermediate body, formed between the parasite and the cissus, which consists of cellular structure, permeated by tubular vessels. (b) Base of the parasite, and upper portion of the intermediate body, destitute of tubular vessels. (c) Longitudinal section of an anther, shewing the cellular structure of its parietes, and the locule containing a little pollen. (e) One of the cellular ovules (or spores), with its pedicel. (1) Section of the walls of the fruit (pseudocarp), shewing the attachment of 2 ovules. (gh) The mass of cellular substance of which the spore consists.

although, from the essay being as yet unpublished, advantage may not be taken of the whole of its contents, it would have been unjustifiable not to have introduced his correction of the generally received opinion of the utter evascular condition of

these plants. And this, the more especially as it so unexpectedly and in so decided a manner corroborates the present scheme of classification, rendering the Selanthi no longer anomalies in the tubivascular region, and confirming the view which has been taken of their intermediate rank between the Rosares and the Fungi.

(4897.) Collectively described, the SELANTHI or Cytinares are parasitic herbaceous plants, consisting wholly or chiefly of cellular structure, and destitute or nearly so of tubular vessels. The axis is sometimes abortive, sometimes developed as a horizontal branched rhizoma, and sometimes as a simple or divided stem. The leaves are absent, the foliage consisting of brown or colourless scales, more or less foliaceous in their appearance; part enclosing the buds, and part rendering the stem squamose.

The inflorescence is solitary or aggregate, usually spadiceous, and the flowers for the most part separate, rarely united.

The perianth is single, the stamens definite, more or less cohering both by stamens and anthers. The germen is inferior, invested by the adherent tube of the calyx, 1- or many-celled, and 1- or many-ovuled.

The fruit is dry or pulpy, a caryopsis or a berry, and indehiscent. The seeds are albuminous, the albumen fleshy, the embryo very minute, divided or undivided, (i.e. mono- or dicotyledonous ?) but sometimes not readily distinguishable from the albumen, the whole nucleus often consisting of a mass of homogeneous grumous matter.

(4898.) Hence, differentially considered, the Selanthi or Cytinares are cellular or subcellular flowering parasites, with spore-like seeds, and furnished with scales instead of leaves.

(4829.) As these plants, in themselves said to be acotyledonous, have been, and still are to the present day, referred by the most accomplished naturalists in part to the monocotyledonous, and in part to the dicotyledonous series, the differences in structure and appearance which have led to such distinctions might reasonably be expected to indicate at least two very different orders, notwithstanding their association in the same class. Two chief subdivisions are indeed recognizable, but there is so much similitude in their diversities, that it is somewhat questionable whether they should be considered as distinct orders, or merely as types or sections of one and the same. The differences, however, do on the whole seem to be sufficient to constitute them separate orders.

(4900.) Cytinus and Cynomorium, the two longest known and most familiar genera, give their names respectively to these two orders, which are all that this class includes; hence the one is called Cytinales, and the other Cynomoriales.

(4901.) The Cytinales are (dicotyledonous ?) Selanthi, with a 1- or many-celled ovary, parietal placentæ, many ovules, and the divided embryo straight, axile, and included within the albumen.

(4902.) The Cynomoriales are (monocotyledonous ?) Selanthi, with a 1-celled ovary, a solitary pendulous ovule, the embryo undivided, globose, and lodged in a superficial excavation of the albumen.

CYNOMORIALES.

(4903.) Cynomorium, Langsdorfia, Helosis, and Balanophora, the presumed monocotyledonous Cytinares which form this order, are distributable into two