FERNS, a large group of cryptogamie plants forming in a wide sense a very naturally limited group. In the article Biology (vol. iii. p. 694) it has been shown that the subkingdom Cormophyta includes three well-marked series, Bryoρhyta, Pteridoρhyta, Phanerogamoe. Of the second of these, Fems, in the most restricted acceptation of the word, are the most conspicuous representatives. It will be convenient, however, to give some account under the present head, not merely of ferns proper, but also of the other smaller groups, closely related to ferns and often spoken of in familiar language as Fern-allies, which are included in the Pteridoρhyta.
The life-history of all these is divisible into two distinct stages or generations, in which not only are the external morphological characters extremely different, but the physiological functions are also sharply contrasted. These stages have been termed in the article already referred to the sporophore and the oophore.
The vigorous vegetative forms which are what are familiarly understood when fems are spoken of are only one phase of their complete life-cycle. They are the sporophores; and the reproductive bodies they bear (spores) germinate without any process of fertilization, and are therefore not to be compared with seeds. The same thing is equally true of the other members of the series Pteridophyta, viz., horse tails, club mosses, selaginellas, &c.
A spore on germination produces a structure which, compared with its immediate parent, is relatively very small, and bears no resemblance to it in form or texture. It is called the prothallium; it is green and membranous, and attached to the surface of the ground by root hairs in ferns and horse tails; in adder's tongues and club mosses it is tuberous and subterranean, while in Rhizοcarpece and Selaginella it always remains more or less included within the cavity of the parent spore. The function of the prothallium is entirely reproductive; it develops sexual organs of two kinds, archegonia and antheridia, either upon the same or different prothallia. Hence, in contradistinction to the sporophore—the function of which in this group of plants is purely vegetative—the prothallial generation is termed the oophore.
History.— The older botanists, starting from the study of flowering plants, endeavoured to recognize in all other cases the same characteristic points of structure. They did not doubt therefore, arguing from analogy, that ferns were provided with true seeds. Gerarde (1597), remarking theappearance of young plants of NephroHum dilatatum in the neighbourhood of old ones, attributed this to the dissemination of seeds, “for I believe all plants have seeds in themselves to produce their kindes.” No one, however, at this time could say anything definite about the “seeds” of ferns, and they came to be regarded as highly mysterious. It was supposed that they were of an invisible sort, and by a transference of properties it came to be believed that the possessor of fern seed would be invisible too. Shakespeare makes Chamberlaine say, “We have the receit of fern-seed, we walke invisible.” Ray ridicules Tragus for spreading linen clothe upon the ground to catch the seed of ferns on the eve of midsummer night, when, as Ray properly observed, it would be indeed astonishing if any fell, seeing that it is not mature till the autumn.
Columna in 1648 compared the fern frond to butcher’s broom, and identified the fructification in the one case with the flowers in the other. Caesius came to his help, and, having examined the sori of a polypody, “telescopii adjumento,” detected what he supposed to be seeds, but which were probably only the spore cases. W. Cole appears to have been the first to microscopically observe the spores themselves (1669), and Ray himself described the hygroscopic movements of the spore-cases which assist the dispersion of the spores. Morison (1715) seems to have been the first to put the matter to the test of actual experience. He sowed the spores of hart’s tongue, and in due course, without, however, distinguishing clearly the two stages, raised plants from them. This important observation fell completely into oblivion, and when in 1789 Dr Lindsay, a Scotch physician settled in Jamaica, responded to a request of Sir Joseph Banks for fern plants from that island by suggesting that spores would be much more convenient to send, the latter was quite unprepared for the suggestion, which he treated as a great discovery. Dr Lindsay accordingly wrote a paper on the subject (which was published in the Linnean Transactions a few years afterwards), in which he figured the prothallial stages of Polypodium lycopodioides. The spores of ferns were still, however, regarded as equivalent to seeds, t.c., as the result of a process of fertilization similar to that which precedes the development of seeds in flowering plants. A variety of attempts were made to see in hairs or glands upon the young fronds, &c., something that would do duty for an anther. As late as 1832 De Candolle maintained the analogy of the prothallium to a cotyledon. In 1844 Nägeli discovered the antherozoids, motile bodies which are set free from the antheridium, and which perform in fems the function of pollen grains. All previous observers had tried to find fertilizing bodies on the vegetative generation, the sporophore, which produces only spores. In 1844 the final step was reached when Suminski found upon the prothallia the archegonia, flaskshaped bodies containing a central cell (oosphere) which, when fertilized by the antherozoids, develops into a sporebearing plant similar to that from which the cycle started. In 1850 Mettenius discovered the details of the reproduction in Isoetes. Since that time the researches upon the Pteridophyta have been very numerous, and are to be found quoted in Sachs’s Lehrbuch.
The whole group of the Pteridophyta may be classified as follows :—
Pteridophyta. —Conuophyta with two distinct stages in the life-cycle. Sporophore with high vegetative differentiation. Oophore inconspicuous and destitute of vascular tissue.
Class I. Filicince.— Leaves highly developed. Sporangia numerous on the fertile leaves.
Sub-Class 1. Filices. — Leaves without stipular appendages. Sporangia epidermal, containing spores of one kind developed in each from a single primary mother-cell.
Sub-Class 2. Stipulate:.— Leaves with stipule-like appendages. Sporangia containing spores of one kind developed in each from many endogenous primary mother-cells.
Sub-Class 3. Rhiaocarpeoe. — Spores of two kinds.
Class IL Equisetinae. — Leaves rudimentary. Sporangia 5-10 on the fertile leaves.
Class III. Lycopodinae. — Leaves small (except Isoetes'}, simple. Sporangia solitary.
Sub-Class 1. Lycopodiaceae. — Spores of one kind.
Sub-Class 2. Ligulatae. — Spores of two kinds.
I. Filicinae. —1. The Filices or ferns proper deserve consideration at greater length, both on account of their numerical preponderance over the other groups, and the popular favour which deservedly attaches to their great beauty and variety of form.
Commencing with the vegetative generation or sporophore, the following remarks briefly touch upon the most essential points in their organization.
In general habit, although with a certain characteristic appearance which, to the experienced eye, is an almost unfailing distinctive character, ferns are as various in stature and aspect as flowering plants. One genus, Cerαtopteris, is remarkable for its aquatic habit. The terrestrial forms vary from the Hymenophyllaceae, which have the small size and delicate texture of mosses, to the larger forms vzhich attain the size of shrubs, while some belonging to the tropics and the southern hemisphere havo a palm-like habit, and are called tree-ferns. Cyathea medullaris reaches 80 feet in New Zealand, and Alsophila australis even greater heights in Australia. The stem creeps on or beneath the surface of the ground as in the common bracken, or climbs up rocks and the trunks of trees {Davalliα canariensis). In the latter case it is usually densely coated with imbricated membranous scales. In stems of this type the internodes or spaces between the insertion of successive leaves (fronds) is considerable, and the growing point is often far in advance of the youngest frond. In a large number of herbaceous ferns, of which the common male and lady fems are good examples, the stem is short and erect or slightly inclined (fig. 1). The fronds are developed spirally, and in close contiguity without obvious internodes; they form, viewed from above, a large rosette with the growing point in the centre. Between the ordinary herbaceous ferns with short erect stems and the tree-ferns there is no hard and sharp line of demarcation. Most of the ferns of decidedly arborescent habit belong to the sub-order Cyatheaceae, which has a sporangium and receptacle a little different from that of the typical Polypodiaceae, but in many of the genera of Poly podiaceae, especially Dicksonia, there are species that assume a more or less decidedly arborescent habit Nearly all Filices are perennial, but we have an instance of annual duration in Gymnogramme leptophylla. Growing Point.— The apex of the stem, whatever the dimensions of the mature plant, is always occupied by a single apical cell, and the whole growth is entirely dependent upon the repeated segmentation of this cell. In form it is either wedge-shaped with a transverse section like that of a bi-convex lens, or it has the form of a three-sided pyramid with a convex base forming the outer face. In the former case the segments are produced from the cell alternately right and left, and the stem, which is gradually built up by the repeated subdivision of the segments so formed, is always distinctly bilateral, and produces its fronds in two rectilinear series. In the latter case the segments are derived in spiral succession from the three interior faces of the apical cell, and the leaves, which are developed from the masses of tissue to which these segments give rise, are gradually displaced from the simple triseriate arrangement due to the segmentation of the apical cell into more complicated forms of phyllotaxis.
Branching is effected by the partition of the growing point; the single stem therefore bifurcates. This scarcely ever happens in ferns with erect stems, such as Nephrodium Filix-mas, or in tree-ferns. New shoots are, however, sometimes formed from the bases of the leaf-stalks, just as in some species they are normally produced from the fronds themselves. When the terminal point is injured, lateral shoots are often formed in this way. A specimen of Dick sonia antarctica has been described with as many as twenty-nine crowns, due doubtless to such a cause.
Structure of the Stem. — The repeated subdivision of the cells of the growing point gives rise to a mass of tissue, at first uniform in character, but which is subsequently differentiated. Particular tracts of cells are modified in the process of growth to form the fibro-vascular bundles. These are more or less flattened plates irregularly elliptical in transverse section. As in all the Pterιdophyta they are closed, and incapable, when once formed, of further increase in diameter. They consist of a mass of “xylem” surrounded by “phloem.” The xylem consiste, in addition to a few spiral vessels at the foci of the elliptical tranverse section, of scalariform vessels—cells of comparatively large diameter, with oblique ends, and with their walls intermittently thickened by transverse projecting ridges (fig. 2). Amongst the vessels are oblong cells containing starch in winter (woodparenchyma). The phloem contains sieve-tubes, and external to these thick-walled bast fibres. The whole fibro-vascular bundle is bounded first by one or two layers of oblong starch-containing cells (phloem-sheath), and external to these by a single layer of narrower cells, often thickened externally and darker in colour—the vascular bundle sheath. In very slender stems the fibro-vasculαr bundles form a central axis tflymenophyllaceae), but when the stems attain any size they are arranged in an anastomosing cylindrical network (fig. 3), in which a mesh corresponds to the base of a frond. This cylinderseparates the remaining or fundamental tissues of the stem into two portions, an internal central medulla and an externaΓcortex. This is shown in fig. 4; the dark lines, however, do not belong to the bundles which correspond to the light spaces enclosed by them; the dark lines themselves consist of a mass of thickened elongated cells termed “sclerenchyma,” which form protecting sheaths to the fascicular tissue. The medulla and cortex are in the main formed of parenchymatous tissue, often loaded with starch. In Cyathea medullaris this starchy pith was used by the Maoris as an article of food. The medulla is not, however, entirely parenchymatous; it is usually permeated by a number of anastomosing filiform bundles (usually also invested with a sclerenchymatous sheath), which eventually pass into the leaves through the meshes in the fibro-vascular cylinder. The greater part of the fibro-vascular tissue of the leaves is, however, continuous with the lower portion of the margins of the meshes, which runs out into bundles arranged round the outer circumference of the leaf stalk, and surrounding therefore the bundles which pass through the meshes from the medulla of the stem. In some cases, however, these medullary bundles contract an adhesion to the margins of the meshes as they pass through them. The surface of the stem, and also of the leaf-stalks,
13 usually strengthened by the sclerenchymatous development of the superficial portions of the fundamental tissue. This coats the surface with a layer which would be impervious to air were it not interrupted by spaces filled with soft parenchyma, by which the internal tissues of the stem are brought into relation with the atmosphere. In Pteris aquilina, the rhizome or horizontal subterranean stem is marked laterally by two lines where the colourless cortical parenchyma is not covered in, and. in many tree-ferns the bases of the petioles have their sclerenchymatous investment perforated by pits filled in the living plant by a soft-celled tissue which speedily becomes pulverulent.
The roots of ferns are always slender and wiry. They are continually formed during the growth of the stem in acropetal succession. In tree-ferns they consequently accumulate on the lower portion as they grow downward, giving it a large apparent diameter. In creeping species they at once attach themselves. They are cylindrical in all cases, scarcely exceeding one-eighth of an inch in diameter, dark-brown approaching black in colour, and clothed with root-hairs.
In ferns the occurrence of a special form of “trichomes” is noticeable in the flat multicellular membranous scales, to which the name of “paleae” is given. They are found in the greatest abundance on the stem and lower part of the stipes, but they often extend to the rachides of the frond, and sometimes to the foliar surfaces, especially in young fronds. There is no definite correlation in the character of these paleae with other points of structure. The most characteristic are the long thread-like bright-brown paleae of EiLdicksonia, which are so abundant that those of Dicksonia Culcita are exported in quantity from Madeira to stuff cushions and mattresses; and the clathrate flat paleae, dull brown in colour and permeated by a distinct lattice-work of thicker cells, which occur in Vittama, Antroρhyum, and some species of Asplénium. By their abundance, position, size, shape, and texture, the paleae often furnish great help in characterizing species.
Stipes. —Although generally distinguished by this special name, the stipes is simply the petiole of the leaf. It is very rare to find a fern in which the blade of the leaf does not possess a distinct stalk. In about two-thirds of the Filices, including not only those with erect, but many with creeping stems, the stipes is continuous with the stem, without any articulation at the point of junction; and when the old fronds die, they do not become disjointed, after the fashion of the leaves of a deciduous tree. In about one-third of the Filices, on the contrary, there is a distinct joint at the point of junction between the stipes and the stem, and when the fronds fade, they fall away from the stem spontaneously, leaving a clean scar. In the classification proposed by John Smith, the Polypodiaceae are subdivided by this character. The ferns in which the leaf-stalk is continuous with the stem, as in Asplénium, Nephrodium, and Pteris, he calls Desmobrya. Those in which the old fronds become disarticulated at the base, as in Oleαndra, Platycerium, and Polypodium vulgäre, he calls Eremobrya; and he states, as the result of his experience as curator for 40 years of the Kew collection, that the Desmobrya may be more easily reproduced from spores, but that the Eremobrya show greater tenacity of life, and that in the importation of ferns from abroad the Eremobrya generally arrive in good condition, whilst the Desmobrya, especially such as have slender stems, often perish in the journey.
Fronds.— Iu Filices we get a more varied range of leafform than in any other order in the vegetable kingdom, and some of the most beautiful and complicated types that exist. In size the fronds vary from under an inch, as in some of the minute Hymenophyllae, to a length and breadth of from ten to twenty feet in the great Cyatheas and Dicksonias. To take examples of the leading types from amongst familiar British species, we have simple fronds in Scolopendrium vulgare, simply pinnate fronds in Asplénium Trichomoenes aηdLomaria Spicant, bipinnate in Nephrodium montanum and Polypodium Phegopteris, tripinnate or tripinnatifid in Cystoptems fragilis and Polypodium Dryopteris, and tecorn^oandinAsplenuimAdiantum-nigrum and Pteris aquilina. The primary divisions of a compound frond are called pinnae, the secondary divisions pinnules. In a deltoid frond, as in Cystopteris montana or Pteris palmata, the pinnae grow regularly smaller from the lower to the upper, and the lower pinnae have the pinnules of the under side exaggerated. A decrescent frond, as in Nephrodiτιm conterminum or Asplénium Trichomanes, has the pinnae largest at the centre of the frond, and gradually smaller both towards its top and bottom. There is no regular relation between the frond and its fructification. Species with fronds similar in texture, cutting, and veining are often found, when the fructification is examined, to stand widely apart in systematic position. · The growth of the fronds is very slow. The whole of the fronds of a rosette of Nephrodium Filix-mas have been in course of formation two years before the lamina begins to unroll. The development is basifugal, and often the fronds continue growing at the top a long time after they are fully unfolded and producing fructification at the base. This may be seen iu Hymenophyllum sericeum and Gymnogramme elongata; iu Lygodium the leaf-stalk resembles a twining stem, growing for a long time. Iu many cases the fronds are of two kinds, developing from the first in a conspicuously different manner, one kind remaining permanently barren, and the other alone producing fructification. Wo have instances of this in Lomaria Spicant and Cryptogramme crispa, and dimorphism is shown still more conspicuously in Platycerium, in which the barren fronds are uncut and spread out flat over the surface on which the plant grows, and the fertile ones are much larger, and stand erect, and are forked, from which it takes its common name of elk’s-horn fern. There is another striking instance of dimorphism in the section Dryηaria of the genus Polypodium, in which the barren fronds are brown in colour, rigid in texture, small in size, and lobed like the leaf of the common oak, and the fertile ones are several feet long, with elongated segments like those of Polypodium vulgare. In some ferns the fronds have a habit of taking root at the tip, and thus producing new plants. This is well shown in the walking-leaf of North America {Scolopendrium rhizophyllum), and amongst tropical ferns in Asplénium rhizoplωrwn and Acrostichum flagelliferum. Some species develop copious adventitious buds, from which the plant may be reproduced, sometimes scattered over the surface of the frond, as in Asplénium viviparum and A. bulbiferum, sometimes in the axils of the pinnae, as in Asplénium proliferwm and Nephrodium cicutarium. Venation.— In the arrangement of the veins of the fronds of Filices there is an extraordinary amount of variety. The different types, and the extent to which they prevail, and the use that may be made in the systematic arrangement of fems of the characters they furnish, were carefully worked out by Presl and John Smith. The two principal leading types are veins that, after once branching, do not again unite, and veins that, on the contrary, after branching, join one another so as to inclose areolae like the meshes of a net. Ferns that exhibit the first plan are said to be free-veined; the other kind are said to have anastomosing venation. The free-veined ferns form far more than half of the order. In the whole frond or its ultimate divisions there may be a midrib alone, as in Monogramme graminea or Hymenoρhyllum tuηbridgense; or the veining may be simply pinnate, that is to say, a midrib - with a series of parallel simple branches on each side, as in Nephrodium patens or N. prolixum ; or there may be a regular midrib from the base to the top of the frond or segment, with the side-branches branched again, a very common type shown in Pteris aquilina and Nephrodium Filix-mas; or there may be no regular midrib, but the veins arranged like the rays of a fan, as in Adiantum Capillus-Veneris and Asplénium cuneatum. Whether the veins are produced to the margin, or end abruptly without reaching it, as Î3 common in the free-veined, and still more so in the net-veined species of Polypodium, is a point to be noted. Of anastomosing venation the following are the principal types :—
1. "Where the veins are free nearly to the edge, but connected by an arch just within it, as in the section Thamnopteris of Asplénium and Olfersia of Acrostichum.
2. Where the veins are arranged in parallel pinnate groups, and the opposite veins of contiguous groups join at the tip, as in the section Eunephrodium of Nephrodium and Goniopteris of Polypodium.
3. Where the veins of contiguous segments are copiously pinnate and free, but the midribs are connected by a single arch and the base, as in the section C'ampleria of Pteris and Pleocnemia of Polypodium.
4. Where the veins imite in regular hexagonal meshes the sori borne at the tip cf a single central vein which these inclose, as id the section Goηiophlebium of Polypodium. 5. "Where distinct main veins run from the midrib to the edge of a frond or its divisions, and these are connected by a series of parallel arches, from the upper side of which several free veins arise, bearing sori at their tips, as in the section Campyloneuron of Polypodium.
6. Where distinct main veins run from the midrib to the edge of a frond or its divisions, with copious small irregular areolae between them, as in the section Pleuridium of Polypodium and Gymnopteris of Acrostichum.
7. Where the whole space between midrib and margin is filled up with copious small irregular areolae, as in the section Phymatodes of Polypodium and Chrysodium of Acrostichum.
Many of the large genera, as characterized by the shape and position of the sori, contain groups which show several of these different plans of veining ; Polypodium and Acrostichum show them nearly all. One school of systématiste, represented in England by J. Smith and Moore, regard a decided difference in veining as a character important enough in itself to mark a genus; and if this plan be followed the number of genera is enormously increased. Our estimate of 50 for the genera of Polypodiaceae is a large one, if genera founded upon fructification be understood; but Moore, using venation in addition, reckons the genera at 143, and John Smith at 194. From this different scale upon which genera have been planned arises the large number of different generic names which have often been used for the same plants, so that the synonymy of ferns has become extremely complicated.
The sporangia of ferns are morphologically trichomes or modified hairs. They originate from epidermal cells, and in the mature state consist of stalked capsules, the walls of which are one cell thick. A ring of cells differently placed in different genera is called the annulus, and by its unequal contraction in drying effects the disruption of the capsule and the dispersion of the spores (fig. 5). In some genera the same object 13 effected by a group of cells instead of a ring. The separate clusters, of few or many individual sporangia, are called sori (fig. 6), and the membranous excrescence of the epidermis which covers a sorus is called an indusium or involucre (fig. 7).
The orders of FUices are founded upon differences in the structure of the individual sporangium. Genera of the same order depend upon differences in the shape and position of the sori and shape and position of the indusium. In a few genera, such as 0smunda, Anemia, and Schizaea, the parenchyma of the lamina is entirely absorbed in the fertile parts of the frond, and the sporangia form distinct panicles or spikes. In Acrostichum the sporangia cover the whole of the under surface of the frond, both veins and interspaces. In many genera, such as Pteris, Vitiaria, and Lind-saya y the sori run in a line along the very edge of the frond, connecting the tips of numerous veinlets. In Blechnum they form a similar line close to the midrib. The commonest type of all is when they form round dots on the veins of the back of the frond, as in Polypodium, Aspidium, Woodsia, and Gyathea. A type only less common than the last is where they run in linβ3 along the veins of the back of the frond, as in Aspleηium, Scolopendrium, and Gymnogramme. In Lygodium the sporangia form spikes on the edge of the leafy segments, each sporangium inclosed in a little pocket. In Dichsonia, Hymenophyllum, and Trichomanes, the sori form round clusters at the tip of a vein on the edge of the lamina. Characters taken from the indusium separate the genera of Polypodiaceae into four tolerably well-marked groups. In a large series of genera, including Polypodium with not less than 400 species, Acrostichum, and Gymnogramme, there is no indusium at all. In another series, including P teris, Pellaea, Cheilaηthes, and Hypolepis, there is what is called a spurious indusium, which consists simply of the edge of the frond modified in texture and more or less completely recurved, so as to protect the sorus in an early stage. The true indusia are distinct scales, and are of two kinds,—those called inferior, which originate beneath the sorus, and may be distinctly cup-shaped, as in Woodsia and Gyathea, or may be one-sided, as in Cystopteris and Daυallia; and those called superior, as in Nephrodium and Aspidium i which are extended over the sorus. As has been already indicated, all the genera of Polypodiaceae, which include in the aggregate not less than 2000 species, agree with one another in the structure of the sporangia, so that the great majority of existing forms of Filices are very closely allied. The classification of ferns is based at present upon characters entirely drawn from the sporophore generation, and takes no account of the oophore. It is undoubtedly to a great extent artificial, but until the development of the prothallium has been carefully worked out in all the different orders, no more satisfactory or more natural arrangement than the following appears attainable.
1. Gleicheniaceoe. — Sporangia sessile, splitting vertically, furnished with a complete horizontal ring. Sori composed of very few sporangia; receptacle not elevated. Fronds with very distinct dichotomous branching. Genera 2; species 30. ·
2. Hymenophyllaceae. — Sporangia sessile, splitting vertically, furnished with a complete horizontal ring. Sori composed of numerous sporangia, inserted on a long filiform receptacle. Fronds of filmy texture, with pinnate branching. Genera 2; species 150-200.
3. Cyatheaceae. — Sporangia nearly sessile, splitting transversely, furnished with a usually incomplete nearly vertical but rather oblique ring. Receptacle prominent, barrel-shaped. Tree-ferns. Genera 3; species 150.
4. Polypodiaceae. — Sporangia stalked, splitting transversely, furnished with a usually incomplete vertical ring. Receptacle not prominent. Genera 50; species 2000.
5. Osmundaceae. — Sporangia stalked, splitting vertically, furnished with only a faint horizontal bar instead of a ring. Genera 2; species 10-12.
6. Schizceaceae. — Sporangia sessile, splitting vertically, crowned by a complete small annular horizontal ring. Genera 5; species 60.
The ooρhore, or sexual generation of ferns, not having been studied to any great extent from the comparative point of view, can only be described in general terms. It is a membranous structure furnished with root hairs, and the cells of which contain chlorophyll; it is therefore capable of obtaining its nourishment independently, and in many respects has a striking resemblance to the simpler Hepaticae. When the spores (for example) of any Polypodiaceous fern are sown, the outer covering bursts, and the inner coat protrudes as a filament, which soon divides and continues to grow into a short row of cells. Cell division next begins to take place laterally as well as longitudinally, and a small cellular plate is formed (fig. 8), which gradually attains a kidney-shaped outline, the growing point being situate in the notch. From the under side are developed the root hairs, antheridia and archegonia.
The antheridium is morphologically a trichome, and is an outgrowth of a superficial cell. The wall when mature consists of a single layer of cells, inclosing a central cell, from which the antherozoids are developed in separate mother cells. The antheridium eventually bursts owing to the absorption of water, and the spirally-twisted motile ciliated antherozoids (fig. 9) are set free (fig. 10).
The archegonia are produced on a cellular cushion immediately behind the notch. They are developed by repeated divisions from a superficial cell, and, when mature, each is a bottle-shaped structure (fig. 11), the rounded base of which is buried in the prothallium, and contains a central cell, the oosphere. The canal of the neck is at first filled by a cell, which eventually deliquesces, and leaves the passage open for the access of the antherozoids to the oosphere, which, after fertilization by blending with them, becomes the oospore. The oospore first divides by a division-wall oblique to the prothallium, and then divides again so as to form four cells. Two of these give rise to a mass of tissue which remains in contact with the prothallium, and is called the foot. Of the other two, one develops into the stem (first leaf), the other into the root. As soon as the young fern-plant (sporophore) is fairly established (fig. 12) the prothallium (oophore) disappears. Its existence is therefore remarkably contrasted by its brevity with the oophore stage in the Bryophyta. There are, however, cases in which the conditions of the one group are approached by the other. In 0smunda the prothallia admit of vegetative reproduction by the detachment of marginal shoots. In Gymnogramme leptophylla, Goebel has shown that the prothallium vegetates in a persistent way, comparable to that of the oophore stage in a moss. The sporophore is very transient in its existence, only living for a few months, and being entirely devoted to the production of spores. Amongst Bryophyta, in Anthoceros the sporophore exhibits a continued production of spores, which places it nearly on a level with Gymnogramme. Farlow has made the remarkable observation that the sporophore may in some cases originate from the prothallium without the intervention of a sexual process, by a process of budding.
2. The Stipulatoe include two orders often included among ferns, but differing from them in some very important points. Ophioglossaceae include Ophioglossum (adder’s tongues) and Botrychium (moon-worts), both widely distributed, and HelmintJωstachys, which is found from the Himalayas to Queensland. The leaves as soon as they have attained a certain age bear a fertile branch from the axial side. The sporangia are developed in a manner totally different from those of other Filicinae. Instead of being pro ducts of single epidermal cells their origin resembles that of the pollen-sacs of anthers. The sexual generation (oophore) is tuberous and subterranean.
Marattiaceae is a group of ferns including about four genera. All are essentially tropical, but run more into the south than the north temperate zone. Hanaea is exclusively American, Kaulfussia and Angiopteτis Old World, Marattia cosmopolitan. The sporangia are placed on lateral veins of the pinnae, and contain two rows of loculi, each of which corresponds to a single sporangium of Ophioglossum, and produces a number of spores originating from many primary mother-cells. The development of the sporangia in Angiopteris, which differs from other Marattiaceae in the sporangia not being confluent, is still a desideratum for the determination of its affinities.
3. Bdùzocarpeae are plants whose affinities have long been obscure, but which are now gradually being seen to be related in many essential points to ferns, though departing from them in other respects. In the growth of the roots, mode of branching, and circinate vernation there is entire agreement, and the fructification, though apparently very different, clearly admits of comparison with the true fern type. For a detailed account of their structure reference must be made to the works referred to below. It will be sufficient here to briefly refer to the fructification. Rhizo- carpeae include two orders, Marsileae and Salvinieae. The former consists of the two genera Pilulama and Marsilea. The former consists of three species of small creeping plants with filiform leaves, one of which is not uncommon in marshy places in Britain; the fruit (sporocarp) is globular, shortly stalked, and springs from the axile of the leaves. The sporocarp is divided by partitions into from one to four cavities, and each of these bears on its outer wall a projecting ridge to which the sporangia are attached. The species of Marsilea are widely distributed mostly in hot countries ; M. macropus is the nardoo of Australia. They all agree in having leaves and long stalks and two pairs of leaflets arranged cross-wise; one pair is developed by lateral branching, the other pair by dichotomy of the growing point. The sporocarp is variously placed, but always in relation to a leaf. Sachs considers that in both Marsilea and Pilularia the sporocarps are modified leaves bearing the sporangia on their upper sides in definite relation to the fibro-vascular bundles. With regard to position he compares them to the fertile fronds of Ophioglossaceae In Marsilea the sporocarp is bean shaped, and the sporangia are attached to transverse ridges in small compartments which are torn away in the peculiar dehiscence of the sporocarp (fig. 13).
The Salvinieae include Salvinia and Azolla. Salvinia is a cosmopolitan genus of five or six species of floating plants distributed over the tropical and warm temperate zones. The sporocarps are developed from teeth of the fertile leaves round which in each case a circular wall grows up covering in the tooth on which the sporangia are developed. This arrangement has been compared to the cup-shaped “involucre” of Hymenophyllum. The sporangia are in all the genera trichomes, and their mode of development is very similar to that in Polypodiaceae. The central mother-cell divides ultimately into eight spore mother-cells. The further progress, however, of these varies according as microspores or macrospore3 are to be formed, each of the eight cells dividing into four in the one case, while in the latter only one does so, and only one of the four daughter-cells attains maturity.
The oophore or prothallial generation in Phizocarpeae is dioecious from the first, the macrospore alone producing archegonia. This, however, only carries to an extreme point a differentiation of which there are indications amongst ferns. In Qsmunda all the spores from a sporangium sometimes produce prothallia with antheridia alone. In Phizocarpeae the differentiation is carried back to the spores themselves. The microspores are male (fig.
14); in Salvinia there is a rudimentary prothallium which produces antheridia, and in Marsilea and Salvinia there is a last trace of such a structure. The macrospores also develop a prothallium from a portion of their contents, which, however, is always for the most part included in the parent spore, which ruptures to expose the archegonia (fig.
15) , and allow of the fertilization of one of them.
II. Equisetinae are plants the habit of which in the asexual stage (sporophore) is very singular. The stems are furrowed and jointed, the joints being hollow and closed at each node by a transverse septum (fig. 16). The leaves are reduced to a whorl of membranous teeth which form a sheath at each node, perforated by the slender branches, which also form a verticil. The teeth of each successive node alternate. The fibro-vascular bundles also alternate in each successive internode, since they are arranged in a ring, and correspond to the ridges between the furrows which above are continuous into the teeth of a leaf-sheath (fig. 17). Each fibro-vascular bundle contains an air-canal, and another series of air-canals alternate with these and correspond to the furrows (fig. 18). The spore-bearing shoots are usually a good deal different in appearance from the barren stems; they terminate in a kind of cone (fig. 19), composed of closely approximated whorls of modified leaves, which are peltate in form, and bear from 5 to 10 sporangia on their under surface (fig. 20). The spores are furnished with an external coat which splits up into four strips, the hygroscopic movements of which actively assist their dispersion (figs. 21, 22). The oophore consists of a prothallium, which is curiously lobed and curled, but otherwise differs little 'from that of ferns.
III. Lycopodinae (Hichotomoe, Sachs).—With a considerable range of morphological variation the different genera agree in some important points. The leaves are always simple and unbranched, and contain only a single fibro-vascular bundle. The branching of stems and roots is always dichotomous, and the dichotomies, with some exceptions, succeed one another in planes at right angles. In their reproductive processes Lycopodiaceae and Ligulatae are related in the same manner as Filices and Phizocarpeae.
1. Lycopodiaceae, besides the club-mosses (Lycopodium), includes three other small genera, Psilotum, Tmesiptems, and Phylloglossum. The first belongs to the warm countries of the new world, and is a singular plant with slender dichotomizing stems, very small leaves, and no true roots. Phylloglosmιm is a minute Australian plant, with a small tuber which produces a rosette of a few long leaves and a stalked spike of sporangia. The species of Lycopodium (fig. 23) attain a more vigorous vegetative development, some almost reaching the size of small shrubs. Their stems contain an axial cylinder, formed by the greater or less coalescence of a number of fibro-vascular bundles forming transverse bands in cross-section. These bands are composed of xylem, with wide scalariform vessels and spiral vessels at each end; between the xylem plates are the phloem common to each, and containing sieve-tubes. The whole surface is surrounded by a phloem-sheath and a fibro-vascular bundle sheath. The cortical tissue surrounding the axial cylinder is often largely converted into sclerenchyma. The sporangia of Lycopodium make their appearance upon the upper surface of the fertile leaves (fig. 24), and, according to Sachs, are developed from the sub-epidermal tissue. The spores are only of one kind (fig. 25), and on germination, according to the observations of Fankhauser, produce a subterranean prothallium, resembling that of Ophioglossaceae, and bearing antheridia and archegonia.
2. Ligulatae includes two orders markedly different in vegetative characters. In both the leaves are furnished at the base with a small appendage called a ligule, wanting in the Lycopodiaceae. Selaginelleae. — The ramification of Selaginella often produces flat frond-like systems of branches, owin~ to the regular suppression of one side of the dichotomies. The leaves, which are arranged in four longitudinal rows, are in these cases often unequal, those attached to the under side of the horizontally spreading stem being larger than those on the upper side. The structure of the stem differs from that of Lycopodium in having no axial cylinder, the fibro-vascular bundles being distinct from one another although transversely expanded like those of Lycopodium. In details of structure they agree with the bundles of ferns. Each is often surrounded by a system of air-cavities. The roots branch dichotomously. The sporangia of Selaginella are shortly stalked, and are inserted in the axil or even in the base of a leaf; the fibro-vascular bundle runs beneath it without sending a branch into it. Macrosporangia (fig. 26), containing usually four macrospores, occur in the same spike with the microsporangia. For a detailed account of the prothallial stage (oophore) of Selaginella reference may be made to Sach’s text-book.
The female prothallium is more distinctly endogenous than even in Rhizocarpeae. It is formed from a portion of the contents of the macrospore (fig. 27), which ruptures partially, exposing the prothallium with its first archegonium. The remaining part of the cavity of the spore becomes filled with a cellular tissue, which may be compared to the endosperm of phanerogams. This is not the only point in which the reproduction of Selaginella approaches that of flowering plants. The oospore first divides transversely, in which respect it differs widely from Filices and Rhizοcarpeae. The upper half forms the suspensor, a structure wanting in all other cryptogams, but characteristic of phanerogams. The lower half forms the embryo, which by the elongation of the suspensor is at first carried down into the endosperm (fig. 28), the cells of which it partially absorbs in its further development. The microspores produce the antherozoids, containing, in addition, traces of a rudimentary prothallium.
The Isoeteae consist of the species of the monotypic genus Isoetes, which, almost always aquatic in habit, are found all over the world. They are all remarkable for the extraordinarily slow growth in length and absence of branching in the stem, which consists of a small tuberous body bearing the rosette of leaves without internodes. It has the unique peculiarity amongst the vascular cryptogams of increasing in diameter by additions to the parenchymatous cortical tissue external to the central fibro-vascular axis. The sporangia of Isoetes are produced at the bases of the leaves, and are unquestionably foliar products; the outer leaves of the rosette produce only macrosporangia, the inner only microsporangia. The macrospores are entirely filled by the female prothallium, the wall being finally ruptured at the apex, as in Selaginella.
Reference may be made to the article Biology for a general view of the relations of the reproductive processes in Pteridophytα to other Cormophyta, and especially phanerogams. See also A Textbook of Botany, Morphological ancl Physiological, by J. Sachs (3d edition), translated by A. W. Bennett and W. T. Thiselton Dyer; Lehrbuch der Botanik, von Dr Julius Sachs, vierte Auflage; Vergleichende Anatomie der Vegetationsorgane der Phanerogamen und Farne, von Dr A. de Bary; Synopsis Filicum, by Sir W. J. Hooker and J. G. Baker. (W. T. T. D.-J. G. BA.)
