ALBERT R. MANN LIBRARY AT CORNELL UNIVERSITY Cornell University Library 523.S65H istoria filicum; an exposition of the na Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924000634448 HISTORIA FILICUM;, AN EXPOSITION OF THE NATURE, NUMBER, AND ORGANOGRAPHY OF FERNS, AND REVIEW OF THE PRINCIPLES UPON WHICH GENERA ARE FOUNDED, AND THE SYSTEMS OF CLASSIFICATION OF THE PRINCIPAL AUTHORS, WITH A NEW GENERAL ARRANGEMENT; CHARACTERS OF THE GENERA}; REMARKS ON THEIR RELATIONSHIP TO ONE ANOTHER}; THEIR SPECIES; REFERENCE TO AUTHORS ; GEOGRAPHICAL DISTRIBUTION ; ETC., ETC., With 30 Ltthographic Plates, by W. H. Frrcu, F.LS., illustrating the Characters of the Tribes. By JOHN SMITH, A.LS., Ex-Cugator oF THE Royat Botanic GarDEN, Kew; AvtHor or ‘' Ferns, BRITISH AND Forrien,”’ ‘‘Domestio Botany,” ETC., ETC. LONDON. MACMILLAN & CO. 1875. i LONDON : H. M. POLC{ETI, HORTICULTURAL PRINTER, 12 10 15 BRIDGEWATER GARDENS, BARBICAN, E.C. PREFACE. —~—_ In submitting this volume to the notice of those in- terested in the study of Ferns I consider it proper, though at the risk of being considered egotistical, to give a brief explanation of the circumstances that have led to its publication. My first introduction to Ferns was in acquiring the names of the common British species. In 1823 the _collection in the Royal Botanic Garden, Kew, came under my care; it then consisted of about forty hardy species, British and Foreign, and about the same number of tender exotics, the latter dispersed in various hothouses. In 1825 I arranged the tender ones in a group at the end of one of the then lean-to houses, the space they occupied being 12 feet by 6 feet; these formed the nucleus of the present great collec- tion. New species were occasionally imported, and others raised from spores, the spores being obtained from collections of dried specimens, chiefly from the West Indies, Brazil, and Australia, also from a collection of d i iv PREFACE, Indian specimens presented to me in 1829 by the late Dr. Wallich, which led to my commencing a Fern Herbarium, that first received a special addi- tion in the form of a collection of Jamaica Ferns, collected in that island by Mr. Robert Heward. In 1841 I purchased a set of the great collection of Ferns made in the Philippine Islands by Mr. Hugh Cuming; of them I drew up a list, published in Hooker’s “Journal of Botany.” I then wrote a Memoir, entitled, “An Arrangement and Definition of the Genera of Ferns,’ which was read before the Linnean Society, and published in Hooker’s “Journal of Botany for 1841-42.” In 1842 Hooker and Bauer’s “Genera Filicum” was completed, wherein I described twenty new genera. Collections of dried specimens were occasionally placed in my hands for naming, among which may be mentioned Schomburgh’s “Ferns of British Guiana,” published in Hooker’s “Journal of Botany” for 1842. The “Ferns of New Zealand,” in Dr. Hooker’s “Flora of New Zealand,’ in 1854. Ferns collected during the voyage of the surveying ship “Herald” by Dr. Seemann, contained in the volume of the Botany of that voyage in 1854. I also drew up a list of tho great collection of Ferns made in India by Drs. Hooker and Thomson, PREFACE. v My Herbarium continued to increase, not only by the addition of specimens from cultivated plants, but also of dried specimens received from the principal Fern regions of the world, either obtained by direct purchase or by presentation. The collection, number- ing about 2,000 species—their various sizes and forms occupying 6,000 large folio sheets—were, in 1866, with the exception of one-third duplicates, purchased for the Herbarium of the British Museum. In 1846 the Kew Garden collection had increased to 400 species, of which I made up a list, published in the volume of the “Botanical Magazine” for that year. This was followed in 1857 by the publication of. a “Catalogue of Cultivated Ferns,’ 600 species being enumerated. In May, 1864, in consequence of failing sight, I resigned the curatorship of the Royal Botanic Garden. I at that time was preparing another catalogue of the Ferns in the Kew collection, which was published in 1866, under the title of “Ferns, British and Foreign,”* consisting of their History, Organograpy, and Classification, including a Treatise on their Cul- tivation. In this work 164 genera are described, each illustrated by a woodcut, and 1,084 species enumerated, including 56 Lycopods and their allies. The Kew * A second edition is now in preparation. vi PREFACE collection at this time occupied three hothouses, cover- ing an area of 6,570 square feet. I should, however, state that the greatness of this collection could not have been attained but for the powerful aid of the late Sir William Hooker, who also took great interest in the study of Ferns, as will be seen by the references to his works in the following pages. In the preface to the “Synopsis Filicum,” speaking of the Garden collection, he says, “it is the finest in cultivation,” and “the formation of this fine collec- tion is mainly due to the exertions and ability of Mr. John Smith, who for forty-three years held an important position among the officers of the Royal Gardens. His knowledge of Ferns, and his writings on them, justly entitle him to rank among the most distinguished Pteridologists of the present day.” The above statement will, I trust, be sufficient to show that I have not written this work without long practical acquaintance and the devotion of much time to the study of Ferns, this leads me to hope that it will be found useful to, and worthy of the patronage of, Pterodologists and Botanists in general. I must, however, state that on account of my loss of sight the work is not so complete as I could wish. Lat- terly my materials have been obtained from books kindly lent me by Dr. Hooker, and from various PREFACE, vii notes furnished by Mr. J. G. Baker, assistant Curator of the Kew Herbarium. The original MS. was nearly ready for publication in 1864, since when it has been rewritten and revised under my direction, and before being sent to press submitted to Mr. James Britten, of the Botanical Department of the British Museum. The proof sheets have been (with the exception of a few of the early ones) corrected by Mr. Robert Heward, F.LS., I therefore trust that the typographical errors, &c., which may have escaped notice are few. I should not have ventured to have added the plates to the work if it had not been that Mr. W. H. Fitch, the eminent botanical artist undertook to prepare the drawings, and to superintend them through the press; from having drawn all the figures of Ferns published by Sir W. Hooker in his several works, he has become well acquainted with the structure and generic character of Ferns. We have also for many years at various times exchanged views regarding special points of struc- ture, I have, therefore, no hesitation in accepting the plates as correct. Although this work is entitled the History of Ferns, it nevertheless must be understood that it does not describe species, nor does it treat of their microscopical structure or chemical products, the chief aim of the vill PREFACE, work being the definition of Genera, and their classifi- cation, founded on the different modes of growth, vena- tion, and fructification. JOHN SMITH. Kew, September, 1875. CONTENTS. PART 1. EXPLANATION OF THE NATURE AND STRUCTURE OF THE PARTS OF FERNS ON WHICH GENERA ARE FOUNDED AND CLASSIFIED. PAGE. 1. Nature and number of Ferns. ‘ s : 1 . Organography.—Definitions of the ibs of Rene on which bo ao genera are founded and classified ‘ 3. On the genera of early authors, and their yates of classi- fication. ‘ : 5 : : ‘ c~ 20 4. On the genera of modern authors, and review of hus systems of classification : . : : : y 5. On the genera and their classification as fatthide on their modes of growth and habit . 3 : f : x Gi PART 2. General arrangement and characters of tribes and genera, with remarks on their relationship one to another, and the number of species. : : : : ‘ 77 to 370 PART 3. 1. On the reproduction of Ferns : ; : é . 371 2. On the geographical distribution of ies : : ; . 877 3. On the properties and uses of Ferns. : : : . 3885 PART 4. 1. Names of writers on Ferns, and their works . ; 396 2. Etymology.—Derivations of generic names adopted in this work . . 4 . 409 3. Alphabetical list of genera of suiinis not eee “ . 422 4, Index to the tribes and genera adopted in this work . 427 NAMES OF THE FIGURES ON THE PLATES, ILLUSTRATING THE TRIBES. PLATE 1. a. Oleandra neriiformis. 6. Oleandra Wallichii. Oss.—With a portion of the caudex of Polypodium vulgare, show- ing the articulate vernation characteristic of the Division Eremobrya. PLATE 2. a. Humata heterophylla. 06. Davallia pyxidata. c. Leucostegia immersa. PLATE 38. a. Polypodium vulgare. 6. Phlebodium aureum. c. Pleuridium crassifolium. PLATE 4. a. Drymoglossum carrosum. 0. Hymenolepis spicata. c¢. Lepto- chilus decurrens. PLATE 5. Platycerium Stemmaria. PLATE 6. a. Elaphoglossum latifolium. 0. Polybotrya osmundacea. c, Gym- nopteris nicotianzfolia. PLATE 7. a. Meniscium serratum. 0. Meniscium simplex. e. Jenkensia undulata (Pecilopteris sp., see Obs., p. 136). PLATE 8. a. Gymnogramma calomelanos. 0. Hemionitis cordata. ¢. Llavea cordifolia. Xil LIST OF PLATES. PLATE 9. Ceratopteris thalictroides, showing sporangia and spores magnified. PLATE 10. a. Vittaria elongata. 8. Vittaria scolopendrina. ¢. Pteropsis angustifolia. d. Dictyoxiphium Panamense. PLATE 11. a. Monogramma fureata. 0. Pleurogramma linearis. ¢. Xipho- pteris serrulata. d. Grammitis furcata. ¢. Calymmodon cueullatus. jf. Calymmodon clavifa. PLATE 12. a. Ctenopteris trichosora. 0. Ctenopteris subtilis. ¢. Ctenopteris moniliforme. d. Ctenopteris peruviana. ¢. Ctenopteris rigescens. f. Ctenopteris discolor, g. Adenophorus tamarisci. /, Adenophorus tripinnatifidus. PLATE 13 A. a. Phegopteris macroptera. 0. Goniopteris crenata. c. Dipteris Wallichii. PLATE 13 B. a. Aspidium trifoliatum. 0b, Nephrolepis exaltata. c. Lastrea patens. PLATE 14. a. Woodsia caucasica. 6. Woodsia obtusa. c. Diacalpe aspidioides. d. Peranema cyatheoides. PLATE 15. a. Cyathea arborea. 0b. Hemitelia horrida. c¢. Alsophila excelsa, PLATE 16. a. Dicksonia arborescens. 6. Cibotium Schiedei. ¢. Cystodium sorbifolium. d. Thyrsopteris elegans. LIST OF PLATES, xi PLATE 17. a. Saccoloma elegans. b. Dennstadtia cicutaria. ¢. Microlepia platyphylla. d. Deparia prolifera. PLATE 18. a. Lindswa trapeziformis. 6. Schizoloma ensifolia. _c. Synaphle- bium recurvatum. d. Odontoloma repens. PLATE 19, a, Adiantum Capillus-Veneris. 6. Adiantum obliquum. c. He- wardia adiantoides. d. Ocropteris pallens. PLATE 20. a. Nothochlena squamosa. 0. Cheilanthes radiata. c. Cheilanthes fragrans. d. Cassebeera triphylla. PLATE 21. a. Pteris nemoralis. 6. Campteria biaurita. c. Doryopteris hastata. d. Lonchitis pubescens. PLATE 22. a. Blechnum occidentale. 0. Woodwardia radicans. c. Lomaria striata. PLATE 238. a. Asplenium serratum. 0. Neottopteris Nidus. ¢. Diplazium plantagineum. d. Hemidictyum marginatum. PLATE 24. a. Platyzoma microphylla. 6. Gleichenia microphylla. c. Gleich- enia gigantea. d. Matonia pectinata. PLATE 25. a. Hymenophyllum tunbridgense. 6. Trichomanes alatum. c. Hymenostachys diversifrons. In the sterile fronds the veins anasto- mose, which has been inadvertently omitted to be shown on the plate. d, Féea spicata. ¢. Trichomanes pinnatum. f. Trichomanes reni- forme, ly 2 IST OF PLATE ,s PLATE 26. a. Schizea dichotoma. 0. Anemia mandioccana. c, Anemidictyon Phyllitidis. d@. Lygodium volubile. PLATE 27. a. Osmunda regalis. 6. Todea Africana. PLATE 28. a. Angiopteris evecta. 0, Marattia purpurascens. c. Danea alata. PLATE 29. a. Ophioglossum vulgatum. 6. Botrychium lunaria. ¢. Helmin- thostachys zelanica. N.B.—In many of the figures the sori, sporangia, and spors are shown more or less magnified. Page 49, 49, 56, 63, 64, 77, 85, 86, 90, 91, 93, 93, 94, ERRATA. line 13, for Lomogramma 7cad Lomagramma. 138, for Miniscium read Meniscium. 24, fom Fadyena read Fadyenia. 9, for Sellegua read Selliguea. 22, for Saromanes read Soromanes. 18, for Davallie read Davalliece. 25, for palmatafid read palmatifid. 26, for obligatum read obliquatum. 2, for hypogaceous read hypogeus. 26, for Synaminia read Synammia. 8, for indusaform read induseform. 28, for papyaceous read papyraceous. 28, for Anexetum read Anaxetum. 18, for adnescens read adnascens. 2, for specie read species. 16, for Thumb. read Thunb. 3, for tenulorus read tenuiloris. 9, for Borg. read Bory. 5, for peltedia read peltidea. 7, for tubicles read tubercles. 22, for ragadiolepis read rhagadiolepis. 25, for to read from. 24, for cordex read caudex. PART TI. EXPLANATION Of THE NATURE AND STRUCTURE OF THE PARTS OF FERNS ON WHICH GENERA ARE FOUNDED AND CLASSIFIED. 1.—NATURE AND NUMBER OF FERNS. Tue plants which form the subject of this volume con- stitute a special Order in the Vegetable Kingdom, termed Filices, popularly known by the name of Ferns; they are plants of special habit and structure, and rank highest in the class Cryptogamia of Linnezus. From circumstances explained hereafter, the number of species representing the present Fern flora is not well ascertained ; it is sufficient here to say that about 2,500 have been described in books by different botanists, speci- mens of nearly that number being preserved in the princi- pal herbaria of Europe, of which about a thousand have been introduced in a living state, and cultivated in the the gardens of this country. They are widely spread over the earth, being found in the tropics and temperate regions, and extending nearly to the limits of vegetable life in both the northern and southern hemispheres. In habit and mode of growth the greater number come under the category of herbaceous perennials, having seasons of erowth and rest according with the nature of the dif ferent climates they inhabit. B bo ON GENERA AND SPECIKS. They vary considerably in size, many not being larger than mosses and liverworts and of as delicate a texture ; a very few partake of the character of shrubs, and about one hundred and fifty assume the aspect of trees, some attain- ing the height of fifty or more feet. As the special character of their structure is fully ex- plained further on, the above is sufficient to give an idea of the nature of Ferns. Although Ferns possess but few economic properties, either as food, medicine, or in the arts, they have never- theless become of great interest to botanists, their study being termed Pteridology, on which many books have been written, and during the last fifty years have also come into special favour with amateur plant cultivators, and lovers of pretty plants. The constant desire for novelty has led both public and private plant collectors and persons living in foreign countries to devote special attention to the col- lecting of Ferns, both for herbarium specimens and also as living plants. By these means the collections of Europe have become greatly enriched. It is recorded that the Royal Herbarium at Kew contains 50,000 specimens of Ferns, which have been communicated by no less than about 150 public, private, and trade collectors from all regions: the Fern collection in the herbarium of the British Museum is also very extensive. These, with the living collection at Kew, which in 1864 contained above one thousand species, have afforded ample means for study and determining the genera and species of Ferns. Many extensive living collections are also formed by amateurs, who in rivalry enhance their value to such an extent that as much as fifty guineas is frequently given for a single plant of fashionable repute at the time. Of late years many books have been published on Ferns, ON GENERA AND SPECIES. 3 but the different views entertained by botanists, respecting the data on which genera and species are founded, renders the study of Ferns a task of no ordinary labour. This is owing to many causes, one of which is found in the fact that many species are normally heteromorphous ; separate plants of the same species presenting different forms at the same or at different periods of growth, although grow- ing under the same climatic influences ; in some cases dif- ferent fronds of the same plant, and even portions of the same frond when seen apart, have been characterized as distinct species, and have even been placed under different genera. It also frequently happens that two or more speci- mens present sufficient characters to be considered distinct species, but when the intermediate gradations of the series are supplied, only the most extreme states of any of the forms appear sufficiently distinct to be so con- sidered. It often happens that plants of the same species at one time produce all simple, and at another time all pinnate or more highly divided fronds; it is therefore not surprising that different forms of individual species should be viewed as representing two or more species when known only from incomplete herbarium specimens. It must also be borne in mind that the descriptions of many species found in books are derived from dried speci- mens, obtained in various parts of the world by trade collectors whose object is often only pecuniary gain, and whose profits are dependent on the number of forms found. Great caution is therefore necessary in accepting a collector’s idea of species; it is only on the word of the independent and truly scientific collector that we can rely for correct information on the various aspects assumed by individual species growing in regions far away from where we write their history ; but even under the most favourable circum- 4 ON GENERA AND SPECIES, stances of observation and discrimination, uncertainty often prevails. For instance, no class of plants has of late years been more written about than the Ferns of Great Britain; and yet much diversity of opinion still exists amongst British botanists regarding the number Of species, although they have the opportunity of carefully and leisurely examining them under every circumstance con- nected with the different aspects they assume in their various places of growth. This surely offers some pallia- tion for the errors committed by the general pteridologist, more especially when we consider that the Ferns of Great Britain are in number as one to sixty of the Fern-flora of the earth. Having had under my observation for nearly half a cen- tury the Fern collection in the Royal Gardens at Kew, which, as already stated, amounted in 1864 to about a thousand species, and having during that period carefully noted their different modes of growth and aspects, I have been induced to consider that the differences seen in a series of closely allied living Ferns, are much more obvious than in examples of the same when seen in the herbarium. It is, however, often difficult to express their differences in words, but the eye readily distinguishes them, and knowing that each form maintains its own peculiar habit from year to year, reproducing its like from spores (seeds), I am, therefore, induced to consider that althongh closely allied, they are nevetheless sufliciently different to be regarded as distinct species. Another point which renders it difficult to arrive at any satisfactory conclusion as to the correct number of species, is the botanical rule which prescribes that the specific name under which a plant is first described, is to be retained in whatever genus it may be referred to by ON GENERA AND SPECIES. 5 subsequent authors; and in consequence of many species of Linnzus, Swartz, and other early authors, being but indifferently described, many being founded on imperfect specimens, and nothing but the meagre description left us for their identification. It frequently happens that some modern author detects or supposes that he has found out that the new species of his contemporary is one of the Linnean or Swartzian doubtful species, and on his decision being confirmed, names long familiar have to be changed. Again, by the distribution of trade and private herbarium collections, sets of the same specimens fall in the hands of different botanists at the same time, and each being eager to describe what he considers new, and not, in some in- stances, having studied the special families, nevertheless does not hesitate to describe new species, or to found new genera. This being done without inter-communication, different names are given to the same plants ; when these are collated by the systematist, it often becomes difficult to decide which to adopt, thus burdening the science with synonyms. Under this state of things few Ferns have now less than three names, many have twenty, thus rendering it in many cases impossible to reconcile one author’s views with an- other’s. For instance, the Fern known by the name of Ceratopteris thalictroides has been described by no less than twelve authors under as many different generic and specific names. But the point upon which pteridologists appear to differ most, and upon which their only agreement seems to be an agreement to differ, is the definition of genera and their limits. I say emphatically appear to differ, for in the works of those most at issue the differences are not so much in the limits of the groups themselves as in the 6 ON GENERA AND SPECIES, relative importance assigned to them. For example, while some, as Presl, Fée, Moore, and myself, break up the old Linnean genera, Polypodium, Aspidium, &c., into a greater or lesser number of smaller genera, upon characters derived from differences in their anatomical structure and modes of growth ; others, as Hooker and Mettenius, prefer adhering to the Linnean genera, without greatly altering their characters, and adopting the modern generic names as sectional ones for such divisions as they find themselves compelled to make. On reviewing what I have now stated it may naturally be asked, What is a species or genus ? or, by what law of nature can this be determined ? As generally under- stood by naturalists, a species is an organised structure specially created and endowed with an essence or quality peculiar to itself, possessing the power of increase and transmitting its primitive essence and anatomical structure and form without change, to its progeny for successive generations. But the difficulty of defining species becomes evident on taking a general view of the numerous forms which connect one with another. It will then be found be- yond human power to ascertain whether the several grada- tions of allied forms are descendants of primitive specific creations, or are, according to the Darwinian theory of the “origin of species,” only deviations from a few primordial creations, endowed with a protean principle which becomes manifest during the lapse of ages, and controlled by the different climatic and local influences under which the progeny of the original have become established, and which now form the flora of the earth. If the latter is admitted to be the case, and we are led to believe that intermediate forms originate during the slow progress of time, then all must be uncertainty, and the number of ON GENERA AND SPRCIES, 7 species will vary according to the evidence possessed by different authors, and the views they take of the apparent transition forms. In verification of what is stated above respecting species, I quote the words of an eminent botanist, that “the pre- liminary exact determination of what is a species no one has yet succeeded in giving.” It is well known that phe- nogamous plants assume different forms and aspects brought about by natural causes or the agency of man; the differ- ence from the original types being often so great, that if evidence of the change were not on record, the botanist of the present day would be justified in describing them as distinct species, and originally created as such. With Ferns, we possess but little evidence of new forms having come into existence in modern times, the chief examples being found in several intermediate states in the genera Gymnogramme and Pteris, which of late years have made their appearance in gardens, some being so different from their supposed parents that, if they had been found wild they would by some botanists be considered sufficiently distinct to rank as species. If such changes do actually take place, and we are to deduce from them that races of intermediate forms originate in the progress of time, and through the causes above alluded to, great difficulty must attend any attempt to define species of Ferns. These observations show the principal conflicting causes that have led to the confusion in the nomenclature of Ferns. This may be considered a sufficient history of species. I will now treat of genera and their classification ; but to make the subject well understood, I deem it best to give first a general exposition of the structure of the different parts of Ferns, on which genera are founded and classified. 8 ON GENERA AND SPECIES, 2.—ORGANOGRAPHY. DEFINITIONS OF THE PARTS OF FERNS ON WHICH GENERA ARB FOUNDED AND CLASSIFIED, Fitices or Ferns are flowerless plants, and form the highest order of the division of the Vegetable Kingdom termed Cryptogamia, which includes all plants having their organs of reproduction invisible to the naked eye. They have no true leaves, but produce leaf-like expansions, called fronds, which not only perform the functions of leaves, but also bear the organs of reproduction. The fronds are succes- sively developed from the apex or sides of an accrescent stem (caudex), and before expansion are spirally coiled inwards (cireinate). They are traversed by veins in various ways, and produce on their under surface, or on special appendages, round linear, or irregular masses of one-celled (sporangia) or many celled (synangia) cases, which contain numerous microscopic germs, called spores (seed). The masses of spare cases are called sori. They are either furnished with a special covering of various forms, called the indusium (indusiate), or they are naked (non-indusiate). VERNATION (stems). The manner in which the fronds are developed from their axis is termed vernation, and their union with the axis is either adherent or articulate. Adherent vernation presents two forms. First—Fasciculote when the fronds are produced in a continuous spiral whorl from the apex of the axis with which their bases are adherent, and thus by their ON GENERA AND SPECIES. 9 successive development form a stem (caudex), which is either erect or more or less decumbent. The caudex varies in size, according to the normal nature of the species. It is often very short, even consisting of a bud only, scarcely rising above the surface of the ground—such is termed acaulose—or it forms a low stem a foot or more in height (arboroid), or it assumes the height of a tree (arborescent). Second—Uniserial. This consists of a creeping rooting stem, termed sarmentuwm, which progresses by the succes- sive evolution of single fronds from its apex, each succeed- ing frond being produced from the base of the preceding one, at more or less distance in advance. It either con- tinues simple or branches, and varies in thickness from that of a thread to an inch or more in diameter, and is either epigeous, hypogeous, or epiphytal. Articulate vernation also presents two forms. First— Fasciculate. This is peculiar to a few Tree Ferns and the order Marattiacee, in which the fronds when mature fall away, leaving marks of rhomboid scars on the stem.* Second — Solitary. This consists of a creeping stem, termed surculum, emitting roots from its under surface, by which it adheres to the earth or other bodies, but generally trees (epiphytal), rarely hypogeous. It varies in size from short and thick to long and slender, and when young is densely covered with scales (squame). It is either simple or branching, and is distinguished from the sarmentum by its apex being always in advance of the forthcoming fronds. These are produced singly from special nodes on its sides, with which the base of the frond (stipes) is articulate, and there separate on becoming mature, leaving the mark of a round, slightly concave scar or Cicatrix. * See Tribe Cyathea. 10 ON GENERA AND SPECIES, FRONDS. Tue fronds before expansion are involutely coiled, in the manner of a watch spring, and gradually uncoil during the period of growth, and with a few exceptions have all their parts formed before they begin to unfold (definite), all of which increase in size with the progress of the frond. Fronds present extreme variation in size and form, vary- ing from less than an inch to 15 or 20 feet in length, and in form, from entire, like a blade of grass or ribbon to being compoundly divided. In describing them, the same terms are used as are applied to the leaves of flowering plants, that is, they are simple, entire, linear lanceolate, ovate, elliptical or sub-rotund, or they are pedate or pal- mately lobed, pinnatifid, or they are pinnate, bi-tripinnate, or more highly divided, and then termed compound multifid. They also vary considerably in texture, some being thin, pellucid, ‘membranous, and flaccid ; while others are thick, fleshy, or hard and rigid. In pinnate, bi-and-tripinnate fronds, the divisions (seg- ments) of the first are called pinnew and the second pin- nules, each of which individually represents a simple frond, The margins of simple fronds, and of pinne and pinnules, are either entire or variously dentate or sinuose repand, regularly pinnatifid or otherwise divided, the divi- sions being called lobes or lacine. The general form of fronds, whether simple or in any way divided, is linear lanceolate, or oblong elliptical, or deltoid ; in deltoid com- pound fronds, the primary pinnew assume the character of branches, the ultimate divisions of which are often divided into small lacine or dents, such being termed multifid. The axis of attachment of the pinne and pinnules is ON GENERA AND SPECIES. 11 termed the rachis, the primary rachis being a continuation of the foot-stalk (stipes); its first branches the secondary rachis, that is the rachis of pinnules. The pinne and pin- nules are as a rule alternate, but are also often opposite or nearly so, and they are either sessile or petiolate, and are either adhering and continuous with the rachis, or arti- culate with it, readily separating at the joint when mature, or earlier by dryness. In such the rachis remains long permanent, and even in some species of Nephrolepis the apex continues to grow, producing pinne. Their leafy part either tapers down to their base, such being termed sessile, or borne on a cylindrical or semi-cylindrical foot- stalk, but which is not special, being an integral part of the frond. They either grow erect or verge in various angles from the axis of development, some being nearly horizontal or bent downwards, their apex in such cases being generally viviparous. Fronds before evolution are densely covered with scales, termed ramente, or articulate hairs, these in most cases are early deciduous, except at the base of the fronds. In Pleo- peltis, Lepecystis, and Ceterach, and some others, they are permanent, and specially characteristic. In some the under side is densely covered with silvery or golden- coloured farina ; others are furnished with various hairs or glands, which are generally viscid. Fronds are in most Ferns of two forms, sterile and fertile, but in the great majority the latter do not differ much from the former, though they are in general narrower in all their parts. In several tribes they are normally of two distinct forms (dimorphous), the sterile presenting the ordinary leafy character, the fertile being specially contracted and thick- ened. In some the contracted segments are very narrow 12 ON GENERA AND SPECIES. (rachiform), in others they are involute, in the form of a pod or siliqua (siliquiform); in others plain, the under surface being densely occupied with sporangia ; in some, the lower portion of the frond is sterile, while the upper portion is contracted and fertile. The duration of fronds is variable, some remaining per- manent for several years, but the greater majority, espo- cially those in temperate and cold climates, develope and come to maturity in one year. It is, however, otherwise with the genera Gleichenia, Lygo- dium, and Salpichlena; in the first the fronds are dicho- tomous; and from the axis of the forks are successively produced other dichotomous branches, and the fronds thus continue long permanent. As regards Lygodium, the fronds so called continue to lengthen and branch (indefinite), rambling over bushes or climbing the tops of lofty trees ; such is also the case with the blechnoid Fern Sulpichlena. The permanent nature of these climbing fronds (?) may be considered analogous to climbing flowering plants, such as Clematis; to these peculiar growing Ferns I apply the term Pterampelids. VENATION. The vascular structure of Ferns consists of cords of indurated tissue termed scalariform, which branches from the base of the previous frond into the nascent bud of the forthcoming frond, and lengthen upwards with the growth of the frond. It consists of one, two, or more cylindrical, flat, or channeled cords, which, on reaching the leafy part of the frond, divide and ramify through all its parts, the ul- timate ramifications ending in the leafy part (lamina), being ON GENERA AND SPECIES. 13 called veins; these veins, which oither radiate from the base of some fronds as also from segments, in a flabellate manner; but are more generally produced from a midrib (costa), which in simple fronds is a continuation of the vascular cords of the stipes; and in segments of fronds, a branch of the rachis from which they are produced. The costa is generally central, or in some excentric, or even quite on one side of the segment (unilateral) ; it generally terminates at or near the apex of the frond or segment, or it becomes undefined before reaching the apex, or is very short, rudimentary, or absent. It is generally elevated in the form of a ridge on one or both sides of the lamina, or it is in the form of a channel on the upper side, or it is imbedded in the substance of the frond, and then often obscure. From the sides of the costa veins are produced at a more or less distance from each other, and generally equal or nearly so, on both its sides; they form a more or less acute, oblique, or nearly a right (patent) angle with the costa, and when free always directed outwards towards the margin or apex of the frond or segment. They are either straicht, curved or flexuose, and are either simple or branch in various ways, the first are termed primary veins, their branches venules, and the branches of these vetnlets, each being a diminutive of the preceding. In many cases the primary vein that rises direct from the midrib is very short, and branches sometimes close to the midrib, or even within the substance of the midrib, as in Neottopteris australiaca ; in others they terminate at or near the margin, and when parallel to one another and very evident are termed coste- form, and the diminutive costwle is applied to the primary veins of laceanee. Veins are termed free when all the branches of each 14 ON GENERA AND SPECIES. fascicle terminate before reaching, or at the margin, their apices being free and unconnected ; and anastomosing when the apices of the branches of each proximate fascicle unite with one another, which they do in various ways, forming regular, square, rhomboid, hexagonal, or unequal sided meshes (areoles). When the primary veins are not evident, and all the anastomoses of about equal size, it is termed reticulate venation, Both free and anastomosing veins are spoken of as follows :—Simple, when each vein proceeds from the costs to the margin without branching. Forked, when they divide at an acute angle into two or more branches, after leaving the costw. Pinnately forked, when the primary veins are scarcely defined, the venules dichotomously branching. Pinnate, when the primary veins that run from the costs to the margin are distinctly defined, and produce venules in regular order on both sides, forming a more or less acute or oblique angle with their axis. The simplest form of anastomosing venation is when the apices of simple or forked veins are combined or connected by means of a vein close to and continuous with the mar- gin, as in the genera Olfersia and Neottopteris. Pinnate and pinnately-forked veins anastomose by the uniting of the apices of the venules of the proximate fascicles, which they do at various angles, from acute to nearly a right angle, or even arcuate as in Menisciwm, or straight asin Blechnum. In most of these cases, the connected venules produce from the apex of their angular junction, or from their exterior sides, one or more outward directed veinlets (excurrent), which are either free or unite with the next exterior anas- tomose, thus forming in the latter case two or more rows of areoles between the primary veins. When anastomose venules form a right or nearly a right angle with the ON GENERA AND SPECIES, 15 primary veins, they thus become parallel with the costa, and are then called transverse. In many cases the venules produce veinlets in an irregular manner, which either anas- tomose, or terminate within the areoles. When directed outwards they are termed eacurrent, and recurrent when directed inwards; this is termed compound anastomose venation. In many cases, both in free and anastomose venation, the apices of the venules and veinlets is thickened (clavate) in the form of a gland. RECEPTACLES. The receptacles are the sites upon which the sporangia are seated, and are generally thickened points, or length- ened portion of some part of the venation. In position they are called terminal when on the apex of the veins or their branches ; lateral when on the sides of the veins ; avillary when on the forking ; compital when on the angular crossing or points of confluence of two or more venules or veinlets. They are superficial or inumersed in the substance of the frond, or elevated above its surface, and then columnar or globose. When in the form of a point or dot they are termed punctiform, and, when lengthened, elongated or linear; and amorphous when of no defined form (as in Acrostichum). SPORANGIA. The spore cases or sporangia are the organs which con- tain the reproductive spores, and are borne in masses upon the receptacles. They are thin and transparent, or horny 16 ON GENERA AND SPECIES. and opaque. In form they are globose, oval or pyriform, pedicellate or sessile, and are either furnished with a more or less complete articulated elastic ring (annulate), or des- titute of a ring (evannulate). This ring is said to be vertical when it passes over the apex of the sporangia, being a continuation of the pedicel; horizontal when it passes horizontally round the sporangium, either at or about its middle, or at the apex (apical); oblique when it passes round the sporangium in some direction interme- diate between vertical and horizontal. When the sporangia arrive at maturity, and are under certain conditions as to dryness, the elasticity of the ring causes them to burst open, and the spores are ejected. The opening takes place at or very near to a right angle with the direction of the ring—that is, those having a ver- tical ring burst horizontally, and those with a horizontal ring, by a vertical slit. Ezxannulate spore cases are in texture coriaceous becom. ing horny. They are either free, or counate and called synangia, which are formed of a greater or lesser number of exannulate sporangia, laterally united, forming a series of cells, disposed in a circle or in two rows, side by side, which either remain united as in Dancee, or separate longitudinally in two valve-like lobes, as in Marattia. The cells open by slits or pores. N.B.—About thirty species which constitute the order Marattiacee have exannulate sporangia. SORI. The sori are masses of sporangia produced on the recep- tacle. Their form and position correspond with, and are ON GENERA AND SPECIES. 17 dependent upon the form of the receptacles, which are their foundations. Thus, when the receptacles consist of a point only, the sori are always round, termed punctiform while linear receptacles bear sori of various forms, being oblong, elliptical, or linear, either simple, straight, or arcuate, forked or reticulated. The sporangia of each sorus are generally crowded, or, in some cases, few, being side by side on the receptacle, as in Ceratopteris and Gleichenia. They are either seated ou the margin (marginal) or near or close to the costa (costal), or on or about the middle of the disc (medial). When a little within the margin they are termed ante-marginal. In some they are in pedicellate dents, exserted on the margin. They are irregularly scattered over the surface or solitary in multifid lacinie, or in regular rows, or con- tinuous lines (serial). When the lines diverge at an angle from the mid-rib they are said to be oblique, and when parallel with either the margin or the costa, transverse. For their position on the veins the same terms are used as those given for the receptacle. As a general rule, each sorus is distinct and well defined, but in many cases the receptacles are so very close together that the sporangium of one sorus runs into that of the con- tiguous one (conjluent), and thus normally punctiform sori become oblong or linear. In the tribe Acrostichec, in which the fertile fronds are contracted, the sori have uo definite form, the sporangia occupying the whole of the under surface or in patches. They are borne on one side of con- tracted spikes, or on simple or compound racemes, as in Schizea, Lygodium, Anemia, and Osmunda. Cc 18 ON GENERA AND SPECIES. INDUSIUM. The sori are either naked or furnished with a special covering called the indusium, or involucrwm by some, which presents many well marked different forms, and is highly important for distinguishing genera; it consists of three kinds— special, accessory, and wniversal. True or special indusia are more or less membranous, or thick and firm in texture, and rise from the receptacles to which they are attached in different ways. Some rise in the form of an orbicular disc, covering the sorus in the form of a shield (peltate) ; more frequently they are attached on the interior or exterior side of the receptacle ; such are termed lateral, which open like a lid (operculéform) when attached on the side next the costa, interior, and when on the side next or on the margin evterior. Lateral indusia partake of the form of the sori, being round, oblong, or linear; in punctiform sori they are attached by a point, or more or less by the base or sides, and are of various forms, being reniform, subrotund, oval, or oblong, sometimes very small (sqguamiform). Their surface is jlat, vaulted, or when attached by their sides cucculate, their free margin being entire or variously lacinated or fringed. Besides these two modes of attachment there is a third kind, in which the indusium is attached all round the recep- tacle—it is at first globose, and includes the sporangia, but ultimately their apex opens and widens, thus assuming the form of a cup (calyciform), the margin being more or less entire or deeply laciniated, even to be like a fringe; some- times the attachment is only half round the receptacle (semicalyctform). ON GENERA AND SPECIES. 19 Accessory indusia are formed of the entire margin of the frond, or a portion of it being changed in texture and form and inflexed, its margin conniving with the free margin of the interiorly attached special indusium, the two thus combined forming a continuous or interrupted groove, opening exteriorly. When the special indusium is pro- duced on a punctiform receptacle, an wreeolate, bilabiate, or tubular cyst is formed, which contains the sporangia, some being exerted beyond the margin (pedicellate) ; while in some cases the cyst is antemarginal, and is then termed cucullate, in all these cases opening exteriorly, but often more or less turned inwards, and then having the appearance of being on the under side of the margin. Universal indusia occurs when the segments of the fertile fronds are contracted. They consist of the margins of the segments being more or less changed in texture, becoming membranous, and folded inwards (involute*), the two opposite margins generally meeting, and thus enclosing all the sori upon the segment, whatever may be their form or position upon the veins, the segments thus assuming the form of a pod or siliqua (siliqguiform) asin Struthiopteris ; or of a berry (bacciform) as in Onoclea. In some Polypodece the sori are furnished with scales of various forms; in Pleopeltis they consist of numerous orbicular peltate scales, imbricate over one another, and covering the sporangia of each sori; in Hymenolepis they are very thin and membranous ; in Lopholepis they consist of numerous slender scales, longer than the sporangia ; in Schellolepis they are irregularly laciniated; club like and other forms are also common to the genera Vittaria and Temtis. Such bodies are considered to be imperfect * That is, the margins turned or rolled inwards, so as to face and meet one another on the under side of the frond or segment. 20 ON GENERA AND SPECIES. sporangia, to which the term paraphysis is given. In Amphidesmium, Lophosoria, Trichopteris, and several species of Alsophila, numerous simple or articulate hairs are mixed with the sporangia, which impart to the sori a woolly or hirsute appearance. The above is a general view of the form and position of the different organisation of Ferns, and furnishes data for the definition of genera and their arrangement; but as in many instances of botanical definitions, there are always exceptions to normal forms and position of parts, even to the sori being produced on the under side of the frond, as in Polystichwin anomalum, a native of Ceylon ; in some of its fronds the sori are as abundantly produced on the upper side as on the under, and bear the characteristic peltate indusium, The same also occurs in Trichiocarpe Moorei, a native of New Caledonia. Such seems to be the normal state of these two species even in native specimens, as also in plants cultivated in Kew. 3.—ON THE GENERA OF EARLY AUTHORS AND THEIR SYSTEMS OF CLASSIFICATION, Havine described the principal organs of Ferns, by means of which genera are defined and classified, I now proceed to give a brief review of the progress of Pteridology from the time of Linneus to the present, noting the works of the principal systematic authors who have from time to time collated and characterised the genera and species of Ferns at their respective periods. It appears that only about one hundred and ninety species were known to Linneus. These he classified under eleven genera—viz. : Osmunda, Onoclea, Acrostichum, ON GENERA AND SPECIES. 21 Hemionitis, Polypodium, Asplentwm, Pteris, Blechnum, Lon- chites, and Trichomanes. The data upon which they were founded were derived, as in flowering plants, from the different modifications and position observed in their so- called organs of fructification. The chief difference con- sisted in the fructification—spore cases or sporangia—being produced either in defined round masses, or in lines, or densely crowded together on the under side of the frond, or in simple spikes or panicles formed by the whole or a portion only of the fertile frond being contracted. Thus, according to Linneus’ characters, all Ferns having their fructification in simple or racemose spikes belong to Osmunda and Onoclea ; in crowded masses to Acrostichwm ; in round masses or spots to Polypodiuin; in lines across one another to Hemionitis; in lines diverging from the mid-rib to Aspleniwm; in a line on each side of the mid-rib to Blechnwm ; in a line or dots on the margin to Pteris and Adiantum; in a sinus on the margin to Lonchites; and inserted on the margin to Trichomanes. These Linnean genera are each composed of a greater or lesser number of species, each of which is presumed to agree in the form and position of the fructification with the respective generic characters as above stated. But the more careful examination by later observers of the same generic forms has revealed additional variations, which will come under notice as we proceed. It is worthy of remark, however, that the whole of the species of Ferns now known are capable of being referred to Linnean genera; but if such were done, species possessing very different appearances as regards habit and mode of growth would be brought together, and the genera would be of unwieldly dimensions. The first author of importance after Linnzus was Dr. 22 ON GENERA AND SPECIES. J. E. Smith,* who, after the death of the younger Linnzus, became by purchase the owner of the Linnean Herbarium. He published many botanical memoirs, and one on Ferns in 1793, in the fifth volume of the “Memoirs of the Academia of Turin,” in which he characterised thirteen new genera—viz.: Darea, Scolopendrium, Woudwardia, Lindscea (Dryander, MS.), Vittaria, Davallia, Dicksonia (L’Heritier, 1788), Cyathea, Hymenophyllum, Schizca, Gleichenia, Marattia, and Dancea. These, with the Linnean genera, raised the number to twenty-four, which he divides under two sections—viz., Annulate and Hrannulate. The technical difference between these consists in the spore cases being furnished with a ring, or destitute of a ring. This mode of arrangement has since been adopted by all systematic writers on Ferns, as defining two distinct orders which will come under special notice hereafter. The next general systematic work was the “ Synopsis Filicum” of Dr. Swartz, published in 1806. In that ex- cellent work nearly 700 species are described, and a great many doubtful species enumerated. The de- scribed species are classified under thirty-six genera, being an increase of twenty-five since those of Linneus. Eleven of them are founded on species removed from Linnean genera, their separation being due to a special organ which I have already described under the name of indusium, by some called involucrum. Although this organ was noticed by Adanson and Necker, yet it was not employed by Linneus in characterising his genera. It was first specially made use of by Sir J. E. Smith, in characterising his new genera, and also about the same time by the German botanist Roth, and was adopted by Swartz in the above- named work, in which he arranges twenty-five of the * Afterwards Sir James Edward Smith. ON GENERA AND SPECIES. 23 genera under two divisions. The first, which he terms “ Nudis,” consists of six genera, characterised by the sori being naked, that is destitute of indusium, by some termed gymmnosorus ; this division contains 167 species. The other genera being furnished with an indusium he terms “ In- dusiates,’ which comprehends 467 species. These two divisions embrace all Ferns which have their sporangia furnished with a vertical ring; as also the genera Hymen- ophyllum and Trichomanes, in which the ring is horizontal. The next systematic arrangement of Ferns appeared in 1810, in the fifth volume of the “Species Plantarum” of Willdenow, wherein 1,010 species of circinate Ferns are described, comprehended under forty-one genera, being an addition of five new genera since Swartz. Three of these new genera are founded on species retained by Swartz in Linnean genera, the other two (Polybotrya and Pleopeltis), are each founded on a single species, for which the authority of Humboldt and Bonpland is given. As the characters upon which Willdenow founded his genera do not bring forward any striking new feature of structure, beyond certain modi- fications in the nature of the indusia and contraction of the fertile frond, I do not deem it necessary to speak further of them in this place. The first pictorial work at that period was the ‘“‘Kryptogamische Gewichse” (plants), by Pro- fessor Schkuhr, of Wurtemburg, published in 1809; itis a quarto volume containing 196 finely executed coloured plates representing 263 species of circinate Ferns, and accompanied by 212 pages of descriptive text. The im- portance of this work is manifested by the fact that all succeeding pteridologists quote the figures as evidence in the identification of species. The above were followed by special works and miscel- laneous writings in journals on Ferns, the principal writers 24 ON GENERA AND SPECIES. being Bernhardi, Robert Brown, Bory, Desvaux, Kaulfuss, Gaudichand, Raddi, Martius, Presl, Don, and others*. In these works a number of new genera are characterised, but it is not necessary to enter into further details regarding them here, more especially as the principal of them are given hereafter, either as accepted genera or as synonyms. These authors were followed by C. Sprengel, who, in the fourth volume of his “Systema Vegetabilium,” published in 1827, brought together and arranged in systematic order the whole of the species of Ferns known to him. Although this work cannot be looked upon otherwise than as a hasty compilation, and is of little value as a work of reference, yet it has the merit of bringing into a brief compass a descriptive enumeration of all the Ferns known at that period, collated from previously published works; the total number of his circinate Ferns being 1,309, which he arranges under fifty genera, This shows that a con- siderable increase of species had taken place since the time of Willdenow, and a glance at the author’s names above mentioned proves that the study of Ferns had not been neglected. The discoveries of Blume and Reinwardt in Java, led to the publication in 1830 of Blume’s enumeration of the Ferns of that and other of the Malayan Islands, in which 460 species are described. Of these 300 were new, the whole being arranged under fifty genera, six of which were previously unknown. This was followed by the splendid folio work, Blume’s “Flora of Java,” which contains finely-coloured plates of Ferns; and in the letter press he gives many new ideas respecting genera. At this period Link and Kunze published several treatises on Ferns, the first an enumera- * See list of authors and books. ON GENERA AND SPECIES. 25 tion and description of Ferns in the Berlin Gardens, entitled ‘‘ Hortus Berlinensis ;” the second a work entitled “ Analects Botanice;” as also a continuation of Schkuhr’s “Filices.” He also published many articles on Ferns in the “ Linnwa,” in which many new species are described. In 1889 Professor Agardh, a Swedish botanist, published a memoir on the genus Pteris, of which ninety-four species are described. 4,—ON THE GENERA OF MODERN AUTHORS, AND REVIEW OF THEIR SYSTEMS OF CLASSIFICATION. It would occupy too much space to enter into a review of the many works and the different views entertained by many of the above authors regarding the characters of genera, especially as all matters of importance relating to genera will be found noticed in their respective places. The chief characters adopted by many of the preceding authors for defining genera consist in differences in the form and arrangement of the sori, and in the different forms of the indusium when present, and even some on the texture and pubescens, and in the contraction of the fertile fronds. By the conjoint labour of authors the number of genera have been considerably increased since the time of Swartz, making the total number up to the time of Sprengel sixty-five, several of which may be viewed as natural genera. Eight are founded each on a single species, six contain two to four each; while the old Linnean genera Polypodiwm, Pieris, and Asplenium, and the Aspidiwm of Swartz collectively contain 648 species, being nearly one half of the whole number enumerated by Sprengel. Although this great number of species agree in the techni- cal characters of their respective genera, as founded on 26 ON GENERA AND SPECIES. their ‘fructification, yet a great many differ so widely in general habit and mode of growth, that any systematic arrangement, founded on genera so constituted, must be received as purely artificial, and only admissible on the principle of its brging together under a few brief cha- racters a number of species possessing such characters in common. This being the case, modern Pteridologists have found it necessary to seek for other characters in order to classify the various groups in accordance with their natural relationship to one another, as exhibited by their general appearance in habit and mode of growth. This became the more necessary on account of the great increase in the number of species from time to time brought into notice by botanical collectors as previously explained. Robert Brown was the first to point out and use an additional character for defining genera. In 1810, in characterising the genera Cyathea, Hemitelia, and Alsophila, he took into account the difference of the position of the sori on the veins, and this he again brought into special notice in the character of his genera Matonta and Hypoderris, first de- scribed in Wallich’s “Plante Asiaticee Rariores.” In Horsfield’s “ Plantee Javanice Rariores,’ published in 1838, he enters more fully upon the importance of employ- ing venation as affording auxiliary generic data. He there sketches out a method for dividing the large genus Polypo- dium into groups of species naturally allied in general habit, on characters derived from the structure of the venation and position of the sori on the veins. About this period several botanists on the Continent directed their attention to the study of the venation of Ferns. The first worthy of notice is Ferdinand Schott, Director of the Imperial Gardens at Vienna, who, in 1834, commenced a work entitled ‘Genera Filicum.” This ON GENERA AND SPECIES. 27 work consists of quarto plates, each plate illustrating a genus by figures, showing a portion of the fronds natural size and also magnified, in which the anatomical structure (venation) and the form and position of the sori and other parts of the fructification are distinctly shown, each plate being accompanied with descriptive letterpress. At the time of the author’s death, only forty plates of this prac- tical and useful work were published. With the works of Schott and Brown commenced a new era in the history of Pteridology. It is, however, to Pro- fessor Presl, of Prague, that the credit is due of being the first to publish a general systematic arrangement of genera founded on venation, which appeared in 1836 in his celebrated work entitled “Tentamen Pteridographes.” In that work about 1,500 species of annulate Ferns are enumerated (exclusive of the tribes Hyienophyllece and Osmundece), which he classifies under 115 genera arranged under two sub-orders, technically distinguished by the dif- ference in the direction of the rig that surrounds the sporangium. The first he terms Helicogyratce, which is characterised by the ring being horizontal or more or less oblique with its point of attachment, it contains two tribes, viz. :—Gleicheniacece and Cyatheacew, the first of which has five genera and twenty-three species; the second eight genera and twenty-four species. The second sub-order is termed Cathetogyrate, and contains all Ferns in which the ring of the sporangia is vertical. The following is an abridgment of Presl’s arrange- ment :— Orprr I.—FILICES. Presi. Sub-Order 1.—Huttcocyrarz, Bernh. Ring horizontal or oblique. 28 ON GENERA AND SPECIES. Tribe 1,.—GLEICHENIACEAE. (Ez. Platyloma, Br., and Gleichenia, Sm.) Tribe TI.—CYaTHEACEAR. (Ex. Cyathea, Sin., Hemitelia, Br., Alsophila, Br.) Sub-Order I..—Catuerocyratan. Bernh. Ring vertical. Cohors I.—Hymenoruores, Pr. (indusiate genera.) Tribe I.—ParaneMacran, Pr. (He. Paranema, Don., Thyrsopteris, Kze, Cibotium, Kaulf.) Tribe I1.—Aspriptacean, Br. Sect. 1. Nephrodiariae, Br. (Zz. Lastrea, Pr., Oleandra, Cav., Nephrolepis, Schott.) 2. Aspidiariae, Br. (Hx, Polystichum, Schoti, Aspidium, Schott, Didymochlaena, Desv.) Tribe IT1.—AspLENIAcEAB, Pr. Sect. 1. Cystopterideae, Pr. (Hx. Cystopteris, Bernh., Leucos- tegia, Pr., Onoclea, Linn.) 2. Blechnaceae, Pr. (Hv. Athyrium, Roth., Wood- wardia, Sm., Blechnum, Linn.) » 8. Aspleniariae, Pr. (Hx. Asplenium, Linn., Hemi- dictyum, Pr.) » 4. Diplazieae, Pr. (Ex. Diplazium, Sm., Oxygonium, Pr.) » 2» Scolopendrieae, Pr. (Hx. Scolopendrium, Sim., Anti- gramma, Pr.) Tribe TV.—Davatitacean, Gawd. Sect. 1. Davallieae, Pr. (Hz, Microlepa, Pr., Saccoloma, Kaulf., Davallia, Sm.) 2. Lindsaeaceae, Pr. (Hw. Lindsea, Dryand., Schizo- loma, Gawd.) Tribe V.—DIckKSONIACEAR, Pr. (Ez. Balantium, Kaulf, Dicksonia, Pr., Culcita, Pr.) ON GENERA AND SPECIES. 29 Tribe VI.—AvIANntTIACEAR, Pr. Sect. 1. Adiantariae, Pr. (Zz. Haplopteris, Pr., Pteris, L., Adiantum, Linz.) . Lonchitideae, Pr. (Hx. Hypolepis, Bernk., Lon- chitis, L.) Cohors II. Gyunogorem, Pr. (non indusiate). Tribe VII.—Virrarwcear, Pr. (Bz. Vittaria, Sm., Prosaptia, Pr.) to Tribe VIII.—Potyropracrag, Pr. Sect. 1. Struthiopterideae, Pr, (Hx. Struthiopteris, Willd.) » 2. Polypodiew, Pr. (Zz. Polypodium, Pr., Goniopteris, Pr., Phymatodes, Pr.) », 8. Lecanopterideae, Pr. (Lecanopteris, Bl., Calymmo- don, Pr.) Tribe [X.—GRaMMITACEAR, Pr. Sect. 1. Grammitideae. (Zz. Monogramma, Schivr., Menis- cium, Schrad., Selliguea, Bory.) ,, 2. Hemionitideae, Pr. (Hx. Heminoites, Linn., Gymno- gramma, Desv.) Tribe X.—TmnitiwBae, Pr. (Ex. Pleurogramma, Pr., Notholwna, Dr., Tonitis, Siw.) Tribe XI.—AcrosticHacuaL, Pr. (Bx. Polybotrya, H. B. K., Platycerium, Desv., Acrosti- chum, Liu.) In 1845, Presl published a supplement to the above work, entitled, “Supplementum Tentamenum Pteridogra- phiw.” It consists of 119 quarto pages (without plates) ; it embraces the tribes Muarattiaceae, Osmundaceae, Schizacaceae, and Lygodiaceae. Of the first he makes two sections— Marattiaceae and Danaeaceae—of which he describes forty- nine species, arranged under ten genera, five of which are 30 ON GENERA AND SPECIES. now being constituted of species separated from the well- known genera Marattia, Angiopteris, Dancea, the technical characters of which are, however, so slight that Ido not consider them worthy of adoption. Of Osmundaceae he describes twenty-four species, two of which he characterises as a new genus under the name of Leptopteris, but which I consider not worthy of being separated from TZodea of Willdenow. Of Schizaeaceae he enumerates sixty-five species, arranged under eight genera, three of which are new—Lophidium, consisting of five species separated from Schizcea, and the two others upon a single species each, founded upon doubt- ful and even mistaken data, one of them (Spathopteris) having no other authority than a figure of Plumier’s “ Filices,” which, in my “ Genera Filicum,” published in 1841, I show to consist of a barren and fertile frond of two distinct Ferns. With regard to Lygodiaceae he describes forty-one species, thirty-nine of which belong to the genus Lygodium of Swartz, and two to Hydroglosswum of Willdenow. The number of species of the three latter genera are also in excess, Anemia alone having thirty-eight. In 1848, Dr. Pres] published another work, entitled “ Hymenophyllacew,” in which he describes 185 species, which had hitherto been comprehended under the well- known genera Trichomanes and Hymenophyllum, but of which Presl makes no less than nineteen genera; and in his “Epimelia Botanica” he characterises three more, making twenty-two in all, about one-third of which are founded upon a single species each. The data upon which these genera are founded is derived chiefly from the size and form of the cellular parenchyma, and the slight modi- fications observable in the form of the indusium and ON GENERA AND SPUECIES. 31 sporangiferous receptacle, differences which, with two or three exceptions, appear to me to be more useful as specific than as generic characters ; indeed, in some cases they are so trifling that I have failed to find them. Under these circumstances I cannot see that any practical advantage will be gained by dividing Hymenophyllacee into so many genera as Pres] has done. Hymenophyllacece also form the subject of a special mono- graph, published in 1858 by Dr. Van den Bosch, of Goes, Holland. It consists of seventy-nine pages octavo, and is entitled “Synopsis Hymenophyllacearum, Monographia hujus ordinis Prodromus.” It contains simply an enu- meration of names of genera and species with references to authors. He enumerates no less than 305 species, being 120 above that of Presl, and 156 above that enumerated by Sir William Hooker in the “Synopsis Filicum,” here- after to be noticed. The whole are arranged under nine genera, of which Hymenophyllum and Trichomanes alone contain 254 species. The remaining fifty-one are divided amongst seven genera, two of which contain a single species each. He does not characterise any new genera, and only adopts five of Presl’s. But his number of species is so much in excess of that of other authors that, if he really had specimens of the whole in his herbarium he must have described his species on the different forms of fronds and other variable structure, which are always to be found in an extensive series of herbarium specimens, many such when carefully collated represent a single species only. In 1849, Pres] published another work, entitled “ Epimelia ” or additions to and emendations of his “ Tenta- Botanica, men Pteridographew.” In this work he characterised no less than thirty-five new genera, and described a consider- e 32 ON GENERA AND SPECIES. able number of new species, many of which appear to be founded on very imperfect, and not well-authenticated materials. A few of his new genera are admissible and will be noticed in their respective places; the greater number, however, are untenable ; for instance, he describes about fifty species of the genus Blechnwm, of authors, which he arranges under five genera, namely, Blechnum, Parablech- num, Distavia, Mesothema, and Blechnopsis; the differential character of these genera is, however, so slight that I do not consider them worthy of adoption (see Blechnum). Another instance of creating genera on what may be called fancied differences, is the genus Niphobolus of Kaul- fuss, of which Pres] enumerates thirty-nine species arranged under eight genera. On examining numerous sets of herb- arium specimens said to be species of this genus, I find the intermediate forms are so numerous that it is impossible to collate them, so as to arrive at any satisfactory conclusion as to what number of them are distinct species, for with all the evidence that has come before me, I have not been able to define more than about a third of the species recorded by Presl, all of which I continue to retain under Niphobolus. Many other of Presl’s genera have as little title to rank as such as those of Blechnum and Niphobolus, and I deem it best not even to notice them as synonyms, as it would only add to the already overburdened Fern nomenclature. In the preface to this book it is shown that it is now more than fifty years since the writer commenced to study the collection of Ferns in the Royal Botanic Garden at Kew, then amounting to about forty exotic species; and, having had the opportunity of profitmg by the verbal observations of Robert Brown, who often directed my attention to the mode of the arrangement of the veins in different species of the genus Polypodiwm, as forming ON GENERA AND SPECIES, 33 auxilliary characters in the grouping of species, which with his published observations above noticed, I, in time, consi- dered myself competent to publish my views. This I did in a treatise entitled “On Arrangement and Definitions of the Genera of Ferns,” which was read before the Linnean Society in 1840, and afterwards published in Hooker’s “ Journal of Botany,” in the volumes for 1841 and 1842, occupying 131 pages, and accompanied by a plate showing the principal forms of venation. In this I characterised 143 genera, accompanied by examples of a few of the most typical species of each genus, and with observations on the natural affinities of the genera. Just as my treatise was completed, but before its publication, I received Presl’s great work “Tentamen Pteridographes,” and it was gratifying to me to find that my views in general coincided with those of that celebrated Pteridologist. I, however, differed from him in some important particulars, chiefly with regard to the relationship of the genera to one another, consequent on the data on which Pres] founded his tribes, placing naturally allied species far apart. My endeavour was to associate the species possessing the same general features in respect to their modes of growth (habit), so as to form natural sequences of genera. This I have endeavoured to complete by drawing up a general arrangement, which forms the second part of this volume. The following is an abstract of my arrangement in 1841 :— Division L-—POLY PODIACE, f. Br. Sporangia furnished with a vertical ring. Tribe 1.—Potyropisx, J. Sm. Sori punctiform, oblong or linear, naked or included under the universal indusium. Scct. 1. Orthophlebiee, J. Sm. Veins free. D 34 ON GENERA AND SPECIES. Ez. Grammitis, Sw., Polypodium, Linn., Struthiopteris, W. Sect. 2. Symplophlebiew, J. Sm. Veins simple or com- pound anastomosing. Ez. Goniopteris, Pr., Meniscium, Schred., Drynaria, Bory. Tribe T1.—AcROSTICHIER. Fertile fronds contracted. Sori amorphous, naked. Sect. 1, Orthophlebiew. Veins free. Ex, Elaphoglossum, Schott, Stenochlena, J. Sm., Polybo- trya, Humb. Sect. 2. Symplophlebiex. Veins anastomosing in various ways. Ex, Aconiopteris, Pr., Acrostichum, Linn., Platycerium, Desv. Tribe TI1.—Preripex, J. Sm. Sori marginal, punctiform or linear, or intramarginal transverse, furnished with a latter exterior attached indu- sium. Sect. 1. Chilosorex, J. Sm. Sori seated on the margin. Ex, Cheilanthes, Sw., Pteris, Linn., Adiantum, Linn. Sect, 2. Metasorm, J. Sm. Sori oblong or linear. Ee. Blechnum, Linn., Lomaria, Willd., Woodwardia, Sm. Tribe 1V.—Aspiuniee, J. Sm. Sori oblique, linear, single or binate, furnished with a lateral indusium. Sect. 1. Orthophlebiew. Veins free. Ex, Asplenium, Linn., Diplazium, Sw., Scolopendrium, Sm. Sect. 2. Symplophlebiew. Veins variously anastomosing. Hz. Neottopteris, J. Sin., Callipteris, Bory., Allantodia, Rk. Br. Tribe V.—Aspivienr, J. Sm. Sori round, furnished with an orbicular, ov lateral iute- riorly attached indusium, Q ON GENERA AND SPECIES. 35 Sect. 1. Symplophlebiew. Veins anastomising in various ways. ; Ez, Aspidium, Sw., Pleocnemia, Pr., Nephrodium, Schott. Sect. 2. Orthophlebiess. Veins free. Ex. Lastrea, Bory., Polystichum, Roth., Didymochlena, Desv. Tribe VI.—Dicxsoniez, J. Sim. Sori marginal, round or linear, furnished with a special and accessory indusium, which united form a cyst or groove including the sporangia, and opening exteriorly. Sect. 1. Lindseew, J. Sm. Sori linear, rarely round. Ex. Lindswa, Dry., Dictyoxiphium, Hook, Odontoloma, J. Sin, Sect, 2. Davalliew, J. Sm. Sori round or vertically oblong. Indusium cystiform, cuculate or tubular. Ex, Microlepia, Pr., Davallia, Sm., Loxoma, R. Br. Sect. 3. Trichomanem, J. Sm. Sporangia seated in a free columnar receptacle included in or inserted along the cyst. Ex. Trichomanes, Linn., Hymenostachys, Bory., Hymeno- phylum, Sin. Sect. 4. Dicksonie. Sporangia included within a concave, bilabiate, or calyciform, reflexed cyst. Ex. Sitolobium, Desv., Dicksonia, L’Herit., Cibotium, Kaulf. Tribe VIL._—Cyatuea, J. Sm. Sori round intramarginal, receptacle elevated. Indusium calyciform, or lateral, trichiform or absent. Ex, Cyathea, Sm., Hemitelia, R. Br., Trichopteris, Pres. Division II.—GLEICHENIACEA, R&. Br. Sporangia furnished with a transverse ring. Sori round intramarginal, Er. Platyzoma, R. Br. Gleichenia, S»., Mertensia, Willd. 36 ON GENERA AND SPECIES. Division W1.—SCHIZAACKA, Mart. Apex of sporangia rayed, analagous to a transverse ring, generally produced in simple spikes, racemes, or panicles. La, Lygodium, Sw., Schizewa, Sw., Anemia, Sw. Division IV.—OSMUNDACEA, Mart. Sporangia globose, reticulated, destitute of stria, opening vertically in two valves, produced on contracted fronds. Hz. Osmunda, Linn., Todea, Willd. Division V.—MARATTIACEA, Kaulf. Sporangia opaque, horny, unilocular, or multilocular, destitute of a ring. Liz, Angiopteris, Hofm,, Marattia, Sm., Danea, Sin. More than thirty years have now elapsed since the above arrangement was drawn up and published, and during that time about 1,000 cultivated species have come under my notice, as well as many additional herbarium specimens new to me, the consequence of which has been to greatly alter my views in the definition of genera and _ their arrangement, which is explained further on. The next general arrangement of this family of plants to be noticed is that of Professor Fée, of Strasburg, who in 1852 published his “ Exposition des genres de la famille des Polypodiacées.” This great work consists of 887 pages of quarto letter press, in which is given the cha- racters of 188 genera, with observations and examples of species of each genus, which are illustrated by figures occupying thirty-two plates, the whole evincing great talent and research. In this work 2,140 species are enu- merated of annulate Ferns; he follows Presl’s method in dividing them into two divisions founded on the direction of the ring, but restricts Helicogyratee to the tribe Cyathew, thus excluding Gleicheniacee, which is not taken ON GENERA AND SPECIES, 37 up in the work. He classifies the genera under twenty-six heads, which bear the name of the leading genus of each, and adopts the different modes of venation and character of the sori as part of his generic character, but he introduces new elements in characterising his genera, such as giving the form and structure of the sporangia, and the number of articulations of the ring, also the different form of the spores ; these form an important feature in his generic as well as in his specific characters; but with regard to their value, I candidly confess that in all my observations and experiences I have failed to satisfy myself that either the number of articulations of the ring, or the form of the spores, can in any way be practically serviceable for defining either genera or species. The little value of such characters become obvious on comparing his descriptions of individual species with their respective generic cha- racters, for example, in the generic character of Aspidiwm the number of articulations of the ring is stated to be 18-20, that is, all his species of Aspidium should not have less or more than the above number; but in ten of his new species of Aspidium, he describes the number of articulations as much below the number assigned in the character of the genus; hence, these species according to his own descriptions, should not be placed in Aspidiwm. As far as regards the characters derived from the spores, I am sensible that their forms are various and in some cases peculiar ; but the spores of Ferns, like the seeds of other plants, will, if gathered before they are mature, shrivel and assume forms that will, in many instances, appear to the microscopic observer to be normal, and figures of such appearances are consequently drawn and published. In my opinion much examination is still re- quired, before either the structure of the sporangium or 38 ON GENERA AND SPECIES. microscopic spores can be made available as auxiliary cha- racters, and this will require the labour of years ; and even when complete, will be of but little service either to the student or general observer; for I am satisfied that, with our present knowledge of the more obvious outward features that distinguish Ferns from one another, these in- teresting plants can be readily arranged in groups forming natural genera and alliances, without requiring much aid from the microscopist. Besides the additional characters derived from the struc- ture of the sporangia, both Presl and Fée attach much im- portance to the difference in the number and forms of the bundles or plates of vascular fibre in the stipes. My observations upon that point tend to show, that, although the bundles are generally uniform in simple fronded spe- cies, yet in many others the number and form appear to vary according to the frond being more or less compound and varying in size, age, and vigour of growth of indi- vidual plants of the same species. In the stipes of a full- sized frond (ten feet in length) of Cyathea arborea, I found thirty-six vascular bundles in the stipes below the lowest pair of pinnae, and this number nearly corresponded with the number of pinne forming the frond ; in fronds of in- termediate sizes the bundles were fewer, apparently formed suited to the more or less compound development of the frond consequent on the size and age of the plant; for instance, in fronds of the same species, a foot in length with few pinne has only three broad vascular plates in the stipes. It is reasonable to suppose that the shape or outward form assumed by every species is based upon a vascular frame-work, developed according to some definite plan inherent to each individual species, but varying according to the circumstances that influence the growth ON GENERA AND SPECIUS. 39 and age of individual plants of the same species. Further investigations, however, are required before the vascular structure of the stipes can be made of service as a distin- guishing character for either genera or species, and this can only be satisfactorily obtained by a series of observa- tions of living plants. M. Fée gives a systematic arrangement of his method of classification, which occupies nearly five double-columned pages; but the many divisions, sub-divisions, figures, letters, and arterisk, renders it necessary to be very care- fully studied before it can be well understood. The fol- lowing is sufficient to show M. Fée’s mode of classifying genera, from which it will be seen that plants most opposite in natural habit are associated, consequent on characters derived from the form and position of the sori, and in being naked or indusiate and the different forms of the latter, and therefore cannot be considered otherwise than as an artificial arrangement of complicated construction, Abstract of Fée’s arrangement :-— OrpEr—POLY PODIACEl. I. CaTHETOGYRATER. Acrosticheae, Gen. 19. (Hz. Acrostichum, Platycerium, Leptochilus.) Lomarieae. Gen. 8. (Hz. Blechnum, Acropteris, Hymon- olepis.) Vittarieae. Gen. 10. (Hz. Pteropsis, Drymoglossum, Lemogramma.) Pleurogrammeae. “Gen. 5. (Hx. Monogramma, Adeno- phorus, Xiphopteris.) Pterideae. 1. Lindsayew. Gen. 5. (Hx. Isoloma, Schizoloma, Dictyoziphium). 40 ON GENERA AND SPECIES, bo . Adiantew. Gen. 4. (Hz. Adiantum, Casebeeria, Ochropteris. ) 3. Pterideae. Gen. 8. (Ez. Pellea, Onychium, Am- phiblestra.) 4, Lonchitideae. Gen. 1. (Lonchitis.) Hypolepidesw. Gen. 2. (Hypolepis, Adiantopsis.) Hucheilantheae. Gen. 6. (He. Myriopteris, Pleucosorus, Nothochlena ) : Jamesonieae, Gen. 1. (Jamesonia.) Hemionitides. Gen. 8. (Ex. Coniogramma, Hemionitis, Syngramma.) Antrophyeae. Gen. 8. (Antrophyum, Colysis, Selliguea.) Leptogrammesw. Gen. 6. (Ez. Pterogonium, Hecistop- teris, Gymnogramma. ) Asplenieae. Gen. 9. (Hz. Neottopteris, Hemidictyon, Woodwardia.) Scolopendriew. Gen. 3. (Scolopendrium, Antigramma, Camptosorus.) Diplazies. Gen. 5. (Hz. Diplazium, Callipteris, Didymo- chleena.) Menisciee. Gen. 2. (Meniscium, Dryomenes.) Struthiopterides. Gen. 3, (Struthiopteris, Onoclea, Cera- todactylis.) Polypodiew. Gen. 19. (Grammitis, Phegopteris, Dry- naria.) Cyclodiew. Gen. 8. (Hz. Polystichum, Cyclodium, Cryto- mium.) Aspidiew. Gen. 13. (Hz. Cystopteris, Nephrodium, Oleandra.) Nephrolepidee. Gen. 5. (Ha. Nephrolepis, Saccoloma, Humata.) Davalliew. Gen. 10. (Hx. Prosaptia, Microlepia, Odonto- loma.) ON GENERA AND SPECIES. Al Dicksoniew. Gen. 6. (Hx. Dicksonia, Hypoderris, Woodsia.) Balantiew. Gen. 4. (Hx. Balantium, Leptopleuria, Cysto- dium.) Cibotierw. Gen. 1. (Cibotium.) 2. HEDICOGYRATER, Cyathee. Gen. 7. (Hz. Alsophila, Hemitelia, Amphides- minum.) Thrysopterides. Gen.1. (Thrysopteris.) Besides his general work on Polypodiacee, M. Fée has also published several memoirs on special genera, accom- panied by figures, one forming a volume devoted to the tribe Acrostiches, in which he arranges the species under nineteen genera. Vittaria and its allies form another memoir ; of Vittaria alone he describes and figures no less than twenty-five species and five doubtful, but the exammation of numerous sets of specimens justifies me in reducing the number to less than a dozen, I have now given a general view of the systematic arrangements of Presl, my own, and Fee’s, by which it will be seen that the first important divisional character is derived from the different directions of the ring that sur- rounds the sporangium. My reason for again reverting to this subject is in order to show that in some Cyathee the so called obliquity of the ring is more apparent than real. In my arrangement of 1841 I have stated my views on this point. I have there explained that the apparent obliquity of the ring in Cyathec is due to mechanical pressure, the sporangia of each sorus being crowded and sessile seated round a columnar elevated receptacle, and each sporangium being inclined upwards, consequently the pressure against each other imparts to them an oblong flattened form especially in the lower part on their inner 42 ON GENERA AND SPECIES. side. This, according to my view, accounts for the ring appearing not to be truly vertical, but it is to be observed that the obliquity is not general, for in the figures of the 8porangia of eight Cyatheaceous genera in Hooker and Bauer’s ‘Genera Filicum,” the ring is shown to be verti- cal, I therefore follow Robert Brown in placing Cyathec in Polypodiacee. The next systematic work to be noticed is that of Professor Mettenius, of Leipzig, who, in 1858, commenced publishing a work entitled “ Uber einige Farngattingen ” (‘On Some Genera of Ferns”), of which five parts have appeared. In order to explain this author’s system of classification of species, I will give a brief outline of the manner in which he treats the genus Polypodiwm. He enumerates, and mostly describes in full, 258 species of this genus, including in it all the forms possessing puncti- form, oval, or linear, naked sori, thus restoring it to nearly the state in which it was left by Swartz, Willdenow, and Sprengel, and even including in it the genus Grammitis of those authors. His reason for placing so many species under one genus is on account of the numerous interme- diate or transition forms, which he says so pass into one another that he finds it difficult to define any group of species as a distinct genus in the manner that Presl and others had done. Notwithstanding this, however, he finds it quite possible to divide the genera into sections and sub- sections, of which he gives an elaborate synoptical table. He first divides the 268 species, according to their veins, being free or anastomose. Those with free veins are placed under four sections, and those with anastomose veins in thirteen sections. These sections and sub-sections (which are numerous) in most cases bear the names that designate the genera of Pres] and others, and consequently ON GENERA AND SPECIES, 43 contain the same species, or nearly so. His table of arrangement, showing the distribution of the species under the different sections, occupies thirteen pages of quarto of his Treatise, of which the following in an abridgment :— g§ 1 § a8 § 4 § 5 § 6 § 7 § 8 § 9 § 10 § 11 § 12 § 18 GENUS POLYPODIUM. . Cenopteris (Adenophorus, Hook. Xiphopteris, Kaulf.). . Ctenopteris, Eupteris, vel Europteris (Grammites, Sw, Blume; Ctenopteris, J. Sm. Polypodium, J. Sm.). . Cenopteris, Ctenopteris, rarissime Eupteris (Pleo- peltis in part, J. Sm.). . Marginaria ; folia conformia (Lepicystis, J. Sm.). . Marginaria-Goniophlebium (Goniophlebium, Blume, J. Sm.). . Cyrtophlebium (Cyrtophlebium, R. Br. Campylonewron, Presl). . Doodya; nacule exappendiculate (Dictymia, J. Sm.). . Doodya ; nacule appendiculate (Phymatodes in part, Presl). . Phlebodium (Pleopeltis in part, J. Sm.). . Marginaria; folia difformia (Lopholepis, J. Sm.). . Phlebodium (Anapeltis, J. Sm. Phyimatodes in part, Pres). . Anaxetum (Pleuridium in part, Presl, J. Sm.). . Anaxetum (Pleuridiwm in part, J. Sm., Selliguea, Blume, Loxogramma, Presl). . Drynaria (Plewridium in part, J. Sm. Colysis, Pres. Drynaria, Bory. Depteris, Reinw. Monosorum Linbr). . Pleocnemia (Niphopsis, J. Sm.). 3. Pleocnemia (Niphobolus, Kaulf., Spreng., J. Sm.). . Pleocnemia appendiculata, s., Drynaria (Niphobolus, Presl, J. Sm.). 44 ON GENERA AND SPECIES, I must state that after much study of this arrange- ment I have not been able sufficiently to understand it to make it practically useful. I do not clearly see what object the author gains by introducing sectional names not before used in the genus, more especially as these names (Ceenopteris, Lupteris, Neuropteris, Doodya, Pleocnemia, &c.) are apparently not given as special sectional names, but used comparatively—that is, all the species under Cvenop- teris, Hupteris, &c., are presumed to have venation similar to Ferns bearing those names. Therefore, in order to ascertain the venation of one Fern, it is first necessary to be acquainted with the venation of that with which it is put in comparison. As, for example, to know Dictymia, it is first necessary to be acquainted with the character of the venation of Doodia and Pleocnemia. Unfortunately too, the analogy in these two cases is far from obvious, the genera in both cases being in every respect of quite dif ferent habits. I therefore see no good reason for making such genera as Doodia and Plevcnemia types for arranging other genera by venation. With regard to his arrangement of Phegopteris and Aspidium, of which he enumerates in consecutive order 299 species—thus viewing them as parts of a single genus —of these sixty-eight belong to the first part, Phegopteris ; the remainder to the second part, Aspidiwm. I do not consider it necessary to enter into further details, I will only add that the sectional names are upon the same principle as those of Polypodiwm, and that, in my opinion, a simple generic name for groups of naturally allied species, would render their study much easier than the complicated comparative similitude with one another. Mettenius has also published a memoir on the genus, Asplenium and several other genera. r ON GENERA AND SPECIES. 45 T have now to notice an important but yet incomplete work, the “ Index Filicum,” a synopsis with characters of the genera, and an enumeration of the species of Ferns with synonyms, and references, &c., &c., by Mr. Thomas Moore, Ourator of the Garden of Apothecaries, Chelsea ; the first part of this work appeared in 1857. The author first gives a synopsis of the orders and tribes, followed by a systematic arrangement and characters of the genera, each genus illustrated by a woodcut showing the character of the venation and position of the sori, also an analytical table of the genera and their synonyms. With regard to the number of genera the author tells us that he has endeavoured “ to hold a middle course between the excessive sub-division and the equally inconvenient non-division of older genera,” and that the classification adopted is based on the plan on which the vascular struc- ture is developed, in conjunction with the nature of the fructification. He enumerates 186 genera*, which he classifies under five orders, namely, Pulypodiacee, Maratti- acece, Ophioglossacece, Lycopodiacee, and Marsileacece. The following abstract will be sufficient to convey an idea of Mr. Moore’s method of classification. FILICALES, SPORE-CASHES. ONE-CELLED. * Spore-cases furnished with a jointed ring, which is usually nearly complete, sometimes rudimentary. Orper I.—POLYPODIACEA. || Ring vertical, nearly complete. Tribe I.—PoLyPoDINnEZz. § 1. Acrostichee (Ex. Elaphoglossum, Schott. Acrostichum, Linn.) * This includes ten genera of the orders Lycopodiacee and Marsilia- cee, which are not taken up in the present work. 4.6 ON GENERA AND SPECIES. § 2. Platyceriex (Ex. Platycerium, Desv., Jenkinsia, Hook.) § 3. Lomaries (Ex. Lomaria, Willd., Blechnum, Lin.) § 4. Pleurogrammes (Ex. Monogramma, Schh., Hymenolepis, Klfs., Gymnopteris, Bernh. § 5. Tenitidee (Ex. Lomogramma, J. Sm., Drymoglossum, Presl, Twntopsis, J. Sm.) § 6. Vittaries (Ex. Vittaria, Sm.) § 7. Lindsees (Ex. Lindsee, Dryand., Dictioaiphium, Hook.) : § 8. Adiantes (Ex. Adiantum, Linn., Hewardia, J. Sm.) § 9. Cheilanthes (Ex. Hypolepis, Bernh., Casebeera, K1fs.) § 10. Pterides (Ex. Onychium, Klfs., Haplopteris, Presl., Amphiblestra, Presi.) § 11. Woodwardiew (Ex. Woodwardia, Sm., includes Doodia, R. Br.) § 12. Menisciese (Ex. Brana, J. Sm., Meniscium, Schreb., Dryomenis, Fée.) § 13. Aspleniee (Ex. Actiniopteris, Link., Asplenium, Presl, Ceterach, Willd.) § 14. Didymochlenee (Ex. Didymochlaena, Desv., Meso- chlena, R. Br.) § 15. Hemionitides (Ex. Polyteniwm, Desv., Anetium, Kifs., Dictyogramma, Fée.) § 16. Gymnogrammeae (Ex. Grammitis, Sw., Stegno- gramma, B1., Selliguea, Bory.) § 17. Platylomes (Ex. Plagiogyria, Kze., Llavea, Lagasca, Cryptogramma, R. Br.) § 18. Polpodies (Ex. Struthiopteris, Willd., Jamesonia. Hook. et Grev., Drynaria, Bory.) § 19. Aspidiea (Gen. 12. Onoclea, Lin., Polystichum, Roth., Nephrolepis, Schott.) § 20. Cystopterides (Gen. 3. Cystopteris, Bernh., Acro- phorus, Presl., Humato, Cav.) ON GENERA AND SPECIES. 47 mn to pane . Davallier (Gen. 4. Microlepia., Presl., Davallia, Sm., Prosaptia, Presl.) § 22. Dicksoniew (Gen. 7. Dicksonia, L’Herit., Dennstediea, Bernh., Cionidiwm, Moore.) § 23, Paranemere (Gen. 5. Diacalpe, Bl., Woodsia, R. Br., Hypoderris, R. Br. |||| Ring more or less obliquely vertical, nearly complete, narrow. Tribe T].—CYaTHEINEs. § 1. Thyrsopteredess (Gen. 1, Thyrsopieris, Kge, T. elegans). § 2. Cyathess (Gen. 3, Cyathea, Sm., Hemitelia, R. Br., Amphicosma, Gard.) § 3. Alsophiles (Gen. 2, Alsophyle, R, Br., Amphides- mium, Schott). Tribe III.—Matonrinex. (Gen. 1, Matonia, R. Br., M. pectinata). Tribe TV .—GLEICHENINER, (Gen. 2. Platyzoma, R. Br., Gleichenia, Sm.) Tribe 5.—TRICHOMANINES. (Gen. 5. Lowoma, R. Br., Hymenostachys, Bory., Hymenophyllum, Sm.) Tribe V1.—ScHIZ#INER, § 1. Lygodiew (Gen. 2, Lygodium, Sw., Hydroglossum, Willd.) § 2. Schizwew (Gen. 5, Anemia, Sw., Mohia, Sw., Tro- chopteris, Gard.) Tribe 7.—CERATOPTERIDINE®. (Gen. 1. Ccratopterts, Brongn., C. thalictroides). Tribe 8.—-OsMUNDINE®, (Gen, 2. Osmunda, Lin., Todea, Willd.) 48 ON GENERA AND SPECIES. ** Spore cases without a jointed ring. (a). Sori dorsal. Orper.—MARATTIACEA. Tribe 1.—MaRatTTingEZz, § 1. Angiopterides (Genus 1, Angiopteris, Hoffm.) § 2. Maratties (Gen. 3, Marattia, Sm., Gymnotheca, Presl, Lupodiwm, J. Sm.) Tribe 2.—KAavLFUSSINER, (Gen. 1, Kaulfussia, Bl., K. esculifolia), Tribe 3.—DanzInen. (Gen. 1, Dancea, Sm.) (b). Fructification marginal, on rachiform fronds or branches (venation straight). Orprr.—OPHIOGLOSSACEA. (Gen. 3. Botrychium, Sw., Helminthostachys, Klfs., Ophioglossum, Lin.) Lycopodales.—Acrogenous plants with axillary radical or one-four-or many-celled spore cases. * Spore cases one-three celled, in the axils of the stem- leaves or bracts. Orper.—LYCOPODIACEA. § 1. Phylloglossee (Gen. 1, Phylloglossum, Kze). § 2. Lycopodiee (Gen. 4, Tmesipteris, Bernh., Psilotum, Sw., Selaginella, P. B.) ** Spore-cases (conceptacles) one-four-or many-celled, radical or petiolar. Orper.—MARSILEACEZL, § 1. Isoetese (Gen. 1, Iscetes, Lin.) § 2. Salvinieae (Gen. 2, Salvinia, Micheli, Azolla, Lam.) ON GENERA AND SPECIES. 49 § 3. Pilulariew (Gen. 1, Pilulariece, Lin.) § 4. Marsilese (Gen. 1, Marsilea, Lin.) The above classification shows that Mr. Moore has be- stowed great ingenuity upon its formation, but I cannot view it otherwise than as an artificial arrangement founded upon certain technical data that normally characterise all true Ferns, whatever size, habit, or mode of growth they assume. We need not therefore be surprised at finding, associated under the same sections, genera of extreme un- natural habit and mode of growth as also differing essen- tially in their vascular structure, such as Jenkinsia and Platycerium ; Monogramma, Hymenolepis, and Gymnopteris ; Lomogramma, Drymoglossum, and Teeniopsis; Brainea, Mi- niscium, and Dryomenes; Didymochlena and Mesochlena; Calymmodon, Stegnogranuma, Gymnogramma, Struthiopteris, Jamesonia, and Drymaria; Onoclea, Polystichum, and Ne- phrolepis ; Cystopteris and Humata; and Woodsia and Hy- poderris : the placing of these in their respective sections being entirely guided by the character of the fructification, without regard to natural habit. Any system of classifi- cation of Ferns based upon these principles may be viewed as analogous to the Linnean classification of pheno- gamous plants. I have now to notice the next part of this work, which consists of an alphabetical arrangement of the whole nomenclature of Ferns, collated from the publica- tions accessible to the author, “and blended with such personal information as the author has been able to bring to bear upon the subject.” In order to appreciate the extent of this work, it is only necessary to state that up to Sep- tember, 1862,* 896 pages were published, and which * Up to the present time, 1874, no addition to the above has been published. E 50) ON GENERA AND SPECIES. extended only to Goniophlebium, being the seventy-third genus in alphabetical order, up to which are enumerated 1,738 species with their synonyms ; from this data it may be calculated that the number of species, when the work is complete, will amount to above 4,000. As an instance of the labour and investigation required, I will select the genus Acrostichum. Under this original generic name of Linneus, no less than 461 specific names are enumerated in alphabetical rotation, with references to their authorities, occupying twelve pages: with the exception of nine names, the whole are in italics, signifying that they are synonyms, the nine being in roman type, indicating that they are the only true species of Acrostichum as now adopted. When complete, this work will be an invaluable treasure to Pteridologists. I have now to notice the works of Sir William Hooker , who in 1820 became Professor of Botany in the University of Glasgow, and in 1841, Director of the Royal Gardens at Kew, but before the first date he had gained reputation as acryptogamic botanist. His first great work on Ferns was in conjunction with Dr. Greville, of Edinburgh, also a famous cryptogamic botanist, entitled “ Icones Filicum,” a work of two folio-sized volumes, containing 230 finely executed plates of new and rare Ferns, each plate con- taining magnified portions of the fructification exhibiting the characters of the genera as then defined, with accom- panying descriptive letter-press. In this work two new genera are characterised, viz., Deparia and Jamesonia, each founded upon a single species. His next work of importance is Hooker and Bauer’s “Genera Filicum,” which commenced with a set of draw- ings begun in 1833 by the late eminent botanical draughts. man Francis Bauer, from specimens from time to time ON GENERA AND SPHCIES. 51 furnished by me, some from living plants from the garden, aud others from herbarium native specimens. The draw- ings being highly appreciated by Sir W. Hooker, Mr. Bauer freely placed them in his hands for publication, and in 1842 a splendid volume consisting of 120 plates was completed,* each plate containing natural and highly mag- nified portions of fronds, exhibiting in different states the fructification and venation, accompanied with corres- ponding letter-press, giving the characters of each genus : there are on the whole 135 genera, many being derived from Presl’s “ Pteridiae Pteridographia,”’ and twenty being named and described by me as new. The learned editor, however, says he reserves for him- self the power of rejecting such of them as on further examination he may consider himself justified in doing. With regard to that point he also says, “so completely do the ideas of Dr. Presl accord with my own in regard to the limits of many gencra that I should do him injustice were I not in many cases to quote his characters verbatim, and indeed the more attentively I study his book and com- pare the descriptions with the plants themselves, the more satisfied Iam that he has produced a work that will not easily be surpassed for accuracy of, and clear and per- spicuous, arrangement.” In 1854 appeared the tenth volume of his work entitled “Tcones Plantarum,” containing 100 figures of new and rare Ferns, which was soon afterwards followed by another volume called a “Century of Ferns.” This was followed by “ Filices Exotice,” consisting of 100 plates, which gave place to another volume entitled “Garden Ferns,” contain- he drawings for the latter part of this work were executed by the equally eminent draughtsman Mr. Walter Fitch ; this was consequent on the health of Mr. Bauer having failed. 52 ON GENERA AND SPECIES. ing sixty-four plates, the drawings of the latter as well as part of the preceding being derived from living plants of the Kew collection. The practical use of these works is in assisting to determine species, and which, either under the same name or as synonyms, will be found in his last and great work the “Species Filicum,” the most important systematic work on Ferns yet published, being a description of all known Ferns, particularly of such as exist in the author’s herbarium (previously noticed), and such as are with sufficient accuracy described and figured in the works of other authors. It consists of five volumes, the first of which appeared in 1846, and the fifth in 1864, its prepara- tion having occupied his leisure time for a period of not less than twenty years. It treats of the orders Gileichent- acece and Polypodiacece only, of which 2,401 species are described, with their synonyms, which amount to about 4,300, as also their native countries and names of collectors, illustrated with 300 plates representing 522 species. He arranges the whole under sixty-three genera, fourteen of which contain only one species each, thus present- ing a great contrast to the mass of species which he continues to retain under the characters assigned by Lin- naus and Swartz, to such genera as Polypodium, Aspidium, Pteris, and Acrostichum, thus avowing his preference for large genera by strictly adopting the characters derived solely from the fructification. With regard to which he says, “Tnereased study has, he must confess, strengthened his conviction that those botanists who have shown them- selves peculiarly addicted to multiplying genera have not always taken nature as their guide, nor succeeded in eliciting a simple nor tangible arrancement, yet have their close and accurate investigations thrown a new light on the study of Ferns, a light which cannot fail to aid the re- ON GENERA AND SPECIES. 53 searches of future writers, and which ought therefore to be gratefully acknowledged; a middle course has been pursued between the highly multiplied genera of Presl and J. Smith, and the too meagre enumerations of Willdenow, Sprengel, Link, Kunze, and others.” By adopting the above view, hundreds of species become massed under one generic name, regardless of their dis- similar habits and anatomical structure. For instance, the genus Polypodium, as characterised in the “ Species Fili- cum,” contains 407 species, amongst which are found the greatest diversity of size and mode of growth, as also in anatomical structure. The small grass-like species cha- racterised by Swartz and others under the genus Gram- mitis, and the many small pinnatifid species represented by Polypodium trichemanoides, Swartz, and P. pectinatuin, Linn., and others, rank with the large fronded genus Drynaria of Bory, and the large decompound subarbores- cent species of the section Phegopteris. This is consequent on all Ferns having round or oblong sori being placed under the genus Polypodium. Upon such data, setting aside habit, the Tree-Fern genus, Alsophila, R. Br., might be correctly referred to Polypodiwm. The same applies to Aspidium and Nephrodiwm, which, between them, contain 227 species of very varied habit and venation. Then, again, all species with linear naked sori diverging from the mid-rib, whether on free or anastomosing veins, and of whatever habit of growth, are placed under Gymno- gramma, of which he enumerates seventy-four species, even including the genus Selliguea of Blume, which differs from the whole of the Gymnogramma alliance in having articu- late venation. With regard to Acrostichum, 167 specics are described, which Fée and other authors have placed under fifteen genera, many presenting a very distinct habit 54 ON GENERA AND SPECIES, of growth and venation, but with the exception of five species of Platyceriwm he retains the whole under the genus Acrostichum. Of the tribe Aspleniece he describes over 300 species, all comprehended under the genus Asplenium, except two species which constitute the two genera Allantodia and Actinopleris, With regard to Svolopendrium, some authors consider it as a genus scarcely separable from Asplenium, but in the “Species Fiicum” it is characterised as a dis- tinct tribe under the name of Scolopendriew. It consists of nine species, which agree in general habit with Asplenvum, and present no special character to justify their separation as a tribe from Aspleniez. We give another instance,—in speaking of Hypoderris, he says :—“ Considering as I do that the primary divisions should be taken as much as possible from the fructili- cation, I have no hesitation in placing Hypoderris among the Dicksonice and near to JVoodsia.” No three genera can be more distinct in natural affinity than these—that is, as regards habit and anatomical structure; the only similitude is between Hypoderris and Woodsia, which have discal punctiform sori, with a very membranous cup indusium, therefore in that character are more properly though artificially related to Cyathea than to Dicksonia, which has marginal sori with cystiform indusia opening outwards. Again, take the case of Matonia, a rare Fern with fan- shaped fronds, in every respect allied to Gleichenia in habit, and also in the ring of the sporangia being hori. zontal, and differing only from Gletchenia in the sori being furnished with an inverted cup-shaped indusium, attached by its centre as in trae Aspidium and Polystichwn, Although such is the case, Sir Wiliam Hooker follows ON GENERA AND SPECIES. 5d Pres] by placing Matonia in alliance with Cyathea, with which in habit and even in technical character no gencric affinity is traceable. In his general arrangement he follows, with a partial exception, the sequence of Pres], arranging the whole under two sub-orders, (leicheniacee and Polypodiacec, the latter consisting of twelve tribes, beginning with Cyathea, including Hymenophyllum, and ending with Acrostichwm. With regard to Asplenium, Polypodium, and Aspidium, he, in a great degree, follows Mettenius and Fee’s arrange- ment and descriptions of Acrostichwm. From what has now been stated it will be seen that, although Sir W. Hooker bestows great praise on his con- temporaries, he nevertheless adopts but few of their genera, thus making it appear that but little advance has been made in our knowledge of the natural habit and anato- mical structure of Ferns since the time of Sprengel (1827). He, however, finds himself compelled to divide his large genera into sections, which are equivalent to the genera of authors ; indeed, it is only by the sections that we are saved the labour of wading through 200 or 300 species in order to find out a single individual. The following is a tabular view of the sequence of the tribes and genera as arranged in the ‘ ‘Species Filicum ” :— Orvex I.—FILICES. Sub-Order I.—GLEICHENIACEE. Gen. 2. Platyzoma, Gleichenia. Sub-Order 11. —PouyPoD1acee. Tribe I,~—-CYATHEE. Gen. 3. Cyathea, Hemitelia, Alsophila, 56 Gen. ON GENERA AND SPECIES. Tribe I1.—Dicxsoniez. Sub-Tribe I.—HypopERRIDEE. 1. Hypoderris. Sub-Tribe 11.—Woopsiem. Gen. 4. Spheropteris, Diacalpe, Arachniodes, Woodsia. Sub-Tribe I1I.—Eupicksoniea. Gen. 7. Thyrsopteris, Dicksonia, Cibotium, Deparia, Loxoma, Hymenophyllum, Trichomanes. Sub-Order TII.—Davatuirg, Gen, 2. Davallia, Cystopteris. Tribe TV,—LInDSHACES. Gen. 2. Lindsea, Dictyoxyphium. Sub-Order [V.—PvErRIDEz. Gen. 12. Adiantum, Cheilanthes, Pellea, Pteris, &c. Gen. Gen. Gen. Gen. Sub-Order V.—LoManicEz. 5. Lomaria, Blechnum, Sadleria, Woodwardia, Doodia. Sub-Order VI.—ASPLENIER. 3. Asplenium, Allantodia, Actinopteris. Sub-Order VIL.—ScoLopenDRIEx, 1. Scolopendrium, species 8. Sub-Order VIII.—Aspriiacex. 7. Didymochlena, Aspidium, Nephrodium, Nephro- lepis, Oleandra, Fadyena, Onoclea. Sub-Order [X.—Potypopiea, Gen. 1. Polypodium, species 409. Gen. Gen Sub-Order X.—GRaMMITEDER. 11. Jamesonia, Nothochlena, Monogramme, Gym- nogramme, Brainea, Meniscium, Antrophyum, Vittaria, Teenitis, Drymoglossum, Hemionitis. Sub-Order XI.— Acrosticurs, . 2, Acrostichum, Platycerium. ON GENERA AND SPECIES. 57 Oss.—The genera founded by Sir William Hooker, of which there are five, are marked with an asterisk, and the number of species named and described by him in the “* Species Filicum ” amounts to about 600, In a notice at the end of the fifth volume of the “ Species Filicum,” he states, that “if his life and health be spared to him, he intends to accomplish a volume to be entitled ‘Synopsis Filicum,’ to contain brief characters of the sec- tions, genera, and species of Ferns, general habitats, and references to the ‘Species Filicum,’ omitting all really dubious species, and including the sub-orders Osmundacece, Marattiacece, and Ophioglossacee,”’ which it is hoped will form “a useful vade mecum for the travelling botanist, and the cultivator of Ferns, and for ready consultation in the herbarium.” Only one part, with the preface of this work, consisting of twenty-two pages, was printed at the time of the author’s death (which took place in August, 1865); the remaining manuscript was placed in the hands of Mr. J.G. Baker, who, shortly after Sir W. Hooker’s death, received the appointment of Assistant Curator in the Kew Herbarium, and who has carried out the author’s views, forming a volume of 482 pages, which was published in 1868, accompanied by nine coloured plates illustrating the characters of the genera, of which there are seventy-five. In the arrangement of the tribes and genera this work does not differ materially from the “ Species Filicum :” the number of species described of Gleicheniacee and Poly- podiacee amounts to 2,122, being 279 less than the “Species Filicum,” while 106 are enumerated under the sub-orders Osmundacece, Marattiacee, and Ophioglossacece, thus making a total of 2,228 species, being chiefly founded on specimens in the Hookerian Herbarium, which, at the 58 ON GENERA AND SPECIES. time of the publication of the ‘¢ Synopsis,” is stated to con- tain 50,000 specimens of Ferns, representing the many different states of the species therein described. As might be expected from a new writer on Ferns, many changes have been made in the nomenclature and synonyms, as given in the ‘‘ Species Filicum,” there being no less than fifty of the names in the “ Synopsis” having Mr. Baker’s initials prefixed, which is principally con- sequent on his having removed species from one genus to another, and in many cases giving new specific names, as also by the addition of a considerable number of new species. {have already noticed that in the “ Species Filicum” species of very distinct habits and characters are placed in the same genus; of this there are several remarkable new examples in the “Synopsis,” the most conspicuous of which will be found noticed under the respective genera in my general arrangement. In December, 1874, a second edition of the ‘“ Synopsis ” was published, containing an Appendix to the first edition prepared by Mr. Baker, regarding which he says “I have endeavoured in this edition to briefly characterize and fit into their places the new discoveries and the plants found upon fuller information to have been inadequately dealt with in the first.” This Appendix contains the names of 438 species, 198 of which bear the initials of Mr. Baker; the principal authorities for the others being Mettenius, Klotzsch, Karstein, Grisebach, Moore, and Kuhn; with the exception of about twenty, the whole are additions to the 2,228 described in the first edition of the ‘‘ Synopsis;” thus on the authority of that work the total number of known species of Ferns amounts to 2,646. Whether the above number is represented by truly dis- ON GENERA AND SPECIES. 59 tinct species may be doubted by some Pteridologists ; but be that as it may, there can be no doubt but that the “Species Filicum” and “ Synopsis” are highly valuable to students of Ferns,* possessing herbaria or cultivated collections, as also to travellers abroad. I have now to notice another important work by Mr. Baker, published in 1870, namely, that part of Martius’s “ Flora of Brazil,” containing the Ferns; a folio work of 317 pages, with fifty plates, each plate containing many figures of portions of fronds, showing the differential characters of the genera and species. Of the latter there are 387 described, the genera being arranged in the same order as in the “Synopsis Filicum.” On taking into consi- deration the extensive territory of Brazil with its various climates favourable to the growth of Ferns, from those growing at elevations that may be termed sub-arctic, to others luxuriating in the lower hot valleys and rocky or forest ravines, the number of 880 species may be con- sidered small; but here again comes the question, what is a species ? and judging from Mr. Baker’s view, it would appear that many plants originally described as species, which successive authors have acknowledged to be distinct, are, nevertheless, in many cases regarded as synonyms ; thus Ferns lone accepted by previous pteridologists cease to be so. When I say long accepted, I go upon the evidence of Link, Kunze, Schott, Mettenius, and myself, who have had for many years under their observation living examples of species all well recognised as being different from one another by some important characters seen only in the living state, but Mr. Baker, with herba- rium specimens, makes no scruple of lumping many of such *The first published by Dulau and Co., 37, Soho Square, London ; the ‘‘ Synopsis” by Hardwicke, 192, Piccadilly. 60 ON GENERA AND SPECIES. under one specific name. For instance, under Polypodium lycopodioides there are no less than twenty-two synonyms, and under P. brasiliensis eighteen. These examples are additional proof of what has already been said of the confusion of the nomenclature of Ferns, brought about by the different views of Pteridologists. I conclude this part by noticing a memoir, published in 1866, by J. E. Bommer, Secretary of the Royal Botanic Society of Brussels, in the bulletin of the Royal Society of Belgium, vol. 5, No. 38, 1866, entitled “ Monographie de la Classe des Fougéres,” being a review of the writings on Ferns by different authors, beginning with Bernhardi, 1799. He gives an abstract of the classification of the principal authors, but as nothing specially new is brought forward above what is noticed in the preceding pages, it is not necessary to enter into details: he concludes by giving an arrangement of his own which also presents nothing new. The memoir is accompanied by six finely executed plates ; the first shows the different forms of the sporangia and synangia ; the other five plates contain portions of the fronds, illustrating the character of forty-one genera. Abstract showing the number of genera of the preceding authors :— Sprengel, “Systema Vegetabilium ” (1827), 65. Presl, “Tentamen Pteridographie,” and other works, including Hymenophyllea 35, 230. J. Smith’s Arrangement (1841), 143. Fée, Polypodiacese only (1852), 181. Moore’s “Index Filicum ” (1857), 178. Hooker’s “ Species Filicum” 5 vols. (1864), 66. Hooker’s ‘‘ Synopsis Filicum ” (1874), ed. 2., 76. J. Smith, in the present work, 220, ON GENERA AND SPECIES, 61 5.— ON THE GENERA AND THEIR CLASSIFICATION, AS FOUNDED ON THEIR MODES OF GROWTH AND HABIT. In the preceding pages I have given a general view of the principal writers on Ferns, and their systems of classi- fication. I will now proceed to explain the reasons that have induced me to attempt to draw up a natural arrange- ment of Ferns, reviewing as I proceed the data upon which the genera are founded. In my first general arrangement in 1841, I characterised the whole of the circinate Ferns under divisions and tribes, as shown in the abstract of my arrangement, given at page 33, the characters of which are in general obvious, so that any Fern with perfect fructification can readily be referred to its proper tribe. A system of classification so based is therefore practically useful, but it must be viewed as purely artificial, for, although many of the genera in any one tribe are naturally allied in general habit, yet others of totally different habit, but possessing the technical character of the tribe, are found associated with them; while on the other hand genera closely allied in natural habit and organisation are placed in separate tribes. For example, the genus Gonispteris has naked sori, while its congener Nephrodium has indusiate sori, they are conse- qtiently placed apart in any arrangement which is based on the absence or presence of indusia. In both these genera the species are numerous, and in many cases are so alike that in the absence of the indusium, specimens of Nephro- dium are referred to Goniopteris, The same applies to the extensive genera Lasérea and Phegopteris, both of which contain closely allied species. When the indusium is absent in normal indusiate genera, it is in general impossible, even by the most careful micro- 62 ON GENERA AND SPECIES. scopic examination to discover that it ever was present; that organ, therefore, cannot be depended upon for determining the genera of the tribe Phegopteridece by herbarium speci- mens. The relation of one genus to another is also governed by the fertile frond being more or less contracted ; for instance, all contracted fronds with naked sporangia closely occupy- ing the whole underside, or in irregular masses, are referred to the tribe Acrostichece ; but in some this character is not always perfect. For instance, in the genus Pecilopteris, the fertile fronds vary in their degree of contraction, and some being but slightly contracted, the veins being distinct, forming obvious well-defined receptacles,* which, when viewed in conjunction with the general habit and mode of venation, indicates the relationship of Peecilopteris to be with Menisciwn. This affinity is further confirmed on comparing it with the original herbarium specimens of Deniscium simplex (Lond. Journ. Bot. v. 1, t. 11), which are strictly characteristic of the genus Meniscium, but since the species was first described it has been introduced in a living state to the garden collection, and has produced fertile fronds so completely contracted, that if such state had been in the first instance observed by the learned author, he would not have hesitated in referring it to Acrostichew. Other instances also occur to show that Meniscium and Peeilopteris are naturally related. I have now said sufEcient to show that the placing of such examples of naturally allied species in different genera must be admitted as only an artificial convenience for arrangement. The various forms and direction of the sori also consti- * On this state Sir William Hooker founded the genus Jenkinsia. — Hook. Gen. Fil, t. 75, B. ON GENERA AND SPECIES. 63 tute an important part in defining gencra; but in many cases if the characters derived from the form are strictly adhered to, species possessing very different appearances are brought together. For instance, the small grass-like Ferns of the genus Grammitis of Swartz are characterised by having linear naked sori; but by adhering to this simple character alone, species of very different habit would be included, as Gymnogramma, Stegnogramma, Loxo- gramma, Plewrogramma, Sellegua, and others, which, as already stated, has to a considerable extent been so done in the “Species Filicum.” The sori are also liable to vary in form and size in different specimens of the same species, or even on the same, or different fronds of the same plant. This usually happens by the contiguous punctiform recep- tacles uniting laterally, when they assume on oval, oblong, or linear form. Examples of this may be found in the genera Pleopeltis, Phymatodes, Pleuridiwm, Drynaria, Aspi- dium, and otbers. A remarkable instance of this occurs in the splendid Fern Drynaria coronans (Polypodium coronans, Wall.), a species with large pinnatifid fronds, which in the presumed normal state as generally observed, has the sori distinctly punctiform, and produced in a single row between each pair of the primary veins. In some fronds, however, the sori are so completely united in a line as to form a true linear sorus, and hence not only agree perfectly with the character of Grammitis, as defined by Swartz, but also with the genus Gymnogranuna, as characterised in the “ Species Filicum.” These examples show that the form of the sori alone cannot be strictly depended upon for defining genera, In previous remarks I have stated that the mode in which the vascular structure traverses the fronds (vena- tion) affords important data in assisting to define the limits of genera. O4 ON GENERA AND SPECIES. This is particularly the case in the extensive tribes Polypodiece, Acrostichece, Phegopteridece, and Aspidiece, each of which contains groups of species possessing similar venation, and agreeing so perfectly in habit and mode of growth, that, in the absence of the fructification, it is often dificult to determine to which genus and tribe they should be referred. For instance, Phegopteris and Lastrea, Poly- betrya and Polystichum, Goniopteris and Nephrodium, Dryo- menes and Aspidiwm, respectively agree both in habit and venation, and all, with the exception of Polybotrya, have punctiform sori, and, as already stated, these and kindred genera can only be known from one another by the presence or normal absence of the indusium, or by the fertile fronds being contracted, as in Polybotrya. There- fore, taking all these points into consideration, it appears to me that, in order to arrive at the true affinity of Ferns, the vascular structure, in conjunction with mode of growth, must be viewed as forming the groundwork of any at- tempted natural classification. It must, however, be under- stood that differences in venation do not always indicate distinct modes of growth or habit ; for example, Phegopteris and Goniopteris, Polybotrya and Saromanes, Pteris and Lito- brochia, Adiantum and Hewardia, Lastrea and Nephrodium, Lindseea and Synaphlebium, respectively contain species perfectly analogous in habit and other characters, and only generically distinguished by the venation being free, or anastomose. Another circumstance to be considered is whether free or anastomose venation is always constant in the same species. In any uncertainty upon this point it is necessary to be first assured that the specimens under examination have arrived at sufficient age to have the normal structure of the species fully developed, for in fronds of young plants the venation is often free while in or ON GENERA AND SPECIES. 6 older or more perfect fronds of the plant, or different plants of the same species, it is anastomose. In some cases, however, the normal structure of the venation is partly anastomose and partly free, as in the genera Goni. opteris, Nephrodium, Henvitelia, &e. In these genera the lower or basal venules of each parcel are anastomose, while the upper are free. In some genera, such as Adian- tum, Lindscea, and Aspleniwn, the venation is normally free ; anastomoses occasionally occur in different segments of the same frond, but such seem more accidental than normal, From what T have now stated it will be seen that neither the form or position of the sori or venation, the absence or presence of indusium, taken singly or combined, can be depended on for determining natural genera. Such being the case, and having had under my notice examples of nearly 1,000 species of living Ferns, has led me to the con- clusion that in order to form natural genera the different modes of growth must be taken into account, that is the manner in which the developing axis (caudex) increases, its direction, and the mode of evolution and attachment of the fronds. I find that Ferns present two very distinct modes of growth, which are comparatively as distinct as Exogens are from Endogens. My views on this subject first appeared in the “ Botany of H. M. 8. Herald,” pub- lished in 1858, but as that work may not be accessible to all engaged in the study of Ferns, I will repeat the sub- stance of what is there given. If Polypodium vulgare be examined, it will be seen that its fronds are produced singly from the sides of a creeping progressing stem, which has its accrescent apex always in advance of its young developing frond, each successive frond being produced singly from special nodes formed at more or tN 66 ON GENERA AND SPECIES. less distance apart, and in an alternate manner on the sidos of the progressing axis, the foot or base of attachment of the frond forming with the node a distinct and well marked articulation or joint, which ultimately becomes the point of separation of the mature frond, the node remaining per- manent in the form of a round more or less elevated cicacrix.. This mode of growth includes all Polypodiaceer, as restricted by me to the tribes Oleandree, Davallice, Poly- podece, and Platyceree. To this mode of growth I have given the name Hremobrya, and in characterising the genera the words vernation articulate are employed, and the pro- gressing axis from which the fronds are developed surculumn. The number of species referable to Hremobrya, however, are few in comparison with those which possess a very different mode of growth. If, for example, a plant of Lastrea Filie-mas be examined, the fronds will be found to be successively produced in a spiral manner from the apex of the stem, in the form of a corona or crown, and thus by successive development the stem is increased in height, All Ferns developing fronds in this manner, even the smallest, the axis of which does not rise above the ground, to lofty tree Ferns, belong to this division as also a numerous class which produce fronds similar to the above, but which form adecumbent stem lying on the ground or but slightly inclined upwards. This I term vernation fasciculate decum - bent, and the preceding vernation fasciculate erect. Another mode of growth is common to many Ferns, such for example in Pleris aquilina, Lastrea serra, Polybotrya osmun- ducea, the fronds are normally produced alternately, in a single series, from the apex of a decumbent or creeping: ster (sarmentum). In some cases this mode of growth resembles that of Eremobrya, but from which it is quite distinct, and is readily known hy the vascular structure of ON GENERA AND SPEUIES, 67 the stipes being continuous and intimately connected with the portion of the stem forming the internode between the fronds, from which they cannot be separated without lacera- tion. In these three modes of growth—that is, whether the fronds are produced in a fascicle, forming an erect or decumbent axis, or in a single series—the principle of at- tachment is identical, and there being no articulation, as in Eremobrya, the bases of the stipes are so united and become indurated, forming the solid portion of the stem. To this mode of growth I have given the term Desmobrya, and in the definitions of the genera the venation is termed adherent. The various modifications of Desmobrya | have found useful as auxiliary characters in indicating natural genera, but there are exceptional cases of species of Lastrea and Nephrodium, with erect fasciculate vernation, not distinguish- able by the character of their fronds from other species of the same genera with decumbent uniserial vernation ; nevertheless the adoption of the vernation and the nature of the accrescent axis as part of the definition of genera is of much assistance in determining the relative affinities of many genera. In general the difference between Hremobrya and Des- mobrya, as above described, is obvious, and there can be no doubt that the two modes of growth are specially dis- tinct, the plants being endowed with very different natures, the vital power and tenacity of life, and power of resisting drought being much stronger in Eremobrya than in Desmo- brya ; and it is not a little remarkable that so far as obser- vations upon cultivated plants enable me to ascertain, the latter are much more freely reproduced from spores than the former. It is also worthy of notice that in the importation of Ferns from distant countries, those of 68 ON GENERA AND SPECIES, the division Eremobrya generally arrive in good con- dition, while those of Desmobrya, especially such as have slender stems. are often dead or in a very enfeebled state on their arrival. The difference in the structure and mode of growth of the species of these two divisions may be admitted to be of equal importance for classification with the difference in structure of Exogens and Endogens in flowering plants. In all attempts to determine natural relationship on any special organ or structure, exceptions are never wanting. In the present case, examples of connecting links are in a few instances found between the two divisions. This is particularly so with the genus Hlaphoglossum, which con- sists of numerous simple-fronded Ferns, and may be viewed asa truly natural genus. In many of the species the stipes have an evident articulation situated in some, more than half an inch above its base, and in many of the species it is marked by a slight swelling ; and also by the part below the swelling being of a dark colour. No change, however, takes place in the structure, the bun- dies of vascular fibre pass uninterruptedly through the apparent articulation, but they are evidently weaker at that point; for, when mature, the stipes separate there, the base portion remaining attached to the developing axis, which is either fasciculate or uniserial and adherent. I therefore place Elaphoglossum in Desmobrya. Another instance of the stipes being jointed above its base is characteristic of the original species of the genus \WWoodsia, as also in species of Arthropteris ; but the base of the stipes being adherent, indicates their position to be in Desmobrya. Also special examples of the articulation of the stipes in fasciculate vernation is exhibited by the genus. Trichopleris and several species of Cyathea; for ON GENERA AND SPECIES. 69 instance, in C, arborea and C. serra, the fronds are terminal, forming the characteristic corona common to tree Ferns. When mature they separate from the stem, leaving a clean scar similar to Eremobrya, thus differing from the adherent vernation characteristic of the mass of the species of the tribe Cyathea; but as they agree in all the essential characters as regards the nature of their fronds and fructi- fication, it is therefore necessary to obtain more evidence as to the mode of evolution and structure of these tree Ferns before any satisfactory conclusion cun be arrived at, as to enable us to view it as a special character to mark distinct genera.* This now leads me to notice another family of Ferns —namely, Marattiacece, which have a very distinct and special habit from other Ferns, and are well marked by the sporangia being destitute of a ring. The evolution of the frond is terminal and articulate to the caudex, as in Cyathea arborea above noticed, but differs in the fronds rising from between two stipuliform appendaves in the form of a socket, analogous to the stipules of phenogamous plants. The speciality of this group of Ferns leads me to view it as constituting a distinct division, which I have named Scaphobrya. This division comprehends but a small number of species as compared with Kremobrya and Desmobrya, the number probably not exceeding thirty or forty, the principal part of which are included under the genera Maratlia, Angi- opteris and Daneu. * Since the above was written, a valuable memoir ‘‘ On the Structure and Formation of the Stems of the Tree Ferns of India,” has been published in the thirtieth volume of the Transactions of the Linnean Society, 1874, by Mr. J. Scott, Curator of the Botanic Gardens, Calcutta. It consists of thirty-nine pages of letterpress, accompanied by eighteen plates. 70 ON GENERA AND SPECIES. T have also to notice that articulation is not confined to the stipes alone ; ; the segments of fronds—lacinie, pinne, and pinnules—being in many instances artica- late with their respective rachis, whether primary or secondary. This is particularly characteristic of the genera Drynaria, Arthromeris, Schellolepis, and Photino- pterts in Eremobrya; and Nephrolepis, Isoloma, Didymo- chlena, Cystodium, and others in Desmobrya, as also in Scaphobrya. By paying attention to these points they are found serviceable in determining the natural affinity of species. Several of the above genera, as also Arthrop teris, Cerato- pteris, Platycerium, and Amphiblestra, possess characters peculiar to themselves, which necessitates their considera- tion as aberrant genera; cousequently the nature of their fructification, venation, and habit, must be conjointly viewed in determining the tribe to which they seem most naturally to belong. On takine a retrospective view of the preceding obser- vations, it will be seen that every defined form of the organs or structure that serve for defining genera, whether artificial or natural, are liable to vary more or less from the adopted typical normal form, and thus afford ample scope for Pteridologists to name and classify genera according to their views of the value of the several organs. On this point two authors seldom agree; hence great confusion is introduced into the nomenclator of Ferns. Some may be called conservative, such as Hooker and Mettenius, who maintain entire the old genera of Linneus and Swartz; while others, as shown in the pre- ceding pages, break up these large genera into a number of smaller ones. These different views are consequent on there being no ON GENERA AND SPECIES. 71 law of nature to mark where a genus begins or where it ends. On this point, in ny “ Avrangement of the Genera of Ferns,” published in 1841, I have said—‘A gradual transition of form is seen to pervade through the whole, not in a lineal, but in a complex reticulated manner, it becomes very difficult to determine in which genus the extreme or transition species of groups should be placed.” On considering that in many cases habit is not excluded in assisting to define the genera of flowering plants, and in numerous instances it is allowed by eminent botanists to constitute the chief distinction between allied genera. For example, the order Liliacece is represented by nearly 1,700 species, arranged under about 150 genera, which all agree as regards their parts of fiuctification, the slight variations of which are made use of for defining genera ; and, indeed, if it was not for the difference in habit, the number of genera are capable of being greatly reduced. This and other instances that might be quoted favour my view, that in order to constitute natural genera of Ferns, habit must play an important part; and so long as any number of species agree in habit, and possessing other characters in common, it matters not whether such groups are called genera, sub-genera, or sections of genera, For my own part, I prefer regarding them as genera, as it saves a great deal of unnecessary trouble in speaking or writing about Ferns, it being much easier to say Drynaria coronans than Polypodium (Drynaria) coronans, or Blapho- glossum. conforme than