Botany

BOTANY

THE science of Botany includes everything relating to the Vegetable Kingdom, whether in a living or in a fossil state. Its object is not, as some have supposed, merely to name and arrange the vegetable productions of the globe. It embraces a consideration of the external forms of plants —of their anatomical structure, however minute—of the functions which they perform—of their arrangement and classification—of their distribution over the globe at the present and at former epochs—and of the uses to which they are subservient. It examines the plant in its earliest state of development, when it appears as a simple cell, and follows it through all its stages of progress until it attains maturity. It takes a comprehensive view of all the plants which cover the earth, from the minutest lichen or moss, only visible by the aid of the microscope, to the most gigantic productions of the tropics. It marks the relations which subsist between all members of the vegetable world, and traces the mode in which the most despised weeds contribute to the growth of the mighty denizens of the forest.

The plants which adorn the globe more or less in all countries must necessarily have attracted the attention of mankind from the earliest times. The science that treats of them dates back to the days of Solomon, for that wise monarch “spake of trees,” from the cedar of Lebanon to the hyssop on the wall. The Chaldeans, Egyptians, and Greeks were the early cultivators of science, and Botany was not neglected, although the study of it was mixed up with crude speculations as to vegetable life, and as to the change of plants into animals. Aescu- )dpius and his priests, the Asclepiades, who studied the art of medicine, had their attention directed to plants in a pharmaceutical point of view. About 300 years before Christ Theophrastus wrote a History of Plants, and described about 500 species used for the treatment of diseases. Dioscorides, a Greek writer, who appears to have flourished about the time of Nero, issued a work on Materia Medica. The elder Pliny described about a thousand plants, many of them famous for their medicinal virtues. Asiatic and Arabian writers also took up this subject. Little, however, was done in the science of botany, properly so called, until the 16th century of the Christian era, when the revival of learning dispelled the darkness which had long hung over Europe. Brunfels, a physician of Bern, has been looked upon as the restorer of the science in Europe. He published a History of Plants, illustrated by figures, about the beginning of the 16th century.

One of the earliest attempts at a methodical arrangement of plants was made in Florence by Andreas Caesalpinus, a native of Arezzo, some time professor of botany at Padua, and afterwards physician to Pope Clement VIII. He is called by Linnaeus primus vents systematic™. In his work De Plantis, published at Florence in 1583, he distributed the 1520 plants then known into fifteen classes— the distinguishing characters being taken from the fruit.

John Ray, a native of Essex, did much to advance the science of botany. He was born in 1628, and died in 1705. He promulgated a system which may be considered as the dawn of the “natural system” of the present day (Ray, Methodus Plantarum, 1682). He separated flowering from flowerless plants, and divided the former into Dicotyledons and Monocotyledons. His orders were founded on a correct idea of the affinities of plants, and he far outstripped his contemporaries in his enlightened views of arrangement.

About the year 1670 Dr Robert Morison of Aberdeen published a systematic arrangement of plants. He divided them into eighteen classes, distinguishing plants according as they were woody or herbaceous, and taking into account the nature of the flowers and fruit. In 1690 Rivinus promulgated a classification founded chiefly on the forms of the flowers. Tournefort about the same time took up the subject of vegetable taxonomy. He was a contemporary of Ray, and was professor of botany at Paris in 1683. He was long at the head of the French school of botany, and published a systematic arrangement in 1694-1700. He described about 8000 species of plants, and distributed them into twenty-two classes, chiefly according to the form of the corolla, distinguishing herbs and under-shrubs on the one hand from trees and shrubs on the other. The system of Tournefort was for a longtime adopted on the Continent, but was ultimately displaced by that of Linnaeus.

Carl von Linné, or, as he is commonly called, Linnaeus, was born on the 23d of May 1707, at the village of Rooshoolt (Râshult),in Smaland, a province of Sweden, where his father, Nicholas Linnaeus, was clergyman. He entered as a pupil at the University of Lund, and about the years 1727-28 was received into the house of Stobaeus, a physician in that city, where he had abundant opportunities of prosecuting natural history. He afterwards proceeded to Upsal, and had to struggle with great difficulties during his studies there. He aided Celsius in his Hierobotanicon, or account of the plants of Scripture, and he became assistant to Rudbeck, professor of botany. He afterwards travelled in Lapland, took his degree in Holland, visited England, and commenced practice in Stockholm, where he lectured on botany aud mineralogy. He finally became professor of botany at Upsal, and was one of the most popular lecturers of the day. He died on the 8th of January 1778, in the 71st year of his age. His herbarium is now in the possession of the Linnean Society.

One of his biographers, in summing up his merits, says, —“Educated in the severe school of adversity, accustomed from his earliest youth to put a high value on verbal accuracy and logical precision, endowed with a powerful understanding, and capable of undergoing immense fatigue, both of body and mind, Linnaeus produced a most important revolution in botanical science. He improved the distinctions of genera and species, introduced a better nomenclature on the binomial method, and invented a new and comprehensive system founded on the stamens and pistils. His verbal accuracy and the remarkable terseness of his technical language reduced the crude matter that was stored up in the folios of his predecessors into a form which was accessible to all men. He separated with singular skill the important from the unimportant in their descriptions. He arranged their endless synonyms with a patience and a lucid order that were quite inimitable. By requiring all species to be capable of a rigorous definition, not exceeding twelve words, he purified botany from the endless varieties of the gardeners and herbalists; and by applying the same strict principles to genera, and reducing every character to its differential terms, he got rid of the cumbrous descriptions of the old writers. It is said of Linnaeus, that, although no man of science ever exercised a greater sway, or had more enthusiastic admirers, yet his merit was not so much that of a discoverer as of a judicious and strenuous reformer. The knowledge which he displayed, and the value and simplicity of the improvements which he proposed, secured the universal adoption of his suggestions, and crowned him with a success altogether unparalleled in the annals of science.”

The system of Linnaeus is founded on the sexes of plants, and hence it is often denominated the sexual system. It is called an artificial method, because it takes into account only a few marked characters in plants, and does not propose to unite them by natural affinities. It is an index to a department of the book of nature, and as such is useful to the student. It does not aspire to any higher character, and although it cannot be looked upon as a scientific and natural arrangement, still it has a certain facility of application which commends it to the tyro. In using it, however, let it ever be remembered, that it will not of itself give the student any view of the true relations of plants as regards structure and properties, and that by leading to the discovery of the name of a plant, it is only a stepping-stone to the natural system. Linnaeus himself claimed nothing higher for it. He says—“Methodi Naturalis fragmenta studiose inquirenda sunt. Primum et ultimum hoc in botanicis desideratum est. Natura non facit saltus. Plantae omnes utrinque affinitatem monstrant, uti territorium in mappa geographica.” Accordingly, besides his artificial index, he also promulgated fragments of a natural method of arrangement.

The Linuean system was strongly supported by Sir James Edward Smith, who adopted it in his English Flora, and who also became possessor of the Linnean collection. The system was for a long time the only one taught in the schools of Britain, even after it had been discarded by those in France and in other Continental countries.

The foundation of Botanic Gardens during the 16th and 17th centuries did much in the way of advancing botany. They were at first appropriated chiefly to the cultivation of medicinal plants. This was especially the case at universities, where medical schools existed. The first Botanic Garden was established at Padua in 1545, and was followed by that of Pisa. The garden at Leyden dates from 1577, that at Leipsic from 1579. Gardens also early existed at Florence and Bologna. The Montpellier Garden was founded in 1592, that of Giessen in 1605, of Strasburg in 1620, of Altorf in 1625, and of Jena in 1629. The Jardin des Plantes at Paris was established in 1626, and theUpsal Garden in 1627. The Botanic Garden at Oxford was founded in 1632. The garden at Edinburgh was founded by Sir Andrew Balfour and Sir Robert Sibbald in 1670, and, under the name of the Physic Garden, was placed under the superintendence of James Sutherland, afterwards professor of botany in the university. The park and garden at Kew date from about 1730. The garden of the Royal Dublin Society at Glasnevin was opened about 1796; that of Trinity College, Dublin, in 1807; and that of Glasgow in 1818. The Madrid Garden dates from 1763, and that of Coimbra from 1773. Gesner states that at the end of the 18th century there were 1600 Botanic Gardens in Europe.

A new era dawned on botanical classification when Antoine Laurent de Jussieu appeared. He was born at Lyons in 1748, and was educated at Paris under the care of his uncle, Bernard de Jussieu. At an early age he became botanical demonstrator in the Jardin des Plantes, and was thus led to devote his time to the science of botany. Being called upon to arrange the plants in the garden, he necessarily had to consider the best method of doing so, and adopted a system founded in a certain degree on that of Ray, in which he embraced all the discoveries in organography, adopted the simplicity of the Linnean definitions, and displayed the natural affinities of plants. His Genera Plantarum, begun in 1778, and finally published in 1789, indicated an important advance in the principle of classification. Jussieu subsequently became professor of rural botany; he died in 1836 at the age of 88.

The system of Jussieu made its way slowly in Great Britain, and it w as not until Robert Brown brought it under notice that it was adopted. It is now the basis of all natural classifications. One of the early supporters of this natural method was Augustin Pyrame De Candolle, who was born in 1778, aud who, after attending the lectures of Vaucher at Geneva, devoted himself to botanical pursuits. He subsequently prosecuted his studies at Paris, and lectured on botany at the College of France. He commenced his publications in 1802, and in 1804 he promulgated his Elementary Principles of Botany. In 1807 he became professor of botany at Montpellier, and in 1816 he was appointed to the chair of natural history at Geneva, with the charge of the Botanic Garden. In that city he carried on his future botanical labours, and began his Prodromus Systematis Naturalis Regni Vegetabilis, which was intended to embrace an arrangement and description of all known plants. He was enabled to complete eight volumes of the work before his death, aud it has since been carried on by his son Alphonse De Candolle, with the aid of other eminent botanists. It now embraces descriptions of the genera and species of Dicotyledonous plants. The system followed by De Caudolle is a modification of that of Jussieu, and it is adopted more or less at the present day. De Candolle’s own herbarium was extremely rich. He had visited and carefully examined many of the most extensive collections, especially those of Paris; and many entire collections, as well as separate families, on which he was specially engaged, were from time to time submitted to his examination by their possessors. He had thus opportunities of comparison greatly beyond what in ordinary circumstances fall to the lot of an individual. His library, too, was stored with almost every important publication that could be required for his undertaking. With such ample materials, aided by his untiring zeal and the persevering energy of his character, he steadily pursued his allotted task, and only ceased to labour at it when he ceased to live. For some years his health declined, and it is to be feared that the severe and incessant attention which he paid to the elaboration of the great family of Compositae had made a deep inroad upon it. As a relaxation from his labours he undertook in the last years of his life a long journey, and attended the scientific meeting held at Turin; but he did not derive from this the anticipated improvement in his health, which gradually failed until his death on the 9th September 1841. Since De Candolle’s time various modifications of his system have been introduced by Endlicher, Lindley, Hooker, and Bentham.

In arranging plants according to a natural method, we require to have a thorough knowledge of structural and morphological botany, and hence we find that the advances made in these departments have materially aided the efforts of systematic botanists.

Robert Brown, a Scottish botanist, was the first in this country to support and advocate the natural system of classification. The publication of his Prodroimιs Florae Novae Hollandiae, according to the natural method, led the way to the adoption of that method in the universities and schools of Britain. Sir William (then Dr) Hooker, in his prelections in the University of Glasgow, and in his numerous writings, ably supported Brown. John Lindley also came into the field, and in 1830 published the first edition of his Introduction to the Natural System. Dr Robert Kaye Greville and Dr Walker Arnott were able coadjutors, more especially in the department of Cryptogamie Botany. From the year 1832 up to 1859 great advances were made in systematic botany, both in Britain and on the continent of Europe. Endlicher’s Enchiridion and Genera Plantarum, De Candolle’s Prodromus, and Lindley’s Vegetable Kingdom became the guides in systematic botany, according to the natural system.

The following remarks embrace the views of Mr Bentham on the change from the Linnean to the natural system of classification :—“The change from the technical to the scientific study of plants was now complete. The Linnean platform, established on the relation of genera and species, had now been so long and so universally adopted as the basis or starting point, that the credit due to its founder was almost forgotten, and it was superseded by the Jussiean method, although it was chiefly by the consistent following out the principles laid down by Linnaeus himself that the change had been effected. Plants were now grouped upon a philosophical study of their affinities, whether morphological, structural, or physiological.”

In all classification it is necessary to define what is meant by species. The usual definition of the term has been that a species (as regards the present epoch of the earth’s history) is an assemblage of individuals having characters in common, and coming from an original stock or protoplast, and their seeds producing similar individuals. It was also supposed that variation in species was restrained within certain limits, and that varieties had a tendency to revert to the parent form. The view, however, adopted by many now-a-days is, that the tendency to variation is continuous, and that, after a lapse of long periods of time, and under the influence of varying external conditions, the descendants from a common stock may exhibit the differences which characterize distinct species. These are the views which are advanced by Darwin, and which imply a complete revolution in our idea of species. This theory is thus stated by Bentham :—

1. That although the whole of the numerous offspring of an individual plant resemble their parent in all main points, there are slight individual differences.

2. That among the few who survive for further propagation, the great majority, under ordinary circumstances, are those which most resemble their parent, and thus the Species is continued without material variation.

3. That there are, however, occasions when certain individuals, with slightly diverging characters, may survive and reproduce races, in which these divergencies are continued even with increased intensity, thus producing Varieties.

4. That in the course of an indefinite number of generations circumstances may induce such an increase in this divergency, that some of these new races will no longer readily propagate with each other, and the varieties become Νeιv Species, more and more marked as the unaltered or less altered races, descendants of the common parent, have become extinct.

5. That these species have in their turn become the parents of ≡of species, that is Genera, Orders, he., of a higher and grade, according to the remoteness of the common parent, and more or less marked, according to the extinction or preservation of unaltered primary, or less altered intermediate, forms.

As there is thus no difference but in degree between a variety and a species, between a species and a genus, between a genus and order, all disputes as to the precise grade to which a group really belongs are vain. It is left in a great measure to the judgment of the systematist, with reference as much to the use to be made of his method as to the actual state of things, how far he should go in dividing and subdividing, and to which of the grades of division and subdivision he shall give the names of Orders, Sub-orders, Tribes, Genera, Subgenera, Sections, Species, Sub-species, Varieties, &c., with the consequent nomenclature.

Such a systematic arrangement is founded on a hypothesis which, so far as the present flora of the globe is concerned, cannot be demonstrated. Conjecture is hazarded as to the present epoch of the earth’s history, by extending back to unlimited ages. If the theory is consistent with what we see around us, and is founded on plausible grounds, then we must think that we have ascertained the plan followed by the great Creator, Designer, and Upholder of all things, that we have been able to ascertain and follow His workings, and the mode an which He has created the diverse plants which have covered our globe in time and space. This new phase of systematic botany, however, requires more definite data to lead to its adoption as an explanation of the plan of creation.

The Physiology of plants did not keep pace with the advance in Classification. Grew and Malpighi were the earliest discoverers in this department of botany. Hales also contributed to it by his observations on the motion of fluids in plants. The subject of fertilization was one which early excited attention.

The idea of the existence of separate sexes in plants was entertained in early times, long before separate male and female organs had been demonstrated. The production of Dates in Egypt, by bringing two kinds of flowers into contact, proves that in very remote periods some notions were entertained on the subject. Female Date I ’ alms only were cultivated, and wild ones were brought from the desert in order to fertilize them. Herodotus informs us that the Babylonians knew of old that there were male and female Date-trees, and that the female required the concurrence of the male to become fertile. This fact was also known to the Egyptians, the Phoenicians, and other nations of Asia and Africa. The Babylonians suspended male clusters from wild Dates over the females; but they seem to have supposed that the fertility thus produced depended on the presence of small flies among the wild flowers, which, by entering the female flowers, caused them to set and ripen. The process was called palmification. Theophrastus, who succeeded Aristotle in his school in the 114th Olympiad, frequently mentions the sexes of plants, but he does not appear to have determined the organs of reproduction. Pliny, who flourished under Vespasian, speaks particularly of a male and female Palm, but his statements were not founded on any real knoλvledge of the organs. From Theophrastus down to Caesalpinus, who died at Rome in 1603, there does not appear to have been any attention paid to the reproductive organs of plants. Caesalpinus had his attention directed to the subject, and he speaks of a halitus or emanation from the male plants causing fertility in the female.

Grew seems to have been the first to describe, in a paper on the A≡toNiy of Plants, read before the Royal Society in November 1676, the functions of the stamens and pistils. Up to this period all was vague conjecture. Grew speaks of the attire, or the stamens, as being the male parts, and refers to conversations with Sir Thomas Millington, Savilian professor at Oxford, to whom the credit of the sexual theory seems really to belong. Grew says that, “when the attire or apices break or open, the globules or dust falls down on the seedcase or uterus, and touches it with a prolific virtue.” Ray adopted Grew’s views, and states various arguments to prove their correctness in the preface to his work on European plants, published in 1694. In 1694 Camerarius, professor of botany and medicine at Tübingen, published a letter on the sexes of plants, in which he refers to the stamens and pistils as the organs of reproduction, and states the difficulties he had encountered in determining the organs of Cryptogamie plants. In 1703 Samuel Morland, in a paper read before the Royal Society, stated that the farina (pollen) is a congeries of seminal plants, one of which must be conveyed into every ovum or seed before it can become prolific. In this remarkable statement he seems to anticipate in part the discoveries afterwards made as to pollen tubes, and more particularly the peculiar views promulgated by Schleiden. In 1711 Geoffroy, in a memoir presented to the Royal Academy at Paris, supported the views of Grew and others as to the sexes of plants. He states that the germ is never to be seen in the seed till the apices (anthers) shed their dust; and that if the stamina be cut out before the apices open, the seed will either not ripen, or be barren if it ripens. He mentions two experiments made by him to prove this—one by cutting off the staminal flowers in Maize, and the other by rearing the female plant of Mercurialis apart from the male. In these instances most of the flowers were abortive, but a few were fertile, which he attributes to the dust of the apices having been wafted by the wind from other plants.

Linnaeus was the next botanical author who took up the subject, and by his sexual system he may be said to have opened a new era in the history of botany. He first published his views in 1736, and he thus writes—“Antheras et stigmata constituere sexum plantarum, a palmicolis, Mill-ingtono, Grewio, Rayo, Camerario, Godofredo, Morlando, Vaillantio. Blairio, Jussievio, Bradleyo, Royeno, Logano, <fcc., detectum, descriptum, et pro infallibili assumptum; nec ullum, apertis oculis considerantem cujuscunque piant« flores, latere potest.” He divided plants into sexual and asexual, the former being Phanerogamous or flowering, and the latter Cryptogamoαs or flowerless. In the latter division of plants he could not detect stamens and pistils, and he did not investigate the mode in which their germs were produced. He was no physiologist, and did not promulgate any views as to the embryogénie process. His followers were chiefly engaged in the arrangement and classification of plants, and while descriptive botany made great advances, the physiological department of the science was neglected. His views were not, however, adopted at once by all, for we find Alston stating arguments against them in his Dissertation on the Sexes of Plants. Alston’s observations were founded on what occurred in certain unisexual plants, such as Mercurialis, Spinach, Hemp, Hop, and Bryony. The conclusions at which he arrives are those of Pontedera, that the pollen is not in all flowering plants necessary for impregnation, for that fertile seeds can be produced without its influence. He supports parthenogenesis in some plants. Soon after the promulgation of Linnaeus’s method of classification, the attention of botanists was directed to the study of Cryptogamie plants, and the valuable work of Hedwig on the reproductive organs of Mosses made its appearance in 1782. He was one of the first to point out the existence of certain cellular bodies in these plants which appeared to perform the functions of reproductive organs, and to them the name of antheridia and pistillidia were given. This opened up a new field of research, and led the way in the study of Cryptogamie reproduction, which has since been much advanced by the labours of numerous botanical inquirers. The interesting observations of Morland, already quoted, seem to have been neglected, and no one attempted to follow in the path which he had pointed out. Botanists were for a long time content to know that the scattering of the pollen from the anther, and its application to the stigma, were necessary for the production of perfect seed, but the stages of the process of fertilization remained unexplored. The matter seemed involved in mystery, and no one attempted to raise the veil which hung over the subject of embryogeny. The general view was, that the embryo originated in the ovule, which was in some obscure manner fertilized by the pollen.

In 1815 Treviranus roused the attention of botanists to the development of the embryo, but although he made valuable researches, he did not add much in the way of new information. In 1823 Amici discovered the existence of pollen tubes, and he was followed by Brongniart and Brown. The latter traced the tubes as far as the nucleus of the ovule. These important discoveries mark a new epoch in embryology, and may be said to be the foundation of the views now entertained by physiologists, which have been materially aided by the subsequent elucidation of the process of cytogenesis, or cell-development, by Schleiden, Schwann, Mohl, and others. The whole subject has been investigated recently with great assiduity and zeal by physiologists, as regards both Cryptogamous and Phanerogamous plants. The formation of germinal vesicles in the ovule, and the development of the embryo in flowering plants, have been fully considered by Griffith, Schleiden, Mirbel, Spach, Meyen, Schacht, Mold, Unger, Naudin, Radlkofer, and others; the embryogénie process in Coniferous plants and in the higher Cryptogams by Hofmeister, Henfrey, Suminski, Mettenius, Strasburger, Eichler, Bâillon, Cohn, Pringsheim, Millardet; and that of the lower Cryptogams by Thuret, Bornet, Decaisne, and Tulasne. The observations of Darwin as to the fertilization of Orchids. Primula, Linum, and Lythrum, and the part which insects take in this function, have opened up a new era in Physiological Botany. He has been followed by Hermann Müller. Darwin’s experiments in reference to the movements of climbing and twining plants, and of leaves in insectivorous plants, have opened up a wide field of inquiry which he has cultivated with eminent success and with most important results. Among other authors who have contributed to the advance of Vegetable Physiology may be named Hoffmann, Sachs, Van Tieghem, Prillieux, Dehe-rain, and Famintzen. We have thus been enabled to come to certain general conclusions on this obscure subject, and future observers have been directed in the proper path of investigation.

In the Physiological department of botany the most important researches have been made by French and German botanists. The laboratories in connection with schools in Germany offer facilities for study which do not exist to the same extent in Britain. Physiological researches demand not only a Botanic Garden with its appendages, but apparatus of various kinds, means of prosecuting histological and chemical investigations, physical experiments, and observations by the spectroscope. Our schools require then not only lecture-rooms, but laboratories well fitted up with all needful appliances, and salaried assistants to aid the teachers in their demonstrations and the pupils in their practical work.

The department of Geographical Botany has made rapid advance by means of the various scientific expeditions which have been sent to all quarters of the globe; and the question of the mode in which the floras of islands and of continents have been formed has given rise to important speculations by such eminent botanical travellers as Darwin and Hooker. The latter has published a valuable paper on insular floras. Under this department the connection between climate and vegetation has been carefully studied both by botanists and by meteorologists. Among the contributors to this department of botany the following authors may be noticed—Humboldt, Schouw, Meyen, Berghaus, Martius, Harvey, Hooker.

The subject of Palaeontological Botany has been much advanced of late by the researches of botanists and geologists. The use of the microscope in the examination of tissues has aided much in the determination of fossil plants. The more accurate study of Organography has also been the means of correcting errors in diagnosis. The nature of the climate at different epochs of the earth’s history has also been determined from the character of the flora. The works of Brongniart, Goeppert, and Schimper have advanced this department of science. Among others who have contributed valuable papers on the subject may be noticed Heer, who has made observations on the Miocene flora, especially in Arctic regions; Saporta, who has examined the Tertiary flora; Dawson and Lesquereux, who have reported on the Canadian and American fossil plants; and Williamson, who has made a careful examination of many of the coal fossils, and whose excellent drawings of structure have opened a new light on the character of many of the genera. Delineations of fossils by Witham, Lindley and Hutton, and Carruthers, have tended much to advance our knowledge of the fossil flora of Britain.

Botany may be divided into the following departments :— 1. Structural Botany, having reference to the anatomical structure of the various parts of plants, including Vegetable Histology, or the microscopic examination of tissues; 2. Morphological Botany, the study of the form of plantsand their organs—(these two departments are often included under the general term of Organography); 3. Physiological Botany, by some termed Organology, the study of the life of the entire plant and its organs, or the consideration of the functions of the living plant; 4. Systematic Botany, the arrangement and classification of plants; 5. Geographical Botany, the consideration of the mode in which plants are distributed over the different regions of the globe; 6. Paloeontological Botany, the study of the forms and structures of the plants found in a fossil state in the various strata of which the earth is composed.

In the present article we shall confine our attention to the Structure and Morphology of Plants. The limits and classification of the Vegetable Kingdom have been partly considered under Biology (vol. iii. pp. 690-696). The Classification of Plants will be taken up in extenso under the heading Vegetable Kingdom, and the Distribution of Plants in space and time will be treated of in separate articles.

STRUCTURAL ELEMENTS OF PLANTS.

The elementary structure which is the foundation of all vegetable tissue is the cell. In the young succulent bud of a growing stem each cell consists of an outer firm, elastic membrane of cellulose constituting a cell-wall; within this, a gelatinous soft mass of protoplasm, of which there may be a portion distinctly marked off as a nucleus; and, enclosed by the protoplasm, a cell-cavity containing a more or less watery fluid, the cell-sap (fig. 1). Such may be taken as the structure of a typical vegetable cell, which is thus a closed vesicle or sac with fluid or semi-fluid contents. Of these elements of the cell the protoplasm is that which is essential for its growth and development. In it are contained all the substances requisite for the formation of the cell-wall and the cell-sap; and the nucleus is merely a differentiated portion of it. From it then all the other parts of the cell are formed, and it is essential to the growth of the cell. Hence it has received the appellation of primordial cell; and, indeed, amongst many Algae it exists for some time as a separate cell without any cellwall or other part. This must be borne in mind when defining the cell as a sac or vesicle. The growth of the cell is usually, at first, uniform throughout, and it has therefore a more or less rounded form; but, according to the function which it is destined eventually to perform, one or other, or it may be all, of the parts of the cell become modified or specially developed. The cell-wall may be greatly thickened; or it may grow more in one direction than another, so as to be elongated and form protuberances; or perforations may occur; or several similar cells arranged in a longitudinal series may, by obliteration of the interposed septa, unite to form a long tube which is then called a vessel. The protoplasm in the process of growth may be completely absorbed; and when this occurs growth ceases and the cell-walls form merely a framework. It may, however, remain a long time, assuming various shapes and often uniting with colouring matters. The cell-sap also may disappear or may remain, containing in solution, or as definite forms in its mass, various assimilative substances, as fat granules, oil globules, starch, mineral crystals, &c.

In some plants, as amongst Algae (Protococcus), one cell alone performs all the functions necessary for the existence of the plant. We have thus in this cell an epitome of vegetable life, and this is the most perfect form of cell. As we pass to the higher forms of plants, where many cells are united, we find a physiological specialization taking place, by which certain cells are set apart for assimilation, some being embryonal, some supporting, and others protective, &c. Amongst such plants as Ferns and ordinary Flowering Plants a further differentiation takes place, and some of these cells unite to form true vessels. We thus have a means of arranging all plants in two groups, viz., those whose tissues consist entirely of cells, cellular plants (including Fungi, Algae, Mosses, &c.); and those in which vessels are present, vascular plants (including Fems, Lycopods, and ordinary Flowering Plants).

1. Cells and Vessels—Cellular and Vascular Tissues.

Cells united together constitute cellular tissue (fig. 2).

It exists in all plants and abounds in fleshy roots, stems, leaves, and in succulent fruits. It constitutes the pith and outer bark of trees, and is very abundant in the centre of the stem of the Aralia (Fatsia) papyrifera, whence Chinese rice-paper is derived by cutting it into thin sheets. By cultivation the Turnip, Carrot, Cabbage, and other esculent vegetables acquire much cellular tissue, and become tender and succulent. The cells of the tissue vary much in size. In a cubic inch of a leaf of the Carnation there are said to be upwards of three millions of cells. They are frequently seen 1 θ 1 00 th, y ⅜ 7 th, a ∏d 7 ⅜ σ th of an inch in diameter. In some of the Cucumber tribe, and in the pith of aquatic plants, cells ∙ s ⅛th and 3 1 σ th of an inch in diameter occur.

In young cells the cell-wall is a thin membrane consisting of cellulose, with some water and a certain amount of incombustible material. It is permeable by water, is slightly extensible and elastic, and is colourless. It dissolves in_ sulphuric acid, and upon addition of iodine and sulphuric acid assumes a deep blue colour. By intussusception of nutrient material, t.β., the interposition of new molecules between those pre-existing, the cell-wall increases both in surface-extent and in thickness. The resulting cellwall is not, however, uniform in its structure, but is composed of lamellae of different refractive power, in which the cellulose is combined alternately with much and with little water. These alternating dense and watery layers,—of which one set is concentric with the cell-wall, whilst two other series are vertical or oblique to the surface of the cellwall, and cut the concentric ones throughout the whole thickness of the wall,—under a high power of the microscope present a series of mutually intersecting lines, and constitute respectively what are termed the stratification and striation of the cell-wall.

Independently of these changes in the structure of the cell-wall, consequent on its increase in surface-extent and thickness, which will be presently noticed, there are other changes of a chemical nature which take place during the growth of the cell, and which so affect its wall as to break it into distinct “shells,” which differ both chemically and physically from the original cell-wall. Thus, in the epidermis or outer cellular covering of plants, the outermost portion of the outer wall of the cells becomes converted into an elastic substance, quite impervious to water, which acts as a protective covering. This substance is known as cork or cuticular matter. Another alteration is the conversion of the layers of the cell-wall into woody matter, by a process of lignification, or formation of wood. Or, again, layers of the cell-wall may be converted into mucilaginous substance, i.e., absorbing water, and becoming gelatinous, as in the cells of pith of Astragalus Tragacantha,—which furnishes gum tragacanth,—and the outer cells of the seed of the Common Flax. Lastly, mineral matters may be deposited in the cell-wall, such as lime, silica, &c., so abundantly in some instances as to constitute, after burning, a perfect skeleton of the cellwall. In all these cases, however, of alteration of layers of the cell-wall, an innermost layer, giving all the reactions of pure cellulose, may be observed. If growth in surfaceextent proceeded uniformly over the whole of a cell-wall the resulting structure would be a more or less rounded vesicle; but at different points portions grow more rapidly than at others, and thus cells, originally oval or spherical, may become cylindrical, conical, &c. The changes consequent on unequal growth in thickness are, however, much more important, giving rise to altered appearances both on the outside and inside of the cell-wall. The external thickenings are most usually projections in the form of spines, knobs, &c., as in some pollen-grains, and in cilia connected with the reproductive cells of many Algae (tig. 3), or clubshaped hygrometric filaments as in Equisetum (figs. 4, 5); whilst the internal ones are more usually ridges—annular, spiral, or reticulate (figs. 11 and 12)—which may proceed so far as almost to obliterate the cavity of the cell-wall.

The p>rotoplasm, which lines the interior of the cell-wall, and which is the essential living portion of the cell, consists of albuminous substance mixed with water and some incombustible materials, and it also contains some organic compounds. It is a homogeneous, soft, gelatinous substance. As we usually find it in cells it has a granular and turbid appearance. This arises from an admixture of formative matters, to which the name metaplasm has been applied. It is coagulated by heat, and is soluble in a dilute solution of caustic potash; iodine solution colours it yellow or brown, whilst strong sulphuric acid at first colours it rose-red, subsequently dissolving it. Usually, at points in the interior, drops of fluid become differentiated as vacuoli (fig. 6), which may subsequently coalesce, and thus the protoplasm may become a sac containing cell-sap; and if growth of the cell-wall continues the protoplasm eventually forms a mere lining of the cellwall constituting the primordial utricle of Von Mohl. The protoplasm in some cells exhibits phenomena of movement within the cellwall of a definite character. Thus in the internodal cells of Characeae (fig. 7) a movement of protoplasm round the longest diameter of the cell is seen, and in the hairs of Tradescantia (Virginian Spiderwort) a circulation of protoplasm occurs. These constitute the phenomena of rotation and circulation.

The nucleus (fig. 1, k) is present in the cells of all the higher plants. It is a small rounded differentiated portion of the protoplasm, and frequently contains vacuoles, which are termed nucleoli (fig. 1, H). It may be in the centre of the cell or close to the sides, but it may change its position. Portions of the protoplasm are also differentiated as grains or granules, to which colouring matters are attached; but these will be noticed hereafter. The protoplasm in old cells may disappear, and then all growth ceases, and the celh consist of a mere framework; or it may remain, and then growth of the cell continues. And it is by a re-arrangement of the molecules of this protoplasm that the formation of new cells begins,—the nucleus entering also into the process.

By the term cell-sap is meant the fluid contained in the vacuoli. It consists in great part of water, in which are dissolved various salts, derived from without, and compounds formed by assimilation in the plant itself. Amongst the latter we may mention inulin, a substance closely allied to starch and sugar, found in Composite plants.

The term parenchyma (areolar, utricular, or vesicular tissue) is a general name for any form of cellular tissue, in which thin-walled cells of a diameter nearly equal in every direction are united to one another by broad surfaces (fig. 2). If the cells are pointed at both ends and have a length greatly exceeding their breadth, there is formed prosenchyma (fig. 8). Both tissues may be complete (figs. 2 and 8) or incomplete (figs. 9 and 10), i.e., the component cells may touch each other on every side and leave no intercellular spaces, or intercellular spaces may exist between the cells. According to the amount of surface-growth and thickening of the cell-wall various forms of parenchymatous and prosenchymatous tissue result. Thus, in the Rush and Bean we have a stellate parenchyma, with large intercellular spaces (fig. 10), in the Elder pith a complete angular parenchyma (fig 2), and in the succulent stem of the Cactus a spherical incomplete parenchyma (fig. 9). Those forms of tissue in which the individual cells have been altered by thickening of the cell-wall are the most important; and the alterations in the cell-wall consequent on growth in thickness may be such as to produce obliteration of the septum between superposed cells, and their cavities, freely communicating, then give rise to a tube or vessel, a combination of which constitutes the vascular tissue of authors. Whilst this is the nature of a true vessel, considerable confusion has arisen from the term being applied to any cell-like structure in which the longitudinal diameter exceeds the transverse; and thus the difference between a cell and a vessel became one of length only. The term vessel ought to be restricted to such as are formed by coalescence of cells.

Under the term pleurenchyma (fig. 8) is included tissue ' composed of such elongated prosenchymatous, flexible, ’ thickened cells, as are found in the bast or phloem layers of ordinary trees. They also occur in the wood portion. Their walls are thickened regularly, and they constitute when united what is commonly known as the woody or bast fibre. The diameter of the woody fibres varies from y7⅛ σ th to 2⅞(j th of an inch. The materials used for ropes and cordage, linen, certain Indian muslins, mummy-cloths, and mats consist of the woody fibre of plants from which the more delicate tissues have been removed by maceration in water. Flax or lint is thus procured from the bark of Linum usitatissimum, hemp from Cannabis sativa, New Zealand flax from Phormium tenax, Pita flax from Agave americana, Sun-hemp from Hibiscus cannabinus, and bass or bast from the common Lime or Linden-tree. Fibres are also procured for manufacture from the Pine-apple plant (Ananassa sativa), from Yucca gloι - iosa, from Büh-meria nivca, which yields the Indian Rheea fibre, from most of the plants belonging to the Mallow and Nettle tribes, and from some Leguminous plants, such as Crotalaria juncea, which supplies a kind of Bengal hemp. If the maceration of the fibre is carried to a great extent, a pulp is formed from which paper is manufactured. Pleurenchyma does not occur in cellular plants, such as Lichens, Sea-weeds, and Mushrooms. The tissues of these plants speedily disappear under the action of water, and hence, perhaps, the reason of their rarity in a fossil state. In the very young state woody cells are delicate, and it is only in proportion as they attain maturity that their walls acquire a thick consistence. In the sap-wood of ordinary trees the woody cells are thickened in their walls, but are pervious; while in the heart-wood they are rendered solid by the thickening matter, which is often variously coloured.

If the thickening of the cell-wall takes place so that a network, ring, or spiral of thickening matter is formed, then the cells are reticulated, annular (fig. 11), or spiral (fig. 12), as in the leaves of Sphagnum, hairs of Cactaceae, and seed-coat of Casuarina. In these cells the spiral thickening frequently becomes loosened from the cell-wall as a spiral fibre, and can be unrolled. Such forms occur in the outer covering of the seed of Collomia linearis and of the fruit of Salvia Verbenaca. In these, when placed in water, the spirals rupture the softened membrane of the cells, and spread outwards. The spongy elastic character of the outer cellular covering of the roots of tropical Orchids and Araceae, of the sepals of Illecebrum verticillatum, of the pericarp of Cachrys Morisoni and C. odontalgica, and of the ribs of the fruit of Aethusa Cynapium, is due to the presence of spiral cells. In the reproductive cells of Hepaticae spiral fibres called elaters are found in connection with the spores. Reticulated or netted cells, produced by fibres forming a sort of mesh or network, occur in the wing of the seed of Swietenia, in the pericarp of Picridium tingitanum and P. vulgare, in the seed-coat of Cucurbita Pepo, in the parenchyma of the leaf of Sanseviera guineensis, and in isolated cells of the pith of Rubus odoratus and of Erythrina Corallodendron.

If spiral, annular, or reticulated cells are arranged in a longitudinal series, and the septum between adjoining cells gives way, then we have a spiral, annular, or reticulated vessel (figs. 13, 15, 16) formed, and to this tissue authors have given the name trachenchyma, on account of its resemblance to the tracheae or air-tubes of animals. These vessels vary from 5 √ ιrσ th to 3 ⅜ σ th of an iuch in diameter. The spiral thickened portion of the wall of the vessel may become loosened from the membrane of the wall and form a spiral fibre in the interior. These fibres are elastic, usually rounded and simple; but sometimes two or more are combined so as to form a flat baud. These flat ribands, consisting of fibres which vary in number from two to twenty-five, or more, are met with abundantly in the stems of Bananas and Plantains, and in the shoots of Asparagus. The spiral in such cases is called compound, and the vessels pleiotracheoe. The spiral fibres have such tenacity that when the vessels are ruptured they can be pulled out. This capability of being unrolled characterizes true spiral vessels (fig. 13). When the spiral is not loosened from the cell-wall and cannot therefore be unrolled, it is said to be closed. On breaking the young shoots or leaf-stalks of the Geranium, Strawberry, and Rose, or the leaves of the Hyacinth, Amaryllis, and Banana, and pulling the parts gently asunder, the fibres can be easily seen in the form of· a fine cobweb. When the aerial stems of the Banana and Plantain are cut across, the spiral fibres may be pulled out in large quantity so as to be used for tinder. Generally, the coils or volutions of the fibre are said to be left-handed, that is, turning to the left of a person supposed to be in the axis. In the garden Lettuce vessels are met with, some having the fibre turning to the left, others to the right. In the Scarlet Bean the coils of the fibres are left-handed, while the plant itself turns to the right in twining. Spiral vessels are abundant in young plants and shoots, while in the hard stems of trees and shnιbs they chiefly surround the pith. Spiral vessels occasionally exhibit a branched appearance. This may arise from the union of separate vessels, or it may depend on a regular division of the fibres, as is seen in the Mistleto, House-leek, and Gourd (fig. 14). Annular vessels are those in which the thickening (or, if it be loosened from the wall of the vessel, the fibre) is in the form of rings (fig. 15). These rings in Mammillaria quadrispina, and in some other plants of the Cactus tribe, are very thick, and leave only a small canal in the centre of the vessel. Annular vessels are from - s ⅜ σ th to τ ⅜ σ th of an inch in diameter. In reticulated vessels (fig. 16) the thickenings take the form of a network. All vessels of this type lose very early their protoplasmic contents, and serve to convey air.

In the process of thickening of the cell-wall, if large spaces of the cell-wall remain thin, and the thickening mass growing in a circular manner projects into the interior of the cell and gradually arches over the thin portion of cell-wall, a dome-shaped cavity is enclosed betwixt the thin cell-wall and the thickening mass. The growing thickening mass gradually contracts the opening into this cavity, but never completely closes it. On front view this presents the appearance of two concentric circles, an outer marking the edge of the original thin portion of the cell-wall, and an inner indicating the under edge of the gradually contracting ring of thickening matter (fig. 17). When this process takes place on opposite sides of the partition wall between two cells, there are then two similar cavities separated by the thin partition wall of the cells, each communicating freely by a small circular aperture with the cell in which it has been formed (figs. 17 and 18). In process of growth the partition wall is absorbed, and then a lenticular cavity is formed, connected by a circular aperture on each side with adjacent cells (fig. 19). When viewed by transmitted light these present the appearance seen in fig. 17 t'; such structures are termed bordered pits, and a collection of such cells constitutes the disk-bearing or punctated tissue of authors. It is well seen in Firs and other conebearing plants. It has been called glandular tissue. In the case of some fossil woods, pieces of silica, like double convex lenses, have been found in the cavities. When a vertical radial section is made of the stem of Fir, bordered pits, arranged in two rows, with individual pits on the same level, are seen. In Araucaria double and triple alternating rows are seen; whilst in the Yew a prominent striation line winding spirally amongst the pits is noticeable. When the thickening begins by the formation of transverse ridges extending right across the wall of the cell, and the inwardly projected ridges gradually arch over the thin membranous portion of the wall, a narrow fissure only is left leading into the cavity enclosed by the thickening masses and the thin portion of the partition wall, on the opposite side of which a similar process lias proceeded. By the absorption of the partition wall a single cavity between two cells is thus produced, communicating with both, just as in the last case. Viewed by transmitted light these present an appearance like rungs of a ladder, and hence the name scalariforrn applied to the cells in which they occur (fig. 20). They are specially seen in Ferns, where they give rise to long prismatic vessels.

When the thickening takes place over nearly the whole of the cell-wall, thin portions may be left here and there, which appear as pits when viewed by transmitted light, or as small canals when the cell-wall is very thick. Such cells are termed porous or pitted or dotted cells (fig. 21). In old cells, after the protoplasm has disappeared, the portiou of the cell-wall which remained thin is often absorbed, and thus there is a true perforation of the cell-wall. These perforations often occur in groups both upon the cell-wall and upon the septum between superposed cells, and give rise to a remarkable sieve-like structure, in which case they are termed sieve cells The latticed cells of some authors are of a similar nature. When superposed porous or sieve cells coalesce by complete obliteration of the septum, then a pitted vessel, sieve tube, or duct is formed (fig. 22). These ducts are usually of a larger size than other ves sels; they are well seen in the inner phloem layers and in the wood of trees, and they constitute the large rounded openings which are seen in the transverse section of the stems of the Oak, Poplar, Willow, àc. They also abound in the Bamboo (fig. 23), and in other plants of rapid growth. The names of bothrenchyma and taplιrenchyma have been given to a tissue composed of such cells. Not unfrequently contractions are visible on the outside of the vessel (fig. 22), indicating its formation by coalescence of superposed cells. To vessels exhibiting contractions of this kind, whether spiral or pitted, the terms moniliform, and vermiform have been applied; and the tissue composed of these moniliform vessels has been denominated phleboidal. In the ducts of many plants a remarkable appearance is produced by the protrusion, through the perforations into the cavity of the vessel, of portions of the adjoining cells, or, before its absorption, of the portion of partition-wall closing the pit. These portions appear as cells filling the interior of the vessel, and are described under the name of tylosis (fig. 24). It is well seen in the Walnut, Chestnut, Oak, &c.

Latιciferous vessels (fig 25) consist of long branchbig tubes or passages, having a diameter of about 1 - i 1 0ι5 th of au inch, forming, by their union, an anastomosis or network, like the veins of animals. They are the milk vessels and the proper vessels of old authors. They receive their name from containing an emulsion called latex, of a granular nature, often milky or coloured. They are seen in the Indiarubber and Gutta-percha plants, the Mudar plant, the Cow-tree, Spurges, Dandelion, Lettuce, Chicory, and Celandine, frequently containing a large quantity of caoutchouc. Usually these vessels are thin-walled, but sometimes slightly thickened. They are found most abundantly in the phloem layers—rarely in the xylem or wood layers (Papayaceae). They are the result of the coalescence of anastomosing or rectilinear rows of cells, and sometimes they seem to have resulted from the conversion of other vessels. In some Araceae they seem to represent spiral vessels. In Asclepiadaceae they are evidently bast-fibres. Some consider them as merely intercellular canals. The milky sap of Euphorbia phosphorea is said to be luminous. The latex exhibits movements which have given origin to the name cinen-chyma applied to laticiferous tissue by some authors. Those movements, classed under the name cyclosis, must not be confounded w ith the motion of protoplasm in cells which is designated rotation.

We have seen that the cellular tissue is sometimes incomplete, that is, the cells do not touch on every side (fig. 9). The intervening spaces are called intercellular spaces, and these may be either circumscribed cavities called lacunoe, or they may extend for some length through the tissue as intercellular canals; but these two structures pass into one another. In the earliest stage of development the tissue is always complete, and these spaces are formed subsequently by a splitting of the partition or common wall of the cells, and they may subsequently be increased in size by an absorption of the investing cells. These lacunae and canals may contain air, especially iu aquatic plants, to give them buoyancy, as in Potamogeton (fig. 26), or they may be receptacles for various secretions, and when they exist as canals they usually aid in conducting sap. The intercellular canals are exceedingly well seen in coniferous plants, where they constitute resin passages, forming a continuous system throughout the plant, and arranged at intervals in concentric circles in the xylem or wood portion of the stem.

Chlorophyll is the green colouring-matter of plants. It occurs in the cells of the superficial parts of plants united with small portions of the protoplasm (chlorophyll bodies), which are combined into grains of various forms. Starch-grains are usually abundant iu the chlorophyll bodies. Chlorophyll is soluble in alcohol and ether. It consists of four substances, two yellow and two green, which possess distinct optical properties. It gives a black band in the red of the spectrum. Physiologically it is very important. It is developed under the action of light, and undergoes changes according to its state of oxygenation. Hence the varied tints of leaves in autumn. Numerous colouring matters occur in plants, especially in flowers, and all such when not green are included under the general term chromule. Starchy and oily matters and albuminoids occur very abundantly in the cells of plants, where they are stored for the purposes of nutrition. Starch, composed chemically of carbon and the elements of water (hydrogen and oxygen), its formula being C 6 H 10 O 3 , occurs in the form of fine grains, more or less oval or rounded, which vary in diameter from the τδ ⅛ σ th to the ^- σ th of an inch. Each grain contains starch in two forms,—one, receiving the name granulose, is easily soluble, and gives a deep blue colour on the addition of iodine; the other form, starch cellulose, is less soluble, and gives a yellow or brown colour with iodine; but the former is most abundant in the grain. The individual grains either lie distinct from each other iu the cells, as in the Potato, Wheat, and Fea, or they are aggregated so as to form compound grains, as in West Indian Arrowroot, obtained from Maranta arundinacea, and Portland Sago, procured from Arum maculatum. Grains of starch present a very characteristic concentrically striated appearance (fig. 27). This is the result of their mode of formation. The poiut round which the striae are arranged is called the hilum. Starch is accumulated in the internal, and ofteu in the subterranean parts of plants. It occurs abundantly in fleshy roots, and in stems, as well as in seeds and fruits, and is easily separated by washing. The ordinary cultivated grains yield starch in considerable quantity; so also do the Potato, Arrowroot, and Cassava plants, the Sago-palms, and Banana fruit. That procured from the Arrowroot plant consists of dull white grains; while that from the Potato, and from various species of Canna supply ing tous-les-mois, is in the form of large shining particles. Sago and Tapioca are granulated forms of starch, the former being procured from the cells of various species of Sagus and Metroxylon, the latter from the Cassava plant The existence of starch in the bark and young wood of trees, such as the Birch and Pine, renders them useful as articles of food in cold countries. Lichenin is a form of starch existing in the cells of Iceland moss and other lichens; while inulin, which occurs dissolved in the cell-sap, is the starchy matter supplied by the roots of the Dahlia, Dandelion, and Elecampane. By the action of prolonged heat, as well as by the addition of dilute sulphuric acid, or of malt, starch is converted into a soluble gummy substance called dextrin. The same change occurs during germination or the sprouting of the seed. Oily matters occur as drops in the interior of cells, usually associated with starchy substances or with albuminoids. These latter exist as small granules or large rounded masses with definite chemical and optical properties, and are termed aleurone grains. They are frequently associated with a crystalline arrangement of portions of the protoplasm of the cell known as crystalloids. Sugar occurs abundantly in the sap of plants. When pure and in a solid state this substance is crystalline and soluble in water; but it also occurs in an uncrystallized form. There are two marked varieties of it. Cane-sugar, C 12 H 4a O 11 , procured from the Sugar-cane, Sugar-maple, Beet, Carrot, &c.; and grape-sugar, occurring iu numerous fruits, as Grapes, Gooseberries, Currants, Peaches, and Apricots The formula for grape-sugar is C β H li O r During the sprouting of the seed, starch is converted into grape-sugar, and a similar change is induced by the action of malt, or of any ferment. A sweet substance (not a true sugar), called Mannite, is procured from the Manua-ash (Ornus europaea) as well as from various sea-weeds, from species of Eucalyptus, and from the Dandelion. Gum or mucilage'is another substance found in vegetable tissues. When pure it is clear, soluble in water and also in dilute acids, but not soluble in alcohol or ether. It is one or the torms through which vegetable matter passes in being applied to the purposes of plant life. It exists largely in the vegetable juices. From the bark of many trees it is procured in the form of an exudation. Two well-marked kinds of gum are met with,—arabin, soluble in cold water, constituting the chief ingredient of gum-arabic, procured from various species of Acacia; and cerasin, insoluble in cold water, but readily soluble in boiliug water, constituting the gummy secretion obtained from the Cherry and Plum. A substance called bassorin, or vegetable jelly, is found in Tragacanth, the roots of some Orchids, as well as in Carrageen (Chondrus crispus), and other sea-weeds. It is allied to gum, but differs iu swelling up and becoming gelatinous when mixed with water. Another gum-like substance called pectin exists in the juice of the Apple, Pear, and other pulpy fruits. It is changed by the action of alkalies into pectic acid, which is found in many fruits and such succulent roots as Carrot, Turnip, Beet, &c. Oils, Fats, and Resins occur in cells of plants, or in special canals or glands as products of assimilation. The oils are either fixed or volatile,—the former being divided into drying, fatty, and solid, while the latter are distinguished according as they consist of carbon and hydrogen alone, or of these elements combined with oxygen or with sulphur. Resinous matter occurs in the form either of fluid balsams, or of the various kinds of solid resin and pitch. In the rind of the Orange and Lemon glands of oil occur (fig. 28). Turpentine canals are met with in the wood of Pines; and Vittae, or oil-canals, in the fruit of Umbelliferous plants, such as the Coriander. In the fleshy covering of the fruit of the Olive there are numerous oil-cells. The fruit of the Guinea-palm yields a solid oil, called palm-oil. The dotted appearance of the leaves of the Orange, Myrtle, Eucalyptus, and St John’s Wort, depends on the presence of numerous cells or cavities containing essential oil.

As allied to these secretions, we may notice caoutchouc, which is found in the milky juice of plants, especially those belonging to the Fig, Spurge, and Dogbane orders. The trees most prolific in this substance are Siphonia elastica, Urceola elastica, and Urostigma elasticum. Guttapercha is the concrete milky juice of the Taban plant (Isonandra Gutta). Wax is also found in the tissues of plants, and it frequently occurs as a secretion on the stems, as in the Wax Palm, and on the surface of fruits, as in the bloom or glaucous secretion of the Plum and the Candleberry Myrtle. Crystals of lime salts occur in the interior of cells, and also in the cell-walls of plants. They consist of lime in combination with carbonic or oxalic acids, and are in many plants very abundant. In Monocotyledons they usually assume the form of needle-like crystals, and are termed raphides; they consist of oxalate of lime (fig. 29). They are remarkably well seen in the Banana. The Squill bulb and the bulb of the Onion exhibit raphidian cells, which are easily separated during the decay of the plants. The crystals are also arranged in a conglomerate form (fig. 30), as may be seen in the root of Turkey Rhubarb, to which they impart grittiness; and in Old-Man-Cactus they constitute 60 to 80 per cent, of the dried tissue. In ä siugle cell of the Poke (Phytolacca decandra) twenty to thirty crystals may be seen. In the epidermal cells of species of Ficus, and other allied plants, prolongations inward of the cell-wall occur, at the extremity of which small crystals of carbonate of lime are deposited (fig. 31); to these the name cystoliths has been applied. Siliceous matter occurs in the walls of cells, as in Grasses and Horsetails, and especially in Diatomaceae.

2. Integumentary System.

A more or less marked division of the tissues into an outer layer bounding an inner mass is visible in all plants. Amongst the lower cellular plants this division is not very distinct; the circumferential cells are perhaps only a little smaller and more compacted than those near the centre. The higher cellular plants, however, exhibit great diversity. In them the cells of the circumference may be arranged in layers so as to constitute a true epidermis, the component cells having a definite relation to one another and to the exterior in the respective families. In all vascular plants an epidermis is found. In many cases, however, it is difficult to recognize it, as in the stems of submerged plants and in most roots. It usually consists of a single thin layer of cells closely compacted, and leaving no interspaces except at definite points {stomata'), where openings lead into intercellular cavities. The cells composing this layer have their outer wall much thickened, the inner wall remaining thin, and they contain no starch or granular matter, and usually are colourless. In many aquatic plants, however, and in Ferns, chlorophyll is present. The apparent colour of the epidermis depends on that of the parenchymatous cells below, from which it can be separated as a colourless layer. Sometimes the cell-sap of the epidermal cells has a red tinge. The outer lamella of the outermost cell-wall of the epidermal cells usually becomes cuticular or corky, and thus is formed an external separable layer or cuticle upon the surface of the epidermis proper. This layer has different chemical properties from the epidermis, being insoluble in sulphuric acid. Upon this cuticle wax is frequently deposited in various forms, serving as a protective from moisture; of this nature is the bloom of the Plum. The cuticle in aquatic plants is very thin; in aerial plauts it is much thicker. The single layer of cells forming the epidermis is not unfrequently strengthened by the addition to the inside of other layers of cells. In the leaves of Begonia, Ficus, and the outer coveriug {velamen) of the aerial roots of Orchids, this constitutes the hypoderma, the cells being of various forms. It is well seen also in vascular Cryptogams, many Bromeliaceae, Ilex, &c. On those parts of the plant •which live long and have vigorous growth in thickness the efficacy of the epidermis as a protective covering is increased by a large formation of cork. Each epidermal cell divides into an outer and an inner cell. The former at once becomes a cork cell, losing all its succulent matter; the latter remains capable of division. When a layer of these merismatic cells occurs we have a cork cambium or phellogen. If several layers of cork cells be formed a cork tissue or periderm is the result, which supersedes the epidermis, and which from variations in the several layers may be stratified. Not unfrequently the phellogen cells, in addition to giving off cork cells outwardly, give rise on the inner face to cells containing chlorophyll; and if a layer of such is formed it is termed a phelloderm. In such cases the phellogen lies between the phelloderm and the periderm. If phellogen lamellae are formed deeper in the tissues of the plant, the internal layers of tissue become dry and constitute the bark. Periderm is thus replaced by bark. One important character of the epidermis is the presence of stomata or breathing-pores. These exist abundantly upon the stems and leaves of plants; they also occur on the parts of the flowers; but they are absent from all root structures, though present on underground axial structures. Each consists of a central pore bounded by two or more cells (guard-cells), which contain chlorophyll, starch, and matters distinct from the surrounding epidermal cells. The pore has various forms, and opens „

into an intercellular cavity ,

(fig. 32). It may be round Ae. '

(Primrose), oval (Liliaceae), quadrangular (Yucca). The arrangement of the stomata on the plant varies much, They may be in lines as in

Equisetum, or they may be lg '

scattered irregularly ns in .≡ Sff≡⅛⅛⅛

Balsam (fig. 33), Orin definite showing the intimate union of the cpi- „„ ,∙ λ ..j dermal cells «, e, the loose subjacent par clusters as in Crassula and encħymap, with intercellular canals m, Saxifraga (fig. 34). In and lacuna l >∙ », stoma.

Equisetum the stomata, which are about 0⅜ σ th of an inch in their greatest diameter, consist of four guard-cells; two of which are arched and thick at their outer convex margin, becoming thin at their inner concave edge, where two other smaller cells overlie them, with their upper walls raised as ridges running at right angles to the stoma. This gives the stoma a fringed appearance, hence called pectinate. The mode of formation of a stoma is an interesting process, resulting from the division of a single epidermal cell. Preparatory to the complete formation of the mother-cell of the stomatic guard-cells, divisions may take place in the epidermal cells by which numerous or few cells are formed surrounding this mother-cell. In the mother-cell a partition wall is developed (fig. 35, s). A thickening of the partition wall occurs, and eventually in the central portion a fissure makes its appearance, which gradually increases in size (fig. 36). It does not, however, pass throughout the whole extent of the partition wall. An opening or stoma is thus formed, bounded on each side by a single guard-cell, leading into an intercellular space in the parenchyma beneath. If we suppose the mother-cell to be divided by numerous vertical septa we should have a stoma surrounded by many guardcells, as occurs in some Hepaticae; and in this family, after the formation of the fissure a series of horizontal septa appear in the guard-cell, and thus the stomatic opening becomes a canal. In Ceratopteris thalictroides the stoma is bounded by three cells,—two of which, in their open condition, are crescentic and concave inwardly, while the third surrounds them, except for a small space at the end of the long axis of the stoma, and has on this account been called peristomatic. In Urostigma elasticum four cells form the stoma. Subsequent changes in the surrounding epidermal cells may cause alterations in their relations to the stomata. Thus, in the Oleander (fig. 37) the epidermis develops so as to sink the stoma within a ciliated cavity; in other cases the stomata are raised on pillars. The guardcells have a power of opening or closing the orifice of the stoma; when moistened they swell and curve outward, and leave a free opening; when dry they collapse and their edges are straightened, and they close the orifice. Stomata are most abundant on the under surface of leaves. Their number varies from 200 to 160,000 or more in a square inch of surface. In the leaves of the White Lily there are 60,000 in a square inch on the under surface, and about 3000 on the upper; on the leaves of the Cherry-Laurel there are 90,000 on the lower surface, and none on the upper.

Upon the epidermis and as outgrowths from it are found certain appendages. These are hairs, scales, prickles, &c., and all have been embraced under the general name trichome. Hairs (pili or villi') are productions each of an epidermal cell, having typically an elongated or conical form, and covered by the cuticle. They are erect or oblique, or they lie parallel to the surface, and are appressed. Sometimes they are formed of a single cell, which is simple and undivided (fig. 38), or forked (fig. 39), or branched; at other times they are composed of many cells, either placed end to end, as in moniliform or necklace-like hairs (fig· 40), or united together laterally, and gradually forming a cone, as in compound or branched hairs. When the branches of a hair proceed from a common centre, it is stellate or radiated (fig. 41, a, δ), as in hairs of the Mallow tribe, in those of Deutzia scabra, and those on the stem of the Rice-paper plant (Fatsia papyrifera). When stellate hairs are flattened out, so as to form a sort of membranous expansion (fig. 44), a scale or scurf is produced. In Bromeliaceae the s curliness of the leaves is a marked character. To such expansions of the epidermis the name lepis is applied, and the surface is said to be lepidote. These scales have sometimes a beautiful silvery appearance, as in Elaeagnus and Sea-buckthorn (fig. 44). Surrounding the base of the leaves of Ferns a brown chaffy substance occurs, consisting of elongated cells, to which the name of ramentaceous hairs, or ramenta, has been given. In Palms also a similar substance but of a fibrous texture occurs, called reticulum or mattulla. Setae are bristles or stiff hairs, and the surfaces on which they occur are said to be setose or setaceous. Some hairs, as those of Drosera

(Sundew), have one or more spiral fibres in their interior. When the cells of hairs are hardened by thickening of the cell-wall, as in the Rose and Bramble, they are called Prickles (aculei). By some these are not considered as hairs, but are termed emergencies, inasmuch as they arise from a collection of cells, not from one epidermal cell. Various names have been given to the different forms of hairs. They are clavate, or club-shaped, gradually expanding from the base to their apex; capitate (fig. 42), having a distinct rounded head; rough or scabrous, with slight projections on their surface; lvoohed or uncinate, with a hook at their apex pointing downwards and to one side; barbed or glochidiate, with two or more hooks around the apex; shield-like or peltate, when attached by their middle, and projecting horizontally on either side, as in Malpighia urens (fig. 45), and in many Cruciferous plants; cilia, when surrounding the margin of leaves. On the pod of the Cowitch (Mucuna pruriens) hairs are produced with projections on their surface, which cause irritation of the skin. In Venus’s Fly-trap (Dionaea muscipula) stiff hairs exist on the blades of the leaf (fig. 46), which, when touched, induce their closure.

Hairs occur on various parts of plants,—on the stem, leaves, flowers, seed-vessels, and seeds, and even in the interior of vessels. In the interior of the spathe of some Palms numerous ovate cells, analogous with hairs, occur in clusters, and can, when the spathe is dried, be shaken out in the form of powder. Cotton consists of the hairs surrounding the seeds of Gossypium herbaceum and other species of the genus. These when fresh are elongated tubular cells; when dried their walls collapse and they appear twisted. Hairs are occasionally developed to a great extent on plants exposed to elevated temperatures, as well as on those growing at high altitudes. When they occur on the organs of reproduction they are connected with fertilization, as the hairs on the style of Goldfussia, and the retractile hairs on the style of Campanula.

Different organa of plants are transformed into hairs,—as may be seen in the flowering stalks of the Wig-tree (Rhus Cotinus), and in the calyx of Compositae.

Names are given to the surfaces of plants according to the presence or absence of hairs, as well as the nature of the hairs which cover them The following are the more important terms ·— Glabrous, smooth, having no hairs; hairy or pilose, furnished with hairs ; pubescent, covered with soft, short, downy hairs; villous, having long, weak, often oblique hairs: sericeous, covered with long, closely oppressed hairs, having a silky lustre; hispid, covered with long harsh or stiff hairs not appressed , hirsute, having long tolerably distinct hairs, not harsh nor appressed; velvety or velutinous, with a dense covering of short down, like velvet; tomentose, covered with crisp, rather rigid, entangled hairs like cotton, which form a sort of felt (tomentum); woolly, with long curled and matted hairs like wool; bearded or stupose, when hairs occur in small tufts. The hairs which are most frequently met with in plants are called lymphatic, from their not being connected with any peculiar secretion. Those, on the other hand, which have secreting cells at their base (fig. 43) or apex, are denominated glandular, and are not to be distinguished from glands. On young roots cellular projections occur, which may be called radical hairs. Young leaves and buds are frequently covered with protecting hairs. On the parts of the flower, as in the Iris, coloured hairs occur which have been called corolline

In connection with the epidermal appendages we may notice glands, although they may occur in any tissue. A gland consists of a single cell or a collection of cells secreting substances different from those contained in the surrounding cells, In the former case the gland is simple, in the latter it is compound. In compound glands it frequently happens that the walls of the inner cells are absorbed, and thus the gland has only a single cavity, as in the glands of the Orange rind (fig. 28); these are termed vesicular. The secretion of the glands may be stored in their interior, as in Orange rind, and in the leaves of Lauras Camphora, or it may be exuded as in Lychnis viscaria, and in the nectaries of Fritillaria imperialis (fig. 50). Hairs serve as ducts through which the secretion of glands is discharged. Such hairs are seen in the Nettle (fig. 43), in Loasa or Chili Nettle, and in Malpighia (fig. 45), and are commonly called stings (stimuli). In the Nettle they are formed of a single conical cell, dilated at its base, and closed at first at the apex by a small globular disk placed obliquely. This disk breaks off on the slightest touch, when the sharp extremity of the hair enters the skin, and pours into the wound the irritating fluid which has been pressed out from the elastic epidermal cells at the base. When a nettle is grasped with violence, the sting is crushed, and hence no injury is done to the skin. The glandular enlarged apex of hairs sometimes exudes a viscid secretion, as in the Chinese Primrose and in the Sundew. The hairs of the latter plant, by this secretion, detain insects which happen to alight on them. The hairs gradually close on the insects, electrical phenomena taking place during the movement, and then the secretion from the glands becoming acid, an action takes place upon the organic matter analogous to that of the pepsin of the gastric juice, by which it is rendered soluble, and is eventually absorbed by the plant. A similar property is possessed by the secretion from the glands upon the surface of the leaf of Venus’s Fly-trap (Dionaea muscipula, fig. 46). The acid in both instances belongs to the formic acid series. Glands may be either internal or external, and they may be situated at the extremity of a hair (when they are stalked), or they may be immersed in the substance of the plant (sessile). In the Dittany (Dictamnus albus, fig. 47) a form of gland is seen intermediate between the sessile and stalked form. The glands in this plant secrete a green oily matter, so also do the stalked glands in the Rose. In the Ice-plant the glands appear as elevations of the epidermis, containing a transparent fluid like ice, which is said to have an alkaline reaction; in the Chick-pea similar superficial cells contain an acid fluid. Clear glands are also seen on the under surface of the leaf of Passiflora lunata. Resinous glands are seen in the Hop (fig. 48) and Hemp plants. At the base of the petals of the Crown-imperial (fig. 50), cavities occur containing a honeylike fluid, secreted by what are called nectariferous glands.

Cavities containing saccharine matter, surrounded by small thin-walled cells, are met with in the leaves of Acacia longifolia, also in Viburnum Tinus, and Clerodendron fragrans. The cavities communicate with the surface of the leaves by means of canals. Peculiar sessile glands secreting a gummy substance are found at the inner side of the base of the petioles of Cinchona and Ipecacuan plants (fig. 49). On the buds of various trees peculiar glandular hairs termed colleters exist, secreting a gummy mucilaginous matter, the blastocolla, which covers the bud. These, however, disappear on the bursting of the bud. They seem to have a protective function. The secretions of glands are very various, oily, waxy, resinous, gummy, saccharine, acid, &c.

ORGANS OF PLANTS.

Having now considered the elementary structures and tissues found iu the Vegetable Kingdom, we proceed to view them in combination to form the plant. The simplest plant is found amongst Algae, where, as in the Red-snow plant (Protococcus nivalis, fig. 51) the whole organism consists of a single isolated cell. Other Algae and all Fungi and Musci are composed of a number of cells united in various ways; whilst in Ferns and their allies and all flowering plants vessels are formed in addition to the cells. The plants in which the tissues are entirely cellular are termed cellular plants; those in which vessels are also found are vascular plants.

That the portions of a plant may be properly maintained two functions have to be performed, namely, nutrition, on the proper performance of which the life of the individual plant depends, and reproduction, by which the perpetuation of the type is provided for. In such a simple form as the Red-snow plant (fig. 51) those functions are performed by the single cell. In the plants composed of numerous cells a differentiation takes place by which special cells are set apart for particular functions, and thus certain organs are formed in the plant. In the higher plants those organs become more complicated from the introduction of the vascular element.

The nutritive organs of plants are generally known as the root, the stem, and the leaves. In all vascular plants and the higher cellular plants an axis or stem having roots and bearing leaves is distinguishable, and such plants have been designated Cormophytes or Phyllophytes. In the lower class of cellular plants, as Fungi and Algae, no such distinction is possible, and there is merely a flattened leafy expansion with dependent filiform processes; this structure has been termed a thallus, and such plants are Thallogens or Thallophytes.

Amongst the higher plants the reproductive organs are, in ordinary language, comprehended under the term flower; and, as they are conspicuous, such plants have been denominated Flowering, Phanerogamous, or Phoenogamoiis. Amongst all cellular plants and in some vascular plants, as Ferns and Equisetum, there are no flowers, and the reproductive organs are inconspicuous, hence they have been termed Flowerless or Cτyptogamous. In all cases the young plant, or embryo, is completely cellular. But as growth proceeds, that differentiation takes place which distinguishes the several classes of plants one from the other. In Phanerogams the first leaves produced upon the embryo plant are termed primary, seed-lobes, or cotyledons. In some cases these are two in number, and are opposite one another. Plants in which this occurs are Dicotyledonous (fig. 52), as our ordinary forest trees. In other plants the lobes alternate and only one cotyledon is formed; such are Monocotyledonous (fig. 53), as Grasses, Lilies. In Cryptogams, on the other hand, no such seed-lobes or cotyledons are produced, and they are Acotyledonous (fig. 54).

In all plants the original cell tissue which gives origin to its parts is of a uniform nature, and is termed the primary tissue. When all the cells of this tissue are capable of multiplication and division the tissue is a meristem or generating tissue. If the cells are not so capable, then it is a permanent tissue. The primary tissue at the growing point of any shoot or root is essentially a meristem, and it has been designated the primary meristem to distinguish it from the secondary meristem, which is applied to a tissue in the older parts of a stem or root which remains or becomes capable of division. The growing point of the apex has been termed the punctum vegetationis, and it not unfrequently forms a conical projection, and is then the vegetative cone. By growth at this punctum vegetationis the shoot or root increases in length, and the mode of addition is in many cases of a definite character. Two chief types of growth are recognized. In one of these a single large cell is always present at the apex, termed the apical cell, which may be regarded as the mother-cell, whence by bipartition in a definite manner the whole meristem below it has arisen, as is well seen in vascular Cryptogams and cellular plants. The other type is seen in Phanerogams, and here no such apical cell is visible, but a number of cells are found at the apex by whose multiplication the subsequent tissues are formed. But in whatever way formed, a primary meristem is the result of all the processes of growth, and by differentiation of its cells the various parts of the shoot or root are formed. The outerlayer of the primary meristem, which extends completely over the punctum vegetationis, is termed the dermatogen; it is the primordial epidermis, being continuous with the epidermis of the shoot and afterwards becoming epidermis. Underneath the dermatogen several layers of cells are distinguished, continuous with the cortical portion of the shoot or root; this is the primordial cortex, and constitutes the periblem. Enclosed by this is a central cellular mass, out of which the fibro-vascular bundles and the structures of the central part of the shoot or root are formed; this has been termed the plerome. If the growing axis be a young root there is in addition developed, usually from the dermatogen, a mass of cells at the extremity, constituting a root-cap, or protective covering, to the delicate merismatic cells beneath; no such structure is formed in a stem. Thus a stem is structurally distinct from a root in having no root-cap.

In the plerome the fibro-vascular bundles are formed. Certain cells become elongated and prosenchymatous and united in bundles leaving no intercellular spaces; this mass is termed the procambium of the fibro-vascular bundle. As growth proceeds changes take place in the cells by thickening, and, their contents disappearing, various kinds of cells and vessels are formed. In this way the whole mass of the procambium may be converted into permanent tissue, and then growth ceases; or an inner portion of the bundle remains merismatic, which is called the cambium, and then growth proceeds, the shoot or root increasing in thickness by the cambium forming new cells on both sides In the former case the bundle is closed, as in Cryptogams and Monocotyledons and some Dicotyledons; in the latter it is open, as in most Dicotyledons and Conifers. In every fibro-vascular bundle a separation into two groups of structures may be distinguished—the wood or xylem layers, and the bast or phloem layers. As long as the bundle is open and cambium present, these layers are separated by the cambium (fig. 55). Their relative position as regards the axis of the stem or root varies. In some cases the phloem is nearer the circum ference, in other instances the xylem is peripheral, and in rare cases amongst Dicotyledons there are phloem layers both on the outer and inner sides of the fibro-vascular bundles. In vascular Cryptogams the phloem layers always surround the xylem portion of the bundle.

In the xylem or wood portion of a fibro-vascular bundle the cells all tend to thicken their walls, and, consequently, numerous kinds of cells and vessels are found, which arc usually arranged in a definite manner. On the side furthest removed from the phloem we find spiral vessels interspersed with wood-cells (figs. 55 and 56), outside are reticulated and scalariform vessels, and then interspersed with wood cells are large pitted vessels. In the phloem or bast layers the cells have not such a tendency to thicken, but usually remain thin-walled, forming ordinary parenchymatous cells, or becoming perforated and forming sieve cells. In certain layers, however, the'cell-walls are thickened so as to become flexible, constituting the bast fibres. Around every fibro-vascular bundle a single layer of cells of the fundamental cellular tissue of the stem is marked off from its surroundings, the cells get filled with starch-grains, and this constitutes what has been termed the bundle sheath or starch-bearing layer (fig. 57). In fig. 57 is seen a transverse section of the closed fibro-vascular bundle from the Maize, and it will be observed that it is essentially the same as that of the open bundle (fig. 55), only that all the cambium cells have passed into per manent tissue.

In all plants a provision is made for branching of the various organs, and two principal forms of it may be recognized. In one of these the generating axis elongates at the apex, producing in succession lateral structures. To this form the term monopodium has been applied. In the second form there is a cessation of growth at the apex in the direction of previous elongation of the axis, and a continuance in two diverging directions. This is dichotomous branching. In their rudimentary state all branchings may be easily referred to one or other of those types; but in the mature system it is frequently difficult to recognize the type, owing to irregular development of the successive branching. Thus in a dichotomous branching only one of the secondary axes may develop strongly, the weaker branch appearing as a small lateral shoot from its base; and an apparent primary shoot is thus produced which in reality consists of the bases of single branches of consecutive forkings. Such an axis is termed a pseudaxis or sympodium. And, again, in monopodial branchings the primary axis may continue to develop more strongly than its lateral axes, which in their turn develop similarly, and a racemose form arises; or the primary axis may be arrested in growth, and the secondary axes develop more strongly and overlap it, when a cymose branching results. More will be said on this subject when considering the inflorescence of Phanerogams, in which the various forms of branching are well seen.

We now proceed to consider the form and internal structure of the various organs of plants.

I. ORGANS OF NUTRITION.

1. Hoot or Descending Axis.

Speaking generally, the root is that portion of the plant which descends into the soil. In all plants the root is at first entirely cellular. It may remain permanently so, or vessels may be formed in it. The Radicle, or young root (fig. 58, r), is the first portion of the embryo protruded from the seed or spore (fig. 54) when germination commences, and resembles very much in structure the young stem. Both are entirely cellular, consisting of a central nucleus of cellular tissue covered by two or more layers of cells. But at the apex of the root a mass of cells is developed, which constitutes what is known as the rootcap or pileorhisa. These cells extend for some distance along the sides of the root, forming a sheath, and in some cases, as Lemna, the cap becomes loosened from the root, remaining attached by a few cells at the apex only, and then it is known as the ampulla. This root-cap distinguishes structurally the root from the stem, and it serves as a protection to the apical growing-point of the root. The roots of Thallophytes, consisting entirely of cells, do not develop a root-cap.

The root is merely a prolongation downwards of the stem, and the part where they unite is the collum or neck. Afterwards the root is distinguished from the stem by the absence of a provision for the development of leaf-buds. It is not always easy to distinguish between a stem and a root. Many so-called roots bear at their upper part a portion called their crown, whence leaf-buds arise. Underground stems and roots are often confounded. Some plants, as the Moutan Paeony, the Plum-tree, Pyrus japonica, and especially Anemone japonica, have a power of forming buds on what are commonly called their roots. The last-mentioned plant develops these buds on every part of its extensively ramifying root-like prolongations, which may be chopped into numerous pieces, each capable of giving rise to a new plant. Such is also the case with the annulated root of Ipecacuan. Roots are usually subterranean and colourless. Externally, they have a cellular epidermal covering of a delicate texture, sometimes called epιblema, in which no stomata exist. In woody plants fibro-vascular bundles are found in the roots, and there is an internal arrangement of tissues similar to that seen in the stem itself, but spiral vessels are rare in the root. The axis of the root gives off branches which divide into radicles or fibrils, the extremities of which, composed of delicate cellular tissue constituting the punctum vegetationis, have been erroneously called sρongioles or spongelets; they are not distinct organs. Hairs are often seen on roots, but no true leaves. These hairs consist of simple elongated cells, which occur singly, and appear to serve the purpose of absorption. Roots increase principally by additions to their extremities, which are constantly renewed, so that the minute fibrils serve only a temporary purpose, and represent deciduous leaves; but in large trees which form thick roots an increase in diameter occurs in the root similar to what is seen in the stem itself. In some plants no roots are formed at all; thus in the Orchidaceous plants Epipogium Gmelini and Corallorhiza ihnata, and also in Lemna arrhiza, no roots occur. Roots may be given off from any portion of a stem, originating as cellular prolongations from the inner portion of the stem, and coming off at any point of the stem, or at small lenticular points to which the name of lenticels has been given by some. When the stem is more or less horizontal the roots given off from it pass directly into the soil; but if the stem be erect they pass for a certain distance downwards through the air, and are called aerial. The latter are well seen in the Screwpine (Pandanus), the Banyan (Ficus indica, fig. 59), and many other species of Ficus, where they assist in supporting the stem and branches. In the Mangrove they often form the entire support of the stem, which has decayed at its lower part. In Tree-ferns they form a dense coating around, and completely concealing, the stem; such is also the case in some Dracaenas and Palms. In Epiphytes, or plants growing in the air, attached to the trunks of trees, such as Orchids of warm climates, the aerial roots produced do not reach the soil; they continue always aerial and greenish, and they possess stomata. Delicate hairs are often seen on these epiphytal roots, as well as a peculiar investment formed by the cells of the epidermis which have lost their succulent contents and are now filled with air. This layer is called velamen radicum, or covering of the roots. The aerial roots of the Ivy are not the nutritive roots of the plant, but are only intended for mechanical support.

Parasitic plants, as the Mistleto (Viscum), Broom-rape (Orobanche), and Rafflesia, send root-like processes into the substance of the plants whence they derive nourishment. In the Dodder (Cuscuta∖ the tissue around the roots swells into a kind of sucker (haustorium), which is applied flat upon the other plant, and ultimately becomes concave, so as to attach the plant by a vacuum. From the bottom of the sucker the root protrudes, which penetrates the supporting body. In the case of parasitic Fungi, such as Mould, there are cellular filaments which spread among the tissues of plants, and which may be looked upon as equivalent to roots and stems united. They form the spaιυn or mycelium of these plants, and in some cases cause rapid destruction of the tissues of plants, as in the disease called Dry-rot.

The forms of roots depend upon the mode in which the axis descends and branches. The mode of branching of roots is almost universally monopodial, only in Lycopodiaceae is it dichotomous. When the central axis goes deep into the ground in a tapering manner, without dividing, a iαp- root is produced. This kind of root is sometimes shortened, and becomes succulent, forming the conical root of Carrot, or the fusiform or spindle-shaped root of Radish, or the najoiform root of Turnip. In ordinary forest trees the first root protruded continues to elongate and forms a long primary root-axis, whence secondary axes come off. In other plants, especially Monocotyledons, the primary axis soon dies and the secondary axes take its place. When the descending axis is very short, and at once divides into thin, nearly equal, fibrils, the root iβ called fibrous, as in many Grasses (fig. 60); when the fibrils are thick and succulent the root is fasciculated, aβ in Ranunculus Ficaria, Asphodelus luteus, and Oenanthe crocata; when some of the fibrils arc developed in the form of tubercules, the root is tubercular, as in Orchis (fig. 61); when the fibrils enlarge in certain parts only, the root is nodulose, as in Spiraea Filipendula, or moniliform, as in Pelargonium triste, or annulated, as in Ipecacuan. Some of these so-called roots are formed of a stem and root combined, and when cut in pieces they give rise to buds and new plants. In some cultivated plants, as Turnip, the central root is sometimes injured, so as to end abruptly, and it then divides into numerous branches, resembling a fasciculated root. This gives rise to the disease called Fingers and Toes, which is very injurious to the çrop. Anbury is a disease where a clubbing of the root takes place. The mode in which the fibres of roots are produced and developed gives origin to different forms of rhizotaxis, or root-arrangement.

Roots either fix the plant in the soil or attach it to other bodies. They absorb nourishment by a process of imbibition or endosmose through their cellular extremities. The elongation of the roots by their extremities enables them to accommodate themselvee to the soil, and allows the extremities of the rootlets to extend deeply without being injured. Roots, in their lateral extension, bear usually a relation to the horizontal spreading of the branches, so as to fix the plant firmly, and to allow fluid nutritive substances to reach the absorbing extremities. As has been already stated, the structure of perennial roots is identical with that of the stem. Thus in Dicotyledons we find a pith, medullary rays, zones of wood, cambium layers, and bark, although no medullary sheath is present. In Monocotyledons we have fibro-vascular bundles distributed in a matrix of cellular tissue. The young primary root in Monocotyledons differs from that in Dicotyledons in that it rises deeply within the embryonal tissue, and on germination this tissue is ruptured and forms a sheath, around the base of the roots, called coleorhiza. Amongst Monocotyledons the primary roots usually soon die, and secondary roots are formed in abundance. In vascular Dicotyledonous plants the structure of the root is similar to that of the stem. In Thallogens the roots consist merely of simple or branching filamentous hair-like structures. In some large tropical Seaweeds the root-like bodies develop to a large extent, but serve only as fixing organs, and take no share in nourishing the plant.

2. Stem or Ascending Axis.

A stem may be defined as an axis bearing leaves. Structurally it differs from a root in having no development of cells forming a cap over the growing point. Under the term caulome (stem structure) are included all those parts of a plant morphologically equivalent in bearing leaves. The stem generally ascends, seeking air and light, and has therefore been termed the ascending axis. Stems have usually considerable firmness and solidity, but sometimes they are weak, and either lie prostrate on the ground, thus becoming procumbent, or climb on plants and rocks by means of rootlets, like the Ivy, being then called scandent, or twist round other plants in a spiral manner like Woodbine, when they are volubile. Twining plants turn either from right to left, as the French Bean, Convolvulus, Passionflower, Dodder, Periploca, and Gourd; or from left to right as Honeysuckle, Twining Polygonum, Hop, and Tamus. Bryony tendrils twine from right to left, and left to right, alternately. In warm climates twining plants (lianas') often form thick woody stems; while in temperate regions they are generally herbaceous. Exceptions, however, occur in tho case of the Clematis, Honeysuckle, and Vine; the twining stem of the vine has been called sarmentum. Some stems are developed more in diameter than in height, and present a peculiar shortened and thickened aspect, as Testudinaria or Tortoise-plant, Cyclamen, Melocactus, Echinocactus, and other Cactaceae; while in many Orchids (fig. 62) the stem assumes an oval or rounded form, and is called a pseudo-bulb.

Names are given to plants, according to the nature and duration of their stems. Herbs, or herbaceous plants, have stems which die down annually. In some of them the whole plant perishes after flowering; in others, the lower part of the stem forming the crown of the root remains, bearing buds from which the stem arises next season. In what are called biennial herbs, the whole plant perishes after two years, while in perennial herbs the crown is capable of producing stems for many years, or new annual products are repeatedly added many times, if not indefinitely, to the old stems. The short permanent stem of herbaceous plants is covered partially or completely by the soil, so as to protect the buds. Plants producing permanent woody stems are called trees and shrubs. The latter are less than five times the height of a man, and produce branches from or near the ground; while the former have conspicuous trunks, which attain at least five times the height of a man. Shrubby plants of small stature are called tinder-shrubs or bushes. The limits between these different kinds of stem are not always well defined; and there are some plants occupying an intermediate position between shrubs and trees, to which the name of arborescent shrubs is occasionally given. The stem receives the name of caulis in ordinary herbaceous plants which do not form a woody stem, culm in grasses, truncus in trees, caudex or stock in Palms and in some Cacti, and stipe in Ferns. The term haulm is probably a corruption of culm; it is used by farmers to designate the stem of grasses and the herbaceous stems of plants. The stem is not always conspicuous. Plants with a distinct stem are called caulescent; those in which it is inconspicuous are acaulescent, as the Primrose, Cowslip, Gentian, and Dandelion. A similar term is given in ordinary language to plants whose stems are buried in the soil, such as Cyclamen or Sowbread. Some plants are truly stemless, and consist only of expansions of cellular tissue representing stem and leaf, called a thallus, and hence are denominated Thallogens, or Thallophytes.

Stems have a provision for a symmetrical arrangement of leaves and branches,— nodes, or points whence leaf-buds are produced, being placed at regular intervals. No such provision occurs in roots. The intervals between nodes are called internodes. The stem, although it has a tendency to rise upwards when first developed, in many instances becomes prostrate, and either lies along the ground partially covered by the soil, or runs completely underneath its surface, giving off roots from one side and buds from the other. Some stems are therefore subterranean, and are distinguished from roots by the provision made for regular leaf-buds. The first rudiment of the young stem in the embryo appears outside the seed or spore after the radicle has been protruded. It is termed the plumule (fig. 58), and differs from the radicle in the absence of a root-cap and in its tendency to ascend. The apical growing portion of the young stem constitutes the terminal bud of the plant, and by its development the stem increases in height; but in addition there is a provision for the production of lateral buds, which develop into lateral shoots more or less resembling the parent stem, and by these the branching of the plant is determined. These buds are found in the axil of previously-formed leaves; or, in other words, in the angle formed between the stem and leaf. They are hence called axillary. They are produced always from the outer portion of the stem except in the case of Equisetaceae (Horsetails), where they have a deep-seated origin. At first they consist entirely of cellular tissue, but in the progress of growth vascular bundles are formed iu them continuous with those of the stem, and ultimately branches are produced, which in every respect resemble the axis whence the buds first sprang. As the axis of the bud increases in length, cellular projections appear at regular intervals upon the primary meristem, which are the rudimentary leaves.

Buds, as has been stated, are either terminal or lateral. By the production of the former, stems increase in length, while the latter give rise to branches (rami), from which others, called branchlets or twigs (ramuli), arise, and add to the diameter of the stem. The terminal bud, after producing leaves, sometimes dies at the end of one season, and the whole plant, as in annuals, perishes; or part of the axis is persistent, and remains for two or more years, each of the leaves before its decay producing a bud in its axil. This bud continues the growth in spring. In ordinary trees, in which there is provision made for the formation of numerous lateral buds, any injury done to a few branches is easily repaired; but in Palms, which only form terminal buds, and have no provision for a lateral formation of them, an injury inflicted on the terminal bud is more likely to have a prejudicial effect on the future life of the plant. In the trees of temperate and cold climates the buds which are developed during one season lie dormant during the winter, ready to burst out under the genial warmth of spring. They are generally protected by external modified leaves in the form of scales (tegmenta or peruloe), which frequently exhibit a firmer and coarser texture than the leaves themselves. They serve a temporary purpose, and usually fall off sooner or later, after the leaves are expanded. The bud is often protected by a coating of resinous matter, as in the Horse-chestnut and Balsam Poplar, or by a thick downy covering, as in the Willow. Linnaeus called leaf-buds hibernacula, or the winter quarters of the young branch. In some plants, as in Platanus, the buds destined to live through the winter are so completely surrounded by the base of the petiole as not to be visible until the leaf has fallen off. These are said to be ιntrapetiolar.

In the bud of a common tree, as the Sycamore (fig. 63), there is seen the cicatrix left by the leaf of the previous year c, with the pulvinus or swelling p, then the scales e > e, arranged alternately in a spiral, and overlying each other in what is called an imbricated manner. On making a transverse section of the bud (fig. 64), the overlying scales e, e, e, e, are distinctly seen surrounding the leaves f, which are plaited or folded round the axis or growing-point. Iu plants of warm climates the buds are often formed by the ordinary leaves without any protecting appendages; such buds are called naked. A bud may be removed in a young state from one plant and grafted upon another by the process of bzιdding, so as to continue to form its different parts; and it may even be made to grow in the soil, in some instances, immediately after removal. In some trees of warm climates, as Cycas, Papaw-tree, Palms, and Tree-ferns, growth by terminal buds is well seen. In these plants the elongation of the stem is generally regular and uniform, so that the age of the plant may be estimated by its height; owing to this mode of growth they do not attain a great diameter. Although provision is made for the regular formation of buds, there are often great irregularities iu consequence of many being abortive, or remaining in a dormant state. Such buds are called latent, and are capable of being developed in cases where the terminal bud, or any of the branches, have been injured or destroyed. In some instances, as in Firs, the latent buds follow a regular system of alternation; and in plants with opposite leaves, it frequently happens that the bud in the axil of one of the leaves only is developed, and the different buds so produced are situated alternately on opposite sides of the stem. Occasionally, after a partial development as branches, buds are arrested and form knots or nodules. The embryo-buds or nodules of the Beech, Cedar, and Olive are apparently of this nature.

When the terminal bud is injured or arrested in its growth the elongation of the main axis stops, and the lateral branches often acquire increased activity. By continually cutting off the terminal buds, a woody plant is made to assume a bushy appearance, and thus pollard trees are produced. Pruning has the effect of checking the growth of terminal buds, and of causing lateral ones to push forth. The peculiar bird-nest appearance, often presented by the branches of the common Birch, depends on an arrestment in the terminal buds, a shortening of the internodes, and a consequent clustering or fasciculation of the twigs. In some plants there is a natural arrestment of the main axis after a certain time, giving rise to peculiar shortened stems. Thus the crown of the root is a stem of this nature, forming buds and roots. Such is also the case in the stem of Cyclamen, Testudinaria Elephantipes, and in the tuber of the Potato. The production of lateral in place of terminal buds sometimes gives the stem a remarkable zigzag aspect. Branches are sometimes arrested in their progress at an early stage of their development, and do not appear beyond the surface of the stem; at other times, after having grown to a considerable· size, they undergo decay. In both instances the lower part of the branch becomes embedded and hardened among the woody layers of the stem.

The mode in which branches come off from the stems gives rise to various forms of trees, as pyramidal, spreading, or weeping,—the angles being more or less acute or obtuse. In the Italian Poplar and Cypress the branches are erect, forming acute angles with the upper part of the stem; in the Oak and Cedar they are spreading or patent, forming nearly a right angle; in the Weeping Ash and Elm they come off at an obtuse angle; while in the Weeping Willow and Birch they are pendulous from their flexibility. The comparative length of the upper and under branches also gives rise to differences in the contour of trees, as seen in the conical form of Spruce, and the umbrella-like form of the Italian Pine (Pinus Pinea). The branching of some trees is peculiar. In the Amazon district many Myristicaceae and Monimiaceae have verticillate branches coming off iu fives. This is also seen in the Chili Pine. Some Amazon trees taper downwards, so as to have a form like an inverted cone or pyramid, as in the Mulatto tree (Euky∙ lista Spruceana), one of the Cinchonaceae.

Branches are sometimes long and slender, and run along the ground, producing buds with roots and leaves at their extremity or apex. This is seen in the runner (flagellum) of the Strawberry. In the Houseleek (Sempervivum) there is a similar prostrate branch of a shorter and thicker nature, producing a bud at its extremity capable of independent existence. It receives the name of offset (propagιdum). In many instances the branch decays, and the young plant assumes a separate existence. Gardeners propagate plants by the process of layering, which consists in bending a twig, fixing the central part of it iuto the ground, and, after the production of roots, cutting off its connection with the parent. A stolon differs from these in being a branch which curves towards the ground, and, on reaching a moist spot, takes root and forms an upright stem, and ultimately a separate plant. This is a sort of natural layering, and the plant producing such branches is called stoloniferous. In the Rose and Mint a subterranean branch arises from the stem, which runs horizontally to a certain extent, and ultimately sends up an aerial stem, which becomes an independent plant. Such branches are denominated suckers, and the plants are surculose. The gardener divides the connection between the sucker and the parent stem, in order to propagate these plants. In the case of Asparagus and other plants which have a perennial stem below ground, subterranean buds are annually produced, which appear above ground as shoots or branches covered with scales at first, and ultimately with true leaves. The young shoot is called a turio. These branches are herbaceous and perish annually, while the true stem remains below ground ready to send up fresh shoots next season. In Bananas and Plantains the apparent aerial stem is a shoot or leaf-bud sent up by an underground stem, and perishes after ripening fruit. Branches are sometimes arrested in their development, and, in place of forming leaves, become transformed into spines or thorns, as in the Hawthorn. Plants which have spines in a wild state, as the Apple and Pear, often lose them when cultivated, in consequence of their being changed into branches; in some cases, as in the Sloe (Prunus spinosa), (fig. 65), a branch bears leaves at its lower portion, and terminates in a spine. Plants bearing thorns (modifications of branches or leaves) are denominated spiny, spinose, or spinescent. A bud is sometimes developed as a slender spiral or twisted branch, called a tendril or cirrus. In the Passion-flower the lateral buds are thus altered, with the view of enabling the plant to climb. In the Vine the tendrils are looked upon as the terminations of separate axes, or as transformed terminal buds, and are sometimes called sarmenta. In the Vine there are no young buds seen in the angle between the stem and leaves, nor between the stem and tendrils; and the latter are not axillary.

Fig. 66 represents the branch of a Vine, in which a' is the primary or first formed axis, ending in υ, a tendril or altered terminal bud, and having a leaf f' on one side. Between this leaf and the tendril, which represents the axis, a leaf-bud was formed at an early date, producing the secondary axis, or branch a! ’ , ending in a tendril v", with a lateral leaf f", from which a tertiary axis or branch a" was developed, ending in a tendril υ"', and so on. Tho tendrils of Ampélopsis Veitchii are terminated by disks which secrete a sticky matter, by means of which they adhere to walls, <fcc. The tendrils, like those of the Vine, are modifications of the axis. Tendrils twist in a spiral manner, and enable the plants to rise into the air by twining round other plants. The direction of the spiral frequently differs from that of the climbing stem which produces the tendril. In the Vine the lower part of the stem is strong, and needs no additional support; the tendrils therefore occur only in the upper part, where the branches are soft, and require aid to enable them to support the clusters of fruit. In the Bryony the tendril in the first part of its course twines from right to left, and in the last from left to right.

In some instances lateral buds are found without being in the axil of leaves. In this case they are extra-axillary. Such buds are produced after the stem and leaves have been formed, and in certain circumstances they are developed like normal buds. What have been called embryo-buds are woody nodules seen iu the bark of the Beech, Elm, and other trees. They are partially developed adventitious or abnormal buds, in which the woody matter is pressed upon by the surrounding tissue, and thus acquires a very hard and firm texture. When a section is made, they present woody circles arranged around a central pith, and traversed by medullary rays. The nodules sometimes form knots on the surface of the stem, at other times they appear as large excrescences, and in some cases twigs and leaves are produced by them.

Buds sometimes become extra-axillary in consequence of the non-appearance or abortion of one or more leaves, or on account of the adhesion of the young branch to the parent stem. In place of one bud there are occasionally several accessory ones produced in the axil, giving origin to numerous branches. Such an occurrence is traced to the presence of latent or adventitious buds. By the union of several such buds, branches are produced, having a thickened or flattened appearance, as is seen in the Fir, Ash, and other trees. These fasciated branches, in some cases, however, are owing to the abnormal development of a single bud. Occasionally adventitious buds are produced on the edges of leaves, as in Bryophyllum calycinum (fig. 67), Malaxis paludosa, and various species of Asplénium, and on the surface of leaves, as in Ornithogalum thyreoideum. These are capable of forming independentplants. Similar buds are also made to appear on the leaves of Gesnera, Gloxinia, and Achimenes, by wounding various parts of them, and placing them in moist soil; this is the method often pursued by gardeners in their propagation. The Ipecacuan plant has been propagated by means of leaves inserted in the soil. In this case the lower end of the leaf becomes thickened like a corm, and from it roots are produced, and ultimately a bud and young plant. Leaves bearing buds on their margin are called proliferous.

The typical form of stems is rounded. They are sometimes compressed or flattened laterally, while at other times they are angular: being triangular, with three angles and three flat sides; trigonous, with three convex faces; triquetrous, with three concave sides; quadrangular or square; quinquangulär or five-angled ; octangular or eight-angled, <fcc. Various terms are applied to the forms of stems, as cylindrical or terete, jointed or articulated, <fcc. The following are some of the more important modifications of stems :—The crown of the root is a shortened stem, often partially under ground, which remains in some plants after the leaves, branches, and flower-stalks have withered. In this case the internodes are very short, and the nodes are crowded together, so that the plant appears to be stemless. It is seen in perennial plants, the leaves of which die down to the ground annually. A rhizome or root-stock (fig. 68) is a stem which runs along the surface of the ground, being partially covered by the soil, sending out roots from its lower eide and leaf-buds from its upper. It occurs in Ferns, Iris, Hedychium, Acorus or Sweet Flag, Ginger, Water-lily, many species of Carex,

Rushes, Anemone, Lathraea, <fcc. By many the term rhizome is applied to stems creeping horizontally, whether they are altogether or only partially subterranean. In rhizomes called definite, the terminal bud gives off flowers, and the lateral buds form the stem; while in indefinite rhizomes the terminal leaf-bud is formed annually. A rhizome such as occurs in Solomon’s Seal (fig. 68) is not a single stem, i,e., the product of a single bud, but is composed of portions of successive axes, the leaves of which have died off and remain as scars (fig. 68, c, c); it is thus an indefinite rhizome. Rhizomes are well seen in British Ferns. A rhizome sometimes assumes an erect form as in Scabiosa succisa, in which the so-called praemorse root is in reality a rhizome, with the lower end decaying. The erect rhizome of Cicuta virosa shows hollow internodes, separated by partitions.

A pseudo-bulb (fig 62) is an eularged bulbous-like aerial stem, common in Orchidaceous plants. It is succulent, often contains numerous spiral cells and vessels, and is covered with a thick epidermis. A soboles is a creeping underground stem, sending roots from one part and leafbuds from another, as in Couch-grass, Carex arenaria, and Scirpus lacustris. It is often called a creeping root, but is really a rhizome with narrow elongated internodes. A tuber is a thickened stem or branch produced by the approximation of the nodes and the swelling of the internodes, as in the Potato. The eyes of the Potato are leafbuds. Tubers are sometimes aerial, occupying the place of branches. The ordinary herbaceous stem of the Potato, when cut into slips and planted, sends off branches from its base, which assume the form of tubers. Tubers frequently store up a quantity of starch as in Maranta arundinacea, whence arrowroot is derived. Another form of thickened underground stem is the corm, as seen in the Autumn Crocus (Colchicum, fig. 69), Gladiolus, &c. Structurally it is composed of a solid more or less rounded axis covered by a layer of thin membranous scales (fig. 70, h, h). A

corm is only of one year’s duration, giving off buds annually in the form of young corms. In autumn the young corm gives origin to leaves, the lower of which (s, s', s") form sheaths round the corm and flower-stalk, the upper (l', I") remaining very small; and in the axil of the uppermost leaves the flowering-stem develops and bears the flowers (δ, δ'). Meanwhile in the axil of the middle leaves on the corm, a bud—the rudiment of a new corm—appears (k"). The flowering-stem dies down, and the corm from which it arose enlarges greatly during the winter at the expense of its parent corm (k), which thus becomes shrivelled. In spring the leaves produced on it (Γ, Γ), which were merely rudiments in autumn, appear above ground as conspicuous large leaves. At the end of spring these leaves die down, the bases of the lower ones alone remaining, and constituting thin brown scales around the corm (as at h). Meanwhile, the -young bud corm (k") in the axil of the middle leaves grows rapidly at the expense of its parent corm (F), but it does not attain a great size. In autumn it produces new leaves, which remain small, but from the axil of the two upper the flowering stem rises up and bears flowers; whilst in the axil of its middle leaves a new bud-corm appears, which will the following autumn produce young leaves, flowering stem, and a new bud-corm, and thus the cycle goes on. The buds or new corms formed from the old corms may be produced either laterally, as in Colchicum autumnale, or terminally, as in Crocus and Gladiolus. The bulb is another form of underground stem or bud. The axis in this case is much shortened, and the internodes are hardly developed. The bases of the leaves rising from the stem are quite close together, and become succulent and enclose the axis. In the Lily the thick and narrow scales are arranged separately in rows, and the bulb is called scaly; while in the Leek, Onion, Squill, and Tulip the scales are broad, and enclose each other in a concentric manner, the outer ones being thin and membranous, and the bulb is tunicated. In the axils of these fleshy scales new lateral shoots arise, forming new bulbs. The lateral buds or cloves sometimes remain attached to the axis, and produce flowering stems, so that apparently the same bulb continues to flower for many years, as in the Hyacinth and Tulip; at other times the young bulbs are detached, and form separate plants. In the axil of the leaves of Lilium bulbiferum, Dentaria bulbifera, and some other plants, small conical or rounded bodies are produced, called bulbils or bulblets (fig. 71, δ). They resemble bulbs in their aspect, and consist of a small number of thickened scales enclosing a growing-point. These scales are frequently united closely together, so as to form a solid mass. Bulbils are therefore transformed leafbuds, which are easily detached, and are capable of producing young plants when placed in favourable circumstances. The scales in bulbs vary in number. In Gagea there is only one scale; in the Tulip and Fritillaria imperialis they vary from two to five; while in Lilies and Hyacinths there are a great number of scales. In the Tulip a bud is formed in the axil of an outer scale, and this gives rise to a new flowering axis, and a new bulb, at the side of which the former bulb is attached in a withered state.

The forms of the stem having been considered, we now proceed to examine its anatomical structure. This structure consists of the elementary tissues combined and arranged in various ways. The arrangement of the fibro-vascular bundles in the mature stem or root is not the same in all plants. But we find that in most plants which have two seed-lobes in the embryo, i.e., Dicotyledons, a characteristic structure is apparent, quite distinct from what is found in the majority of plants in which only one seed-lobe is present in the embryo, i.e., in Monocotyledons; and these, again, have a different structure from that found in. Acotyledons, or plants with no seed-lobe in their embryo. The three forms of stem here refeιτed to have been usually distinguished as follows :—(1.) Exogenous stem, in which the fibro-vascular bundles are produced indefinitely in an outward direction, and the stem increases in diameter by the annual formation of a new layer of woody matter on the outside of the preceding layers. This is the form found in most Dicotyledons, and they have hence been called Exogens, or Outward-growers. Ordinary trees, such as tho Oak and Ash, furnish instances. (2.) Endogenous stem, in which the fibro-vascular bundles are definite, and are formed towards the centre, which becomes filled up with them in the progress of growth, so that the diameter of the stem increases in a great measure by the new matter pushing out that previously formed. This stem characterizes many Monocotyledons, which have thus been called Endogens, or Inward-growers. Palms supply examples. (3.) A erogenous stem, in which the bundles of vessels are simultaneously developed, and the additions to the stem take place at the summit by the union of the bases of the leaves. Plants having this kind of stem are called Acrogens, or Summitgrowers, and are Acotyledons. Tree-ferns furnish an example. Recent research, however, has shown that these terms cannot always be used as synonymous with Dicotyledon, Monocotyledon, and Acotyledon, as we find amongst Dicotyledons stems where the formation of new fibro-vascular bundles is distinctly endogenous, and again amongst Monocotyledons stems with a provision for exogenous growth, and also amongst Acotyledons examples are not wanting in which a continuous increase in diameter is provided for.

We shall consider the structure of the stem in Dicotyledons, Monocotyledons, and Acotyledons successively.

In the young stem of a Dicotyledon the fibro-vascular bundles first appear as a circle of wedge-shaped masses, by which the stem is divided into a central cellular portion, the pith, and a peripheral cortex,— the space between the bundles being occupied by cellular tissue constituting the medullary rays, and uniting the pith and cortex. Each fibro-vascular bundle increases by division of its own cambium cells. If eventually all the cambium cells become permanent tissue then the bundle becomes closed, and all further growth ceases. This is the complete structure of an annual herbaceous dicotyledonous stem, which thus consists of a central cellular pith, a circle of fibro-vascular bundles, a cellular cortex united with the pith by medullary rays, and outside all an epidermis (fig. 72). In trees and shrubs with permanent woody stems, the young shoots given out annually have a structure similar to that of annual herbaceous stems; but as the shoot grows, further changes occur by which the diameter is increased, and the stem becomes more dense. After the first year’s growth, the cells of the medullary rays, stretching between the fibro-vascular bundles and continuous with their cambium cells, become converted into secondary meristem, and then an interfascicular cambium is formed, which eventually coalesces with the cambium cells of the fibro-vascular bundles, and thus a complete cambium ring is formed (fig. 73). From this cambium ring new xylem or wood is formed on the inside, whilst new cortex is formed on the outside; and in this way the stem increases in thickness. This growth in thickness, however, ceases periodically, and is renewed with each new period of vegetation, and thus the wood is formed in concentric layers, sharply marked off from succeeding layers, each being an annual ring of wood, and the same is seen in the cortex (fig. 74). The rings of wood are thus formed successively outside those pre-existing, while in the cortex the new layers are produced inside those already formed.

The inner vessels of the primary fibro-vascular bundles immediately surround the pith and often project into it, and form what is termed the medullary sheath, which consists of spiral vessels, and through this sheath the primary medullary rays pass. These medullary rays extend from the pith to the cortex, but as new zones of wood and cortex are produced, new rays are formed in them, which increase by additions from the cambium layer. The secondary cortex, formed from the cambium ring, constitutes what is commonly known as the inner bark or endophloeum ; the primary cortex, which forms the outer bark, consists of two layers of cells, which have been respectively termed the mesophloeum and the epiphloeum. Outside all is the epidermis. This, however, does not remain as thin-walled cells, but is usually converted into periderm, and this may in turn be completely supplanted by the bark. Thus, if a transverse and a longitudinal section of a Dicotyledonous stem be made, the following structures will be seen as represented in fig. 74. In the centre is the cellular pith a, a, immediately surrounding it comes the medullary sheath d, then the secondary layers of wood b, c, of successive years’ growth, outside this the cambium ring e, separating the woodlayers from the cortical layers. Of these last, the inner,/, is the bast layer or endophloeum, g is the middle cellular layer or mesophloeum, and d is the outer epiphloeum. The latter two constitute the outer bark. Connecting the mesophloeum with the pith is seen a medullary ray i. Outside all are the epidermal tissues.

Let us now examine the different parts of an Exogenous stem proceeding from the centre to the circumference :—

The Pith, or the central part of a Dicotyledonous stem, is composed of cellular tissue. In the young stem it is succulent, the cells being full of fluid and frequently of a greenish hue; but in process of time it becomes palecoloured, dry, and full of air. Thesé changes take place first in the central cells. Sometimes the pith is broken up into cavities, which have a regular arrangement, as in the Walnut, Jessamine, and Cecropia peltata; it is then called discoid or disciform. At other times, by the rapid growth of the outer part of the stem, the pith is ruptured irregularly, and forms large cavities, as in the fistular stem of Umbelliferous plants. In some cases fibro-vascular bundles are found in the pith, as in Elder, Pitcher-plant, and Ferula, and occasionally its cells are marked by pores indicating a thickening of the cell-wall. The extent of pith varies in different plants and in different parts of the same plant. In Ebony it is small, while in the Elder it is large. In Aeschynomene aspera (Shola plant, the Ricepaper plant of India), the interior of the stem is almost entirely composed of cellular tissue or pith; from this a kind of paper is made, and light hats. The same kind of tissue occurs in the Papyrus of the Nile. Large pith is also seen in Fatsia (Aralia) papyrifera (Tung-tsaou or Chinese rice-paper plant), and in Scaevola Taccada of the Malay archipelago. When the woody circle of the first year is completed, the pith usually remains stationary as regards its size, retaining more or less its dimensions, even in old trunks, and never becoming obliterated.

The Medullary Sheath is the fibro-vascular · layer immediately surrounding the pith. It is the inner layer of the fibro-vascular bundle of the first year (fig. 74, d), and consists chiefly of true spiral vessels, with annular and reticulated vessels, intermixed with long woody fibres, which continue to exercise their functions during the life of the plant, and which extend into the leaves. Between the vessels of the sheath the medullary rays from the pith pass.

The Wood.— The layers of wood (fig. 7 4, b, c) are formed outside the medullary sheath in concentric rings in the manner already described. On account of this mode of formation of wood-layers successively outside pre-existing layers the stem increases indefinitely. There are no annular or spiral vessels present; these have been replaced by pitted and punctated vessels along with wood-cells. The stems have been called exogenous and also indefinite, and Dicotyledonous plants have sometimes received the name of Cyclogens, in consequence of exhibiting concentric circles in their stems. On a transverse section each zone or circle is usually seen to be separated from that next to it by a well-marked line of demarcation. This line, as in the Oak and in the Ash, is indicated by holes which are the openings of large pitted vessels,—the remainder of the tissue in the circle being formed by pleurenchymatous tubes with thickened walls and of smaller calibre. In some trees, as the Lime, Hornbeam, and Maple, the line is by no means so well marked, as the openings are smaller and more generally diffused; but there is usually a deficiency of pitted vessels towards the outer part of the circle. In cone-bearing plants, as the Fir, in which the woody layers consist entirely of punctated tissue, without any large pitted vessels, the line of separation is marked by the tissue becoming dense and often coloured. In some kinds of wood, as Sumach, the zones are separated by a marked development of cellular tissue. The separation between the zones is owing to the interruption in the growth of the tree during autumn and winter, and hence it is well defined in trees of temperate and cold climates. But even in tropical trees, the lines, although often inconspicuous, are still visible,<— the dry season, during which many of them lose their leaves, being their season of repose.

The woody layers vary in their texture at different periods. At first all the tissues are pervious and full of fluid; but by degrees they become thickened, and the channels of the vessels get filled up and obliterated. The first-formed layers are those which soonest become thus altered. In old trees there is a marked division between the central heart-wood or duramen, and the external sapwood or alburnum,— the former being hard and dense, and often coloured, with its tubes dry and thickened, while the latter is less dense, is of a pale colour, and has its tubes permeable by fluids. The difference of colour between these two kinds of woods is often very marked. In the Ebony tree the duramen or perfect-wood is black, and is the part used for furniture, while the alburnum is pale; in the Beech, the heart-wood is light-brown; in the Oak, deep-brown; in Judas tree, yellow; in Guaiacum, greenish. The alteration in colour is frequent in tropical trees. In trees of temperate climates, called white-wood, as the Willow and Poplar, no change in colour takes place; this is also the case in the Chestnut and Bombax. The relative proportion of alburnum and duramen varies in different trees. The heart-wood is more useful than the sap-wood, and is less liable to decay.

From the mode in which the woody layers are formed, it is obvious that each vascular zone is moulded, upon that which precedes it; and as, in ordinary cases, each woody circle is completed in the course of one year, it follows that, by counting the concentric circles, the age of a tree may be ascertained. Thus fig. 75 represents an oak eight years old, having eight woody layers b. This computation can only be made in trees having marked separations between the circles. There are, however, many sources of mistake. In some instances, by interruption to growth, several circles may be formed in one year, and thus lead to an erroneous estimate. Care must be taken to have a complete section from the bark to the pith, for the circles sometimes vary in diameter at different parts of their course, and a great error might occur from taking only a few rings, or circles, and then estimating for the whole diameter of the tree. When by the action of severe frost, or other causes, injury has been done to the tender cells from which the young wood is developed, while, at the same time, the tree continues to live, so as to form perfect woody layers in subsequent years, the date of the injury may be ascertained by counting the number of layers which intervene between the imperfectly formed circle and the bark. Inscriptions made in the wood become covered, and may be detected in after years when a tree is cut down; so also wires or nails driven into the wood. As the same development of woody layers takes place in the branches as in the stem of an exogenous tree, the time when a branch was first given off may be computed by counting the circles on the stem and branch respectively. If there are fifty circles, for instance, in the trunk, thirty in one branch and ten in another, then the tree must have been twenty years old when it produced the first, and forty when it formed the other.

In exogenous stems the pith is not always in the centre. The layers of wood on one side of a tree may be larger than those on the other, in consequence of their fuller exposure to light and air, or the nature of the nourishment conveyed, and thus the pith may become excentric. Zones vary in size in different kinds of trees, and at different periods of a plant’s life. Soft wooded trees have usually broad zones, and old trees form smaller zones than young ones. There are certain periods of a plant’s life when it seems to grow most vigorously, and to form the largest zones. This is said to occur in the Oak between twenty and thirty years of age.

C ’ cmδ⅛7n.—External to the layers of wood, and between them and the bark, there is a layer of thin-walled cells in which the protoplasm and cell-sap remain, and consequently they are capable of division and growth. To this layer the name of cambium, has been given. This cambium layer marks the separation between the wood and the bark, and may be regarded as constituting the active formative tissue of Dicotyledonous stems. It constitutes the thickening zone, by means of which the stem is enlarged,— the cambium cells situated most internally being subservient to the purposes of the wood formation, while the external ones give origin to the new bark. When these cells are carrying on the process of growth with activity, during the flow of the sap in spring, the bark can be easily separated from the wood.

The Bark or Cortical System, lies external to the wood, and, like it, consists of several layers. In the early state it is entirely cellular, and is in every respect similar to the pith; but as the fibro-vascular bundles are developed, the bark and pith are separated, and the former gradually becomes altered by the formation of secondary deposits. We find in the cortex, as in the wood portion of the stem, fibro-vascular along with cellular tissue. But the position and relative proportion of these two systems is reversed. In the bark the cellular system is external, and is much developed, while the fibro-vascular is internal, and occupies comparatively a small space. The cellular portion of the bark consists of an external layer, or epiphloeum, and the cellular envelope, or mesophloeum, while the vascular system forms the internal portion called liber, or endophloeum.

The endophloeum, liber, or inner bark, is formed from the secondary cortex of the young stem. It consists mainly of thick or thin walled woody fibres, commonly known as bast-fibres, mixed with elongated cellular tissue and frequently with laticiferous vessels. It is separated from the wood by the cambium layer. The pleurenchymatous tubes are thickened so as to be flexible, but are not lignified, and are thus very tenacious. The endophloeum of the Lime-tree and of Antiaris saccidora (the Sack tree of Coorg) is used to form mats, cordage, and bags; and the toughness of the fibres of the inner bark of flax, hemp, and of many of the Nettle and Mallow tribes, render them fit for various manufacturing purposes. The endophloeum is sometimes, from its uses, called the bast-layer. Occasionally it is continuous and uninterrupted, as in the Vine and Horse-chestnut; at other times, as in the Oak, Ash, and Lime, the fibres are separated during the progress of growth, and form a sort of network, in the interstices of which the medullary rays are seen. The fibres of the Lace-bark tree (Lagetta lintearia) are similar. In fig. 76 is represented the bark of Daphne Laureola,— f indicating the woody fibres of liber, and r the medullary rays. The endophloeum increases by layers on its inside, which are thin, and may be separated like the leaves of a book. The outer layer of bastfibres betwixt the endophloeum and the outer bark has been termed the cortical sheath, corresponding to the medullary sheath on the inside of the stem.

The outer bark is formed from the primary cortex; it is always cellular, and is divided into two layers, the epiphloeum with the mesophloeum underneath. The cellular envelope, or mesophloeum, lies immediately on the outside of the liber. It consists of polyhedral, often prismatical cells, elongated vertically to the surface, usually havingchlorophyll, or green colouring matter, in their interior, but sometimes being colourless, and containing raphides. They are distinguished from those of the epiphloeum by their form and direction, by their thicker walls, their green colour, and the intercellular spaces which occur among them. This covering is usually lese developed than the outer suberous layer, but sometimes, as in the Larch and common Fir, it becomes very thick, and separates like the epiphloeum. In the cellular envelope laticiferous vessels occur. The epiphloeum is the outer covering of the bark, consisting of cells which usually assume a cubical or flattened tabular form. The cells have no chlorophyll in their interior, are placed close together, and are elongated in a horizontal direction; and thus they are distinguished from the cells of mesophloeum. In the progress of growth they become often of a brown colour. This covering may be composed of a single layer of tabular cells; but in some trees it consists of numerous layers, forming the substance called cork, which is well seen in Quercus Suber, the Cork-oak; hence the name suberous, or corky layer, which is given to it. The form of its cells varies in some instances, being cubical at one part, and more compressed or tabular at another, thus giving rise to the appearance of separate layers. On the exterior of the epiphloeum is situated the epidermis, which has already been described. It is formed of a layer of cells, which in woody stems serve only a temporary purpose, becoming ultimately transformed in various ways.

The bark, in its increase, follows an order exactly the reverse of that which occurs in the woody layers. The layers of liber owe their increase to the cambium cells, which, by their constant reproduction, mark the separation between the xylem and phloem portions of the stem. These layers are often so compressed and united together as to be counted with difficulty, while at other times they are separated by rings of cellular tissue, and thus remain conspicuous. As the additions are made to the woody layers on the outside, and to the bark on the inside, there is a constant distension going on, by which the bark becomes compressed, its layers of liber are condensed, the fibres are often separated (fig. 76) so as to form meshes, its epidermis is thrown off, and the epiphloeum is either detached along with it, or, when thick, is ruptured in various ways, so as to give rise to the rugged appearance presented by such trees as Elm and Cork-oak. In some instances the bark is very distensible, and its outer cellular covering is not much developed, so that the surface remains smooth, as in the Beech. The outer suberous layer sometimes separates with the epidermis, in thin plates or scales; in the Birch these have a white and silvery aspect. There is thus a continual destruction and separation of different portions of the bark. The cellular envelope and liber may remain while the epiphloeum separates, or they also may be gradually pushed off—the parts which were at first internal becoming external. In the case of some Australian trees both the cellular and fibrous portions are detached in the form of thin flakes, and occasionally each annual layer of liber pushes off that which preceded it. The epidermis separates early, and no renewal of it takes place. It is, however, replaced by the cork layer, which then covers the outer part of the stem. To this covering the name penderm is given.

From the mode in which the outer layers of bark separate, it follows that inscriptions made on them, and not extending to the wood, gradually fall off and disappear. A nail driven into these layers ultimately falls out. In consequence of the continued distension of an exogenous stem, it is found that woody twining plants cause injury, by interrupting the passage of the fluids. Thus a spiral groove may be formed on the surface of the stem by the compression exercised by a twining plant, such as Bush-rope (Bauhinia, fig. 77) or Honeysuckle. From what has been stated relative to the changes which take place in the bark, it will be understood that it is often difficult to count its annual layers, so as to estimate the age of the tree by means of them. This may, however, be done in some cases, as shown at fig. 75, where there are eight layers of bark e, and eight woody layers b. Medullary Rays or Plates.— While the bark and pith become gradually separated by the intervention of vascular bundles, the connection between them is kept up by means of processes called medullary rays (figs. 78 and 79). These form the silver grain in wood, so conspicuous in the maple; they communicate with the pith and the cellular envelope of the bark, and they consist of cellular tissue, which becomes compressed and flattened so as to assume a muriform appearance. At first they occupy a large space (fig. 72, st); but as the vascular bundles increase they become more and more narrow, forming thin laminae or plates, which separate the woody bundles. On making a transverse or horizontal section of a woody stem, the medullary rays present the aspect of narrow lines running from the centre to the circumference (fig. 74); and on making a vertical section of a similar stem through one of the rays, the appearance represented in fig. 78 will be observed, where a medullary ray mr, composed of flattened muriform cells, passes from the pith p to the cellular envelope ce, crossing the tracheae of the medullary sheath t, the ligneous tissue I, the pitted vessels of the wood b, and the fibres of the liber cf. The laminae do not by any means preserve an uninterrupted course from the apex to the base of the tree. They are broken up by the intervention of woody fibres, as seen in a vertical section of a woody stem (fig. 79), tangentially to the medullary rays mr, mr, mr, which are separated by interlacing fibres ll. The primary medullary rays extend completely from the pith to the bark; but in the secondary wood and secondary cortex new rays are formed which, therefore, extend only through a portion of the stem. These are secondary medullary rays. All may increase by division of the merismatic cells of the cambium. Medullary rays are conspicuous in the Cork-Oak, Hazel, Beech, Ivy, Clematis, Vine. They are not so well marked in the Lime, Chestnut, Birch, Yew.

The medullary rays are in some cases, as in Clematis and Aristolochia, large and broad, while the woody wedges are comparatively small.

The stems of Dicotyledonous plants occasionally present anomalous appearances in the structure and arrangement of their wood, bark, and medullary rays. In place of concentric circles there are sometimes only a few rows of wedge-shaped vascular bundles produced during the life of the plant, additions being made by the annual interposition of bundles of a similar kind, resembling in this respect the formation of woody bundles in the early growth of herbaceous plants. In Piperaceae, Aristolochiaceae, and Menispermaceae, these anomalous stems occur. In Gnetum (fig. 80), the vascular bundles, δ, b, b, b, b, form zones, which are each the produce of several years’ growth, and are separated by layers, I, I, I, I, I, I, which may be considered as representing different bast-layers. In some of the Menispermaceae the separating layers are of a cellular and not of a fibrous nature. Many of the Malpighiaceae, Sapindaceae, and Bignoniaceae of Brazil exhibit stems in which the woody layers are arranged in a very irregular manner. In some of them (fig. 81) there is a central woody mass with from three to ten small secondary ones round it. Each of the masses contains true pith, derived either from the cortical cellular tissue, or from the original medullary centre. Around these separate collections of pith there is a medullary sheath and spiral vessels. No annual rings have been detected in the secondary masses, but medullary rays exist usually in their outer portion. In some anomalous Sapindaceae, the central and lateral woody masses are enclosed in a common bark, with a continuous layer of liber. Some have supposed that the lateral masses are undeveloped branches united together under the bark.

In some Bignoniaceae (fig. 82) the layers of wood are divided in a crucial manner into four wedge-shaped portions by the intervention of plates differing in texture from the ordinary wood of the plant, and probably formed by the introversion, or growing inwards, of the liber. In Aspidosperma excelsum (Paddle-wood) of Guiana, and in Heteropterys anomala, the stem assumes a peculiar lobed and sinuous aspect; and in some woody climbing plants pressure causes the stems to become flattened on the side next the tree on which they are supported, while from being twisted alternately in different directions, they present a remarkable zigzag form, having the woody layers developed only on one side (fig. 83). In Firs the wood is occasionally produced in an oblique in place of a perpendicular manner, thus injuring the timber and causing it to split in an unusual way. The young plants produced from the seed of such twisted-wooded firs are said to inherit the peculiarity of their parents. Occasionally the Dicotyledonous stem becomes swollen at certain places, especially near the root, and thus exhibits a tuberous appearance. This peculiarity is said to be liable to occur in Coniferous plants grown from cuttings. In some of the lower class of plants a cellular stalk is produced, which on a transverse section presents an appearance like that of a Dicotyledonous stem. Thus Lessonia fuscescens, a species of sea-weed, has stems which are often 5 to 10 feet long, and as thick as the human thigh, and which show concentric elliptic cellular rings. Such is also the case with Usnea melaxantha, a tree like lichen. In these plants, however, the structure is entirely cellular, and quite distinct from that of Dicotyledonous plants.

In the young stem of a Monocotyledonous plant the fibro-vascular bundles appear scattered in an irregular manner through the fundamental parenchyma, becoming more numerous towards the periphery. There is thus no primary separation in the stem into pith, cortex, and medullary rays, although the central cellular mass may be considered as representing the pith. Thus, if a section be made across a young Monocotyledonous stem, an appearance is observed such as is represented in fig. 84, where the vessels are seen as points, scattered through a cellular matrix, the circumference not having any marked cortex, and the whole covered by an epidermis. A similar section of a further advanced stem, as of a Palm (fig. 85), shows numerous bundles of vessels dispersed irregularly in cellular tissue; those near the centre, m, being scattered at a distance from each other, while those towards the outside are densely aggregated, forming a darkish zone b, and are succeeded at the circumference by a paler circle of less compact vessels I, with some compressed cells, covered by an epidermis e. The central cellular mass has no medullary sheath. In some cases its cells are ruptured, and disappear during the progress of growth, leaving a hollow cavity; but in general it remains permanent, and is gradually encroached upon by the development of the vascular system. The peripheral portion differs from true bark in not being separable from the rest of the tissue. It has received the name of false bark, and consists of the epidermal cells e, and what has been called the cortical integument, I. This portion of the stem is often very inconspicuous, but sometimes it is much developed, as in Testudinaria Elephantipes, in which it is rugged, and is formed of a substance resembling cork in many respects. The fibro-vascular bundles of the stem very soon become closed, and thus all growth in them ceases, the cambium cells being converted completely into permanent tissue. Fig. 57 represents a transverse section of a closed fibro-vascular bundle from the stem of the maize (Zea Mays). In it the several elements seen in the section of the fibro-vascular bundle from a Dicotyledonous stem may be recognized, but there is no cambium, the cells marked v, v representing those cambiform cells which have last become permanent tissue. This absence of cambium necessarily curtails the growth of the bundle, and hence the limited diameter characteristic of the stems of Monocotyledons. In fig. 86 a diagram represents a transverse and longitudinal section of a Monocotyledonous stem.

All the fibro-vascular bundles of the stem are common bundles, that is, they pass out into the leaf, and in these stems each bundle has a definite arrangement. At the point where the bundle curves out into the leaf it has its greatest thickness, gradually becoming attenuated a3 it passes upwards into the leaf and downwards into the stem. Iu some instances, however, the bundles as they descend increase at different parts of their course, probably by interstitial growth, and give rise to irregular swellings of the stem, as in Ceroxylon Andicola. The distension takes place occasionally at the base of the stem, as in Euterpe montana. This downward prolonged portion does not, however, run a straight course, but first passing inwards towards the centre of the stem, it then gradually arches outwards towards the periphery. This passing inwards at first of the fibro-vascular bundle gave origin to the idea that the first formed fibres were gradually pushed towards the circumference by those which succeeded them, and that the wood portion of these stems was increased by additions to the centre; hence the term endogenous applied to them, meaning internal growth. But, as has been shown, the fibro-vascular bundles really become external at the base, although internal above. On making a vertical section of the endogenous stem, as of a Palm (fig. 87), there is observed an interlacing of fibres, the fibro-vascular bundles are first directed towards the centre, and then curve outwards towards the circumference, so that those last formed ultimately become external. The term endogenous will, therefore, only apply strictly to the fibres at the early part of their course. The true distinction between exogenous and endogenous stems is, that in the former the fibro-vascular bundles remain open, a cambium ring being eventually formed from which the stem increases indefinitely in diameter. In the latter the fibro-vascular bundles soon become closed, and being scattered irregularly through the cellular tissue, and not in a circle, no cambium ring can be formed, and thus growth in a transverse direction is soon arrested, and the stem is of a comparatively uniform diameter throughout. The investing bark of the former permits an unlimited extension of woody growth beneath it; the fibrous cortical layer of the latter, by maintaining an intimate union with the subjacent tissue, prevents unlimited increase in diameter. Hence we find that the stem does not attain the enormous diameter exhibited by some Dicotyledonous trees, such as Sequoia (Wellingtonia) gigantea and the Baobab,—the former of which has been measured 116 feet in circumference. In consequence of this mode of formation, the outer part of a Palm-stem is the hardest and densest, and after acquiring a certain degree of firmness it resists all further distension, and frequently becomes so hard as to withstand the blow of a hatchet, and therefore a woody twining plant does less injury to it than to trees of exogenous growth (fig. 88). The growth of endogenous stems may be said to resemble an upward growth of an exogen by the terminal bud only, for there is no cambium layer, and no peripherical increase. The terminal central bud is called a phyllophor or phyllogen. As growth only proceeds in this manner, no annual rings of wood are formed. From the absence of concentric circles the age of a Palm cannot be estimated in the same way as that of an exogenous tree. The elongation, however, of each species of Palm is pretty regular, and by this some idea may be formed of its age. There are many herbaceous plants in Britain, as Lilies, Grasses, &c., having Monocotyledonous stems, but there are no British Monocotyledonous plants with permanent aerial woody stems. All the British trees are Dicotyledonous. Illustrations cf Monocotyledonous stems must be taken from trees of other countries, and of these Palms furnish the best examples.

Although this limited growth in a transverse direction is characteristic of most Monocotyledons, we find instances where a true thickening ring is formed and an indefinite increase in diameter takes place, as in Dracaena, Yucca, &c. This thickening ring, however, originates in a manner different from that in Dicotyledonous plants. A layer of the fundamental cellular tissue parallel to the surface of the stem becomes merismatic, and produces new closed fibro-vascular bundles and new cellular tissue between them. The new woody tissue so formed thus corresponds to the secondary wood in Dicotyledonous stems, and the new fibro-vascular bundles are all cauline, that is, do not pass out into the leaves. By this means an enormous increase in diameter of the stem may take place. And not unfrequently the primary central portion of the stem gives way, and thus a hollow cylinder formed by this secondary wood remains. This was well seen in the famous Dracaena Draco, or Dragon tree of Orotava, in the Canary Islands, which had a hollow stem capable of holding several men.

The composition of the vascular bundles in different parts of their course varies. Thus, at the upper part, tracing them from the leaves towards the centre, they contain spiral vessels, pitted vessels with some cellular tissue, a few laticiferous vessels, and woody fibres resembling those of liber. As we descend to the older portions of the bundle, where the tissues have become permanent throughout, the spiral vessels disappear, then the pitted vessels, and at the periphery nothing but pleurenchymatous tissue remains, forming a complicated anastomosis or network. Not unfrequently at the nodes of the stem a network of horizontal vessels occurs. They are well seen in many Grasses, where the central portion of the stem has given way, and there they serve to strengthen the stem. The branching of Monocotyledonous stems is originally the same as in Dicotyledons, namely, a monopodial form; frequently however, the axillary buds do not unfold, and the form becomes much varied. In Palms this is well seen, and usually no lateral shoots are formed. In some Palms, however, as the Doum Palm of Egypt (Hyplwene thebaica), the lateral shoots are developed in such a way that the stem appears to fork. Other examples of the development of branches are seen in the case of the Screw-pine {Pandanus), in Grasses, as the Bamboo, in Asparagus, Cordyline, Dracaena, &c. In some instances the axillary shoots detach themselves from the parent stem and form an independent plant, as in Lilium bulbiferum (fig. 71). The bases of the leaves produced from the stem, remain attached to the stem in many Palms, being surrounded by a fibrous substance, the mattulla or reticτιhιm.

When the internodes of the caudex of a Palm are not much elongated, the scars of the leaves are seen forming spirals on the stem, as in the Coco-nut and Date. In Xanthorrhoea Hastile the same arrangement is observed. In this plant also a curious internal structure of the stem occurs. On making a vertical section the structure appears to be that of a Dicotyledon. The woody part is formed of vertical loose fibres as in Palms, and there are other fibres, radiating from the centre, and cutting the preceding at right angles. These horizontal fibres resemble the medullary rays, but differ in their structure. They probably serve for the origin of leaves, which are numerous, and are disposed throughout the whole length of the stem. In Palms, such as species of Chamaedorea, the internodes are much lengthened, and rings are seen on the stem at distant intervals, showing thickened node-like joints. Some Palm stems, as those of Calamus Rudentum, the common Caue, are very thin and slender. In many Monocotyledonous plants the stem remains below ground, developing shoots which are simple, as in Banana and Plantain, or branched, as in Asparagus. In the former the stem above ground is an herbaceous shoot, composed of the sheaths of the leaves. It dies after fruiting, and is succeeded by other shoots from the subterranean stem. The shoots or budβ from such stems occasionally remain in part below ground in the form of bulbs, as in Lilies, Tulips, and Hyacinths; or as corms, as in Colchicum, Crocus, Gladiolus, and Arum. In some instances the aerial stem has the usual Monocotyledonous structure, while in the underground stem the fibro-vascular bundles are arranged in a circle, enclosing a central cellular mass or pith, and thus resembling in structure a Dicotyledonous stem. This structure has been remarked in the Smilax or Sarsaparilla family. Lindley calls these plants Dictyogens, from their netted leaves, by which they differ from most Monocotyledons.

Amongst Acotyledonous plants there are some which possess stems consisting entirely of cellular tissue, whilst in others the stems have a well-developed vascular system. Of the former we have examples in Mosses, Hepaticae, and Characeae; the latter we find represented in Ferns, Equisetaceae, and Lycopodiaceae. The term Acrogenous has been used as descriptive of the stems of Acotyledonous plants, as they were supposed to be formed by additions to the summit, and by the elongation of vessels already formed. They are also sometimes called Acrobrya. But as in the case of the terms exogenous and endogenous applied to the stems of Dicotyledons and Monocotyledons respectively, recent research has shown that the term acrogenous cannot have the significance formerly attached to it; for in some Acotyledonous plants a true cambium ring is formed by which layers of tissue are successively added, and the stem increases in diameter, and also in many instances the fibro-vascular bundles develop from above downwards. The characters of the stem, however, enable us easily to distinguish it from that in Dicotyledons or Monocotyledons. With merely a slight exception there is no provision for lateral growth of the stem, and all increase in size takes place by an elongation of the terminal growing-point. No cambium ring is as a rule formed; where it does exist, it is not produced in the same way as in Dicotyledons, but rather resembles the formation of the cambium ring in Dracaena amongst Monocotyledons. When a permanent woody stem occurs amongst Acotyledonous plants it resembles in general aspect the stem of a Monocotyledonous plant, having nearly a uniform height, and being usually unbranched and producing a tuft of leaves (fronds) at the summit. Tree-ferns furnish the best example of this kind of stem. In them it is denominated a stipe, and it often attains the height of 120 feet. The stem in Acotyledonous plants is distinguished from that in Dicotyledons by the absence of annual rings of wood, of (with only slight exception) a cambium ring, of a separable bark, and by the fact that the fibro-vascular bundles are all closed; in this latter character they agree with Monocotyledons, but are distinguished from them by the arrangement of the vascular bundles.

In Acotyledonous stems growth takes place by division of a single apical cell, situated at the extreme end of the punctum vegetationis. By divisions of this cell two portions of tissue are marked out in the stem,—an inner portion, from which arise the fibro-vascular tissues when they exist, and an outer or cortical portion. The primary fibro-vascular bundles originate from the cellular tissue of the stem in a manner analogous to what occurs in Dicotyledons and Monocotyledons,—a procambial bundle being first formed, which differentiates subsequently into xylem and phloem layers, but the bundles always become closed. The character of the stem varies very much in the several families of Acotyledonous plants.

In Characeae the stem consists of a series of joints (internodes), each composed of a single much elongated cell. Interposed between successive internodes are the nodes, each composed of a whorl of small cells, from which the leaves are developed, one leaf from every cell of the node. In the genus Chara a cortex is found completely investing the internodal cell. It is formed by the development, from every cell, of the nodes of an ascending and descending lobe. The ascending lobes of one node and the descending lobes of the next higher node meet in the middle of the intervening internode, and there interlock in a prosenchymatous manner. Subsequently by cell-division a complicated structure is developed in this investing mass. No fibro-vascular bundles are found in Characeae.

In Mosses and many Hepaticøe we find no fibro-vascular system, but a slender leafy cellular stem. In it there is sometimes a differentiation into an inner and a peripheral mass of tissue. In the former the cells have usually thin walls, and are not coloured, though in some cases a thickening of the walls of the central cells is observed, and a sort of rudimentary bundle is formed. The circumferential cells have usually thickened walls, and are not unfrequently coloured.

In Ferns we have a familiar example of an Acotyledonous stem. Ordinary ferns of Britain seldom attain any height, but usually creep along the ground, forming rhizomes. But in the Tree-ferns of warm climates the caudex frequently attains a great height. A transverse section of the stem of a fern (fig. 89) exhibits an irregular circle of fibro-vascular bundles, composed of masses/, v, of various forms and sizes, situated near the circumference,— the centre m being formed of cellular tissue, frequently with solitary fibro-vascular bundles scattered through it, and often becoming hollow. On the outside of the vascular circle cells exist, p, covered by an epidermal layer or cellular integument e, often of hard and dense consistence, and marked with the scars of the fronds.

The vascular system is of greater density than the rest of the tissue, and is distinguished by the dark colour of the outer portion (fig. 89, ∕), which surrounds the paler vessels in the centre. In a very young stem only a single fibro-vascular bundle is found in the axis of the stem, surrounded by a mass of cellular tissue; but as growth proceeds a network of anastomosing bundles is formed in place of the central bundle, constituting a wide meshed hollow cylinder (fig. 90), and separating the fundamental cellular tissue of the stem into an outer cortical and an inner medullary portion. In addition, however, scattered through the pith, other bundles occur, as in Pteris aquilina, where two large cauline bundles are found; and in Treeferns numerous small bundles are found which pass out through the meshes of the cylinder into the leaves. The primary bundles are ribbon-shaped and broad, and curve outwards at their margins, from which the bundles for the leaves pass off—a leaf arising always from an opening of the network. The xylem portion of the fibro-vascular bundles is characterized by the great abundance of scalanform vessels, spiral vessels also being present, and intermixed with them thin-walled prosenchymatous cells, containing in winter a large amount of starch. This constitutes the inner, lighter - coloured portion of the bundles. The phloem portion, which completely surrounds the bundle, has also starch-cells, and, in addition, sievetubes and elongated bast-like thick-walled fibres are found at the periphery of the bundle. The whole bundle is usually enclosed by a distinct sheath of elongated cells; these cells are frequently much thickened and of a dark brown colour, hence the appearance presented on transverse section of a dark layer enclosing a lighter-coloured part. The central cellular portion of the stem frequently becomes ruptured and absorbed in old stems, which are thus hollow. The bases of the leaves remain long attached, but ultimately fall off, leaving marked scars (fig. 91, c), which are at first close together, but often separate afterwards by interstitial growth. On these scars or cicatrices the markings of the vessels are easily seen, arranged in the same manner as those of the stem. From all points of the woody cylinder stem-roots may be given off, and frequently they are formed so abundantly as completely to invest the stem and conceal if.

In Equisetaceae or Horsetails we have another example of an Acotyledonous stem. The true stem in those plants is a rhizome underground, from which aerial shoots are sent up annually. Every aerial stem consists of a series of joints (internodes), which are usually hollow and closed at their base by a thin septum (fig. 92). Each internode passes up into a leaf-sheath embracing the next internode, and this sheath is usually split into teeth at its upper margin. In transverse section a circle of fibro-vascular bundles is seen (fig. 93 ) marking off an inner medullary from an outer cortical portion, and separated one from the other by a mass of parenchyma. These bundles are all common bundles. From each tooth of the leaf-sheath the bundle may be traced, passing vertically down into the internodes as far as the node beneath, and at the lower end dividing into two, uniting by each branch with a bundle of the next lowest internode (fig. 94). In transverse section the bundle resembles much that of a Monocotyledon. One peculiarity, however, must be noticed. The spiral, annular, and reticulated vessels at the inner side of the bundle soon disappear, and an air-canal is formed in the bundle. This canal must be distinguished from the vallecular canals found in the cortex, with which they alternate. Round the bundles a single layer of usually thickened and coloured cells forms a sheath, and this may either run continuously round the stem, and thus shut off completely the medullary from the cortical portion of the stem, or may invest separately each bundle, in which case the medullary and cortical portions of the stem are continuous. In some cases a layer of thickened cells is found within the vascular bundles. The cortical portion is divisible into two layers, an inner layer of parenchymatous cells with air canals, the vallecular canalβ, and an outer layer of thickened cells immediately underlying the epidermis. The structure of the rhizome resembles that of the aerial stem, differing only in the fact that the thickened cells of the outer cortical layer are dark coloured, and that the central caual of the internodes is often absent. In Calamites, a fossil plant which flourished abundantly during the Carboniferous Period, we have a plant which in all particulars is closely allied to Equisetum of the present day. But whilst the young stem of Calamites presented a zone of vascular bundles separating it into a medullary and cortical portion (just as in an Equisetum), new vessels were added to the exterior of the pre-existing bundles as the plant grew, so that a series of woody wedges were formed, increasing in growth to an enormous extent.

In Lycopodiaceae, or Club Mosses, the stem presents differences from that in Ferns and Equisetaceae. In most members of the family there is an axial cylinder formed of either one or more fibro-vascular bundles. In some cases (Selaginella) the fibro-vascular bundles remain separate and do not form a cylinder. They are surrounded by a mass of prosenchyιnatous cellular tissue, the cells of which may be thin-walled, leaving no intercellular spaces, or the cellwalls may be much thickened. Immediately surrounding the central axial bundle in Selaginella are large air lacunae, which are wanting in Lycopodium. The fibro-vascular bundles consist of tissues very similar to what occurs in Ferns, and the bundles are all cauline. In the leaves, bundles arise which pass into the stem and coalesce with the cauline bundles. In Isoetes we find a most anomalous structure. Here there are no internodes in the stem; consequently, it is much shortened. The central axial fibro-vascular bundle is very poorly developed, and round it a meristem is formed, which, by adding new parenchyma outside, increases the diameter of the stem. Thus we have here a thickening ring, but it differs from that structure in Dicotyledons, inasmuch as it is not formed by a union of cambium cells of the fibro-vascular bundles. It resembles more the thickening ring formed in Dracaena, Yucca, and other Monocotyledons.

In Rhizocarpeae the structure of the stem is very simple, consisting of an axial fibro-vascular bundle surrounded by parenchyma full of air-cavities.

In plants with Acotyledonous stems the mode of branching is various. In Lycopodiaceae we have an example of a dichotomous branching, and the same is also seen in Ferns. In the Equisetaceae a mode of branching occurs çeen nowhere else in the Vegetable Kingdom. The branches arise from deeply-seated lateral buds, which, upon developing, break through the superincumbent tissue; these have been termed endogenous buds. In Characeae the branching is of the normal monopodial type.

Amongst Thallogens, as a rule, there is no evident distinction between stem and leaf. The vegetative organs consist of a number of cells separate or combined into a more or less flattened expansion termed the thallus. In this group are included all the Algae, Fungi, and Lichens, and many Hepaticae. The vegetative structures vary much in each of these groups. In many Hepaticae, as Marchantia, the vegetative structure is of a thalloid nature, and appears as a flattened, broad, dichotomously-branched plate, composed of several layers of cells. The upper and lower layers consist of colourless cells, and form a sort of epidermis. The upper epidermal layer has stomata, while the lower gives off root-like filaments, which attach it to the stratum on which it grows. The stomata are, by repeated division of the guard-cells, converted into canals. The central layer of cells contains green matter, and between the cellules there are large intercellular spaces, which communicate with the stomata on the upper surface. In what are usually termed the foliose Hepaticae, such as Junger-manniae, a distinct stem is visible bearing leaves. All are cellular, no differentiation occurring beyond a compacting of the peripheral layers of cells. The mode of branching is usually dichotomous.

In Fungi the vegetative organs consist of a mass of cellular filaments, anastomosing and branching in various ways, constituting the mycelium or spawn (fig. 95, wi.), from which diversely-shaped portions of cellular tissue are developed, forming receptacles for the support of the reproductive organs. The mycelium creeps along in or upon the stratum whence it is nourished, sending up here and there a receptacle. The structural elements which form these are colourless cellular filaments, or lvyphoe, divided usually by many transverse septa. These sometimes combine to form a dense mass of parenchyma. In some cases the hyphae of the mycelium are densely interwoven, and form masses of definite shape, the outer layer of which (pseudo-parenchymatous) forms a hard shell or skin; such a mass is termed a sclerotium. The reproductive receptacle formed from the mycelium is also composed of hyphae. When there is only a single hypha the reproductive bodies are borne on the ends of its branches; but more usually the branches of the hyphae bearing the reproductive bodies unite to form a flattened expansion, the hymenium. As they contain no chlorophyll, Fungi take up assimilated matter from other organisms, and therefore are frequently parasites.

In Algae the simplest forms present nothing but a cellwall, containing a coloured protoplasmic substance, as in some Nostocaceae. In some cases firmness is attained by deposition of silica as in Diatoms, or by lime carbonate as in Acetabularia. Lignification never occurs. More usually, however, the cells are combined into a tissue, but the forms which this may assume and the modes of combination are more various than in any other class of plants,—linearmasses, strings, globes, laminae, <fcc., being formed. There may be in some cases a slight differentiation into distinct organs. Thus, in some of the higher forms, as Fucaceae, a distinct stem is formed, from which flattened thalli resembling leaf structures arise, and at the base root-like structures (rhizoids) are formed. These parts, however, have none of the internal structures which characterize higher plants, but consist entirely of cellular tissue. At most there is a condensation of cellular tissue at the periphery, forming an epidermis, and a similar condensation in the axis. The mode of branching amongst Algae is either monopodial or dichotomous.

In Lichens the thallus (fig. 96) consists of a hyphal element of anastomosing and interlacing filaments, amongst which are distributed rounded unicellular coloured bodies, the gonidia (fig. 97, g). These gonidia are either arranged in one layer in the thallus, which is then said to be heteromerous (the hypha layer being divided into a cortical and deeper part), or they are scattered equally throughout its substance and are homoeomerous. The gonidia are separable, and may form independent plants. When they are being detached, the separation begins at the centre of the thallus, so that the middle of the Lichen becomes pulverulent, while its circumference may remain foliaceous or crustaceous. By the continuation of this process it sometimes happens that the whole Lichen becomes a mass of greenish or yellowish powder. The thallus may be crustaceous, in the form of an incrustation upon rocks, trees, &c., with which it becomes closely connected, as in Graphis; or it may be foliaceous, forming a flattened expansion easily detachable from the substratum to which it is connected by small rootlike bodies or rhizines, as in Peltigera; or the lichen may be what is termed fruticose, i.e., composed of a much-branched thallus rising from a single point of attachment, as in Usnea. Recent observations upon Lichens tend to show that they are in reality composed of two distinct organisms,—one, a Fungus, being the hyphal part, and the other, an Alga, the gonidial portion, and that the Fungus is parasitic upon the gonidia.

3. Leaves and their Modifications.

In popular language all the green expanded organs borne upon an axis are designated leaves. Investigation, however, has shown that many other parts of a plant which externally appear very different from ordinary leaves are, in their essential particulars, very similar to them, and are in fact their morphological equivalents. Thus the scales on the bulb of the Onion, the various parts of the flower, &c., are all leaves. Assuming, then, that the structure ordinarily termed a leaf is the typical form, these latter are designated changed or metamorphosed leaves; and all structures morphologically equivalent with the leaf are included under the general term phyllome (leaf-structure). Leaves are produced as lateral outgrowths of the stem. This character, common to all leaves, distinguishes them from other organs. In the higher plants we can easily recognize the distinction between stem and leaf. Amongst the lower plants, however, it is found that a demarcation into stem and leaf is impossible, but that there is a structure which partakes of the characters of both,— such is a thallome. The leaves always arise from the outer portion of the primary meristem of the plant, and the tissues of the leaf are continuous with those of the stem. Every leaf originates as a simple cellular papilla, which consists of a development from the cortical layers covered by epidermis; and as growth proceeds, the fibro-vascular bundles of the stem are continued outwards, and finally expand and terminate in the leaf. The increase in length of the leaf by growth at the apex is usually of a limited nature. In some Ferns, however, there seems to be a provision for indefinite terminal growth, while in others this growth is periodically interrupted. It not un frequently happens, especially amongst Monocotyledons, that after growth at the apex has ceased, it is continued at the base of the leaf, and in this way the length may be much increased. Amongst Dicotyledons this is very rare. In all cases the dimensions of the leaf are enlarged by interstitial growth of it 3 parts.

The simplest leaf is found in some Mosses, where it consists of a single layer of cells. Usually it consists of several layers, and amongst vascular plants is distinguishable into an epidermis and a central parenchyma with fibro-vascular bundles distributed through it.

The epidermis (fig. 98, es, ei), composed of cells more or less compressed, has usually a different structure and aspect on the two surfaces of the leaf. It is chiefly on the epidermis of the lower surface (fig. 98, ei) that stomata, st, are produced, occupying spaces between the veins, and it is there also that hairs, p, usually occur. The lower epidermis is often of a dull or pale-green colour, soft, and easily detached. The upper epidermis is frequently smooth and shining, and sometimes becomes very hard and dense. Many tropical plants present on the upper surface of their leaves several layers of compressed epidermal cells. These appear to be essential for the preservation of moisture in the leaf. In leaves which float upon the surface of water, as those of the Water-lily, the upper epidermis alone possesses stomata.

The parenchyma of the leaf is the cellular tissue surrounding the vessels, and enclosed within the epidermis (fig. 98, ps, pi). It has sometimes received the names of diachyma, mesophyllum, and diploë. It is formed of two distinct series of cells, each containing chlorophyll or green-coloured granules, but differing in form and arrangement. Below the epidermis of the upper side of the leaf there are one or two layers of oblong blunt cells, placed perpendicularly to the surface (fig. 98, p>s), and applied so closely to each other as to leave only small intercellular spaces (fig. 98, m), except where stomata happen to be present. On the under side of the leaf the cells are irregular, often branched, and are arranged more or less horizontally (fig. 98, pi), leaving cavities betweeu them, I, which often communicate with stomata. On this account the tissue has received the name of cavernous. In leaves having a very firm texture, as those of Coniferae and Cycadaceae, the cells of the parenchyma immediately beneath the epidermis are very much thickened and elongated in a direction parallel to the surface of the leaf, so as to be fibre-like. These constitute a hypodermal layer, beneath which the chlorophyll cells of the parenchyma are densely packed together, and are elongated in a direction vertical to the surface of the leaf; this has been termed palisade tissue. The form and arrangement of the cells, however, depend much on the nature of the plant, and its exposure to light and air. Sometimes the arrangement of the cells on both sides of the leaf is similar, as occurs in leaves which have their edges presented to the sky. In very succulent plants the cells form a compact mass, and those in the centre are often colourless. In some cases the cellular tissue is deficient at certain points, giving rise to distinct holes in the leaf, as in Monstera Adansonii; such a leaf has been called pertuse. In Victoria regia perforations in the leaf seem to be subservient to the purposes of nutrition, by permitting the gases collected beneath the large expanded leaf to escape, and thus allowing its under surface to be brought into immediate contact with the water. The fibro-vascular system in the leaf constitutes the venation. The fibro-vascular bundles from the stem bend out into the leaf, and are there arranged in a definite manner. They usually form two layers, which may be separated by maceration. In skeleton leaves, or leaves in which the parenchyma is removed, these layers are well seen. In some leaves, as in the Barberry, the vessels forming the veins are hardened, producing spines without any parenchyma. The hardening of the extremities of the vascular tissue is the cause of the spiny margin of many leaves, such as the Holly, of the sharp-pointed leaves of Madder, and of mucronate leaves, or those having a blunt end with a hard projection in the centre.

Submerged leaves, or leaves which are developed under water, differ in structure from aerial leaves. They have usually no fibro-vascular system, but consist of a congeries of cells, which sometimes become elongated and compressed so as to resemble veins. They have a layer of compact cells on their surface, but no true epidermis, and no stomata. Their internal structure consists of cells, disposed irregularly, and sometimes leaving spaces which are filled with air for the purpose of floating the leaf. When exposed to the air these leaves easily part with their moisture, and become shrivelled and dry. In some instances there is only a network of filamentous-like cells formed, the spaces between which are not filled with parenchyma, giving a peculiar skeleton appearance to the leaf, as in Ouvirandra fenestralis (Lattice plant). Such a leaf has been called fenestrate. A leaf, whether aerial.or submerged, generally consists of a flat expanded portion (fig. 99, I), called the blade, limb, or lamina, of a narrower portion called the petiole or stalk (fig. 99, p), and sometimes of a portion at the base of the petiole, which forms a sheath or vagina (fig. 99, y), or is developed in the form of leaflets, called stipules (fig. 123, s). The sheathing portion is sometimes incorporated with the stem, and is then called tigellary. These portions are not always present. The sheathing or stipulary portion is frequently wanting, and occasionally only one of the other two is developed. When a leaf has a distinct stalk it is called petiolate; when it has none, it is sessile, and if in this case it embraces the stem it is said to be amplexicaul. The part of the leaf next the petiole or the axis is the base, while the opposite extremity is the apex. The surfaces of the leaf are called the pagin<x, and its edges or margins form the circumscription of the leaf. The leaf is usually flattened and expanded horizontally, i.e., at right angles to the longitudinal axis of the shoot, so that the upper pagina is directed towards the heavens, and the lower pagina towards the earth. In some cases leaves, as in Iris, or leaf-like petioles, as in Australian Acacias and Eucalypti, have their plane of expansion parallel to the axis of the shoot; or the leaf may have a cylindrical or polyhedral form, as in Mesembryanthemum. In other instances, as in Aiströmeria, the leaf becomes twisted in its course, so that what is superior at one part becomes inferior at another. The upper angle formed between the leaf and the stem is called its axil, and everything arising at that point is called axillary. It is there that leaf-buds are usually developed. The leaf is sometimes articulated with the stem, and, when it falls off, a scar or cicatricula remains; at other times it is continuous with it, and then decays, while still attached to the axis. In their early state all leaves are continuous with the stem, and it is only in their after growth that articulations are formed. When leaves fall off annually they are called deciduous; when they remain for two or more years they are persistent, and the plant is called evergreen. The laminar portion of a leaf is occasionally articulated with the petiole, as in the Orange, and a joint at times exists between the vaginal or stipulary portion and the petiole.

The arrangement of the fibro-vascular system in the lamina constitutes the venation or nervation. In an ordinary leaf, as that of the Elm, there is observed a large central vein running from the base to the apex of the leaf, this is the m'(Zrιδ (fig. 100); it gives off veins laterally (pri mary veins'), which either end in a curvature within the margin (curve-veined), as in Lilac and Belladonna, or go directly to the edge of the leaf (feather-veined), as in Elm (fig. 100) and Chestnut. If they are curved, then external veins and marginal veinlets are interspersed through the parenchyma external to the curvature. There are also other veins of less extent (costal veins) given off by the midrib, and these give origin to small veinlets. A leaf with only a single midrib is said to be unicostate (fig. 100). In some cases, as Sycamore and Cinnamon, in place of there being only a single midrib there are several large veins (ribs) of nearly equal size, which diverge from the point where the blade joins the petiole or stem, giving off lateral veins. The leaf in this case is multicostate (fig. 101). When there are three prominent ribs, as in Cinnamon and Cassia, the leaf is tricostate; when five, quinque cost ate. When the midrib gives off two ribs a little above the base, the leaf becomes triplicostate; when it gives off five, quintuplicostate.

In a leaf having many ribs they may converge towards the apex, as in Cinnamon, or they may diverge, as in Sycamore and the Castor-oil plant (fig. 101). Thus the primary veins give off secondary veins, and these in their turn give off tertiary veins, and so on until a complete network of vessels is produced, and those veins usually project on the under surface of the leaf. To a distribution of veins such as this the name of reticulated or netted venation has been applied. In the leaves of some plants there exists a midrib with large veins running nearly parallel to it from the base to the apex of the lamina, as in Grasses (fig. 102); or with veins diverging from the base of the lamina in more or less parallel lines, as in Fan Palms (fig. 103), or with veins coming off from it throughout its whole course, and running parallel to each other in a straight or curved direction towards the margin of the leaf, as in Plantain and Banana. In these cases the veins are often united by cross veinlets, which do not, however, form an angular network. Such leaves are said to be parallel-veined. The leaves of Monocotyledons have generally this kind of venation, while reticulated venation most usually occurs amongst Dicotyledons. Some plants, which in most points of their structure are Monocotyledonous, yet have reticulated venation. Such have been called Dictyogens. In vascular Acotyledonous plants there is frequently a tendency to fork exhibited by the fibro-vascular bundles in the leaf; and when this is the case we have fork-veined leaves. This is well seen in many Ferns. The distribution of the system of vessels in the leaf is usually easily traced, but in the case of succulent plants, as Hoya, Agave, Stonecrop, and Mesembryanthemum, the veins are obscure, and the leaves are said to be hidden-veined. In the cellular leaves of Mosses a median vein of several layers of cells is often visible, but as there are no fibro-vascular bundles present this is considered a false venation, and they are styled veinless {Avenid). [table] [table]

In all plants, except Thallophytes, leaves are present at some period of their existence. In Cuscuta (Dodder), however, we have an exception. The forms assumed by leaves vary much, not only in different plants but in the same plant. It is only amongst the lower classes of plants,—Mosses, Characeøe, Ac.,—that all the leaves on a plant are similar. As we pass up the scale of vegetable life we find them becoming more and more variable. The, structures in ordinary language designated as leaves are! considered so par excellence, and they are frequently spoken' of as foliage leaves. In relation to their production on the stem we may observe that when they are small they are always produced in great number, and as they increase in size their number and rapidity of growth diminish correspondingly. The cellular process from the axis which develops into a leaf is simple and undivided; it rarely remains so, but in progress of growth becomes segmented, in various ways, either longitudinally 7 or laterally, or in both ways. By longitudinal segmentation we have a leaf formed consisting of vagina, petiole, and lamina; or one or other of these may be absent, and thus stalked, sessile, sheathing, &c., leaves are produced. Lateral segmentation affects the lamina, producing indentations, lobings, or fissuring of its margins. In this way two marked forms of leaf are produced—(1.) Simple form, in which the segmentation, however deeply it extends into the lamina, does not separate portions of the lamina which become articulated with the midrib or petiole; and (2.) Compound form, where portions of the lamina are separated as detached leaflets {foliola}, which become articulated with the midrib or petiole. In simple leaves, then, there is never more than one articulation, which is at the point of their insertion on the stem. In compound leaves there are one or more articulations beyond the point of insertion on the stem. In both simple and compound leaves, according to the amount of segmentation and the mode of development of the parenchyma and direction of the fibro-vascular bundles, many forms are produced.

Simple Leaves.— When the parenchyma is developed symmetrically on each side of the midrib or stalk, the leaf is equal; if otherwise, the leaf is unequal or oblique (fig. 100). If the margins are even and present no divisions, the leaf is entire (fig. 104); if there are slight projections of cellular or vascular tissue beyond the margin the leaf is not entire; when the projections are irregular and more or less pointed, the leaf is dentate or toothed; when they lie regularly over each other, like the teeth of a saw, the leaf is serrate (fig. 100); when they are rounded the leaf is crenate. If the divisions extend more deeply into the lamina than the margin, the leaf receives different names according to the nature of the segments; thus, when the divisions extend about half-way down (fig. 105), it is cleft (fissus}, and its lines of separation are called fissures: when the divisions extend nearly to the base or to the midrib the leaf is partite, and its lines of separation are called partitions.

If these divisions take place in a simple feather-veined leaf it becomes either pinnatifid (fig. 106), when the segments extend to about the middle and are broad; or pectinate, when they are narrow; or pinnatipartite, when the divisions extend nearly to the midrib. These primary divisions may be again subdivided in a similar manner, and thus a feather-veined leaf will become bipinnatifid or bipinnatipartite; and still further subdivisions give origin to tripinnatifid and laciniated leaves. If the divisions of a pinnatifid leaf are more or less triangular, and point downwards towards the base, the extremity of the leaf being undivided and triangular, the leaf is runcinate (fig. 105), as in the Dandelion. When the apex consists of a large rounded lobe, and the divisions, which are also more or less rounded, become gradually smaller towards the base, as in Barbarea, the leaf is called lyrate, from its resemblance to an ancient lyre. When there is a concavity on each side of a leaf, so as to make it resemble a violin, as in Rumex pulcher, it is called panduriform. The same kinds of divisions taking place in a simple leaf with radiating venation, give origin to lobed, cleft, and partite forms. When the divisions extend about half-way through the leaves, they may be three-lobed, five-lobed, seven-lobed, many-lobed, or trifid, quinquefid, septemfid, multifid, according to the number of the divisions. The name palmate, or palmatifid (fig. 101), is the general term applied to leaves with radiating venation, in which there are several lobes united by a broad expansion of parenchyma, like the palm of the hand, as in the Castor oil plant, Rheum palmatum, and Papaw. The divisions of leaves with radiating venation may extend to near the base of the leaf, and the names bipartite, tripartite, quinquepartite (fig. 107), septempartite, digitipartite (Plate II. fig. 1), are given according as the partitions are two, three, five, seven, or more. In Drosera dichotoma bipartite and tripartite leaves are seen. The term dissected is applied to leaves with radiating venation, having numerous narrow divisions, as in Geranium dissectum.

When in a radiating leaf there are three primary partitions, and the two lateral lobes are again cleft, as in Helleborus

(fig. 108), the leaf is called pedate or pedatifid, from a fancied resemblance to the claw of a bird. In all the instances already alluded to the leaves have been considered as flat expansions, in which the ribs or veins spread out on the same plane with the stalk. In some cases, however, the veins spread at right angles to the stalk. If they do so equally on all sides, and are united by parenchyma, so that the stalk occupies the centre, the leaf becomes orbicular, as in Hydrocotyle; if unequally, so that the stalk is not in the centre, the leaf is peltate, as in Indian Cress (fig. 109). The edges or margins of orbicular and peltate leaves are often variously divided.

Without attempting to notice all the forms of leaves, the following are enumerated as the most important. When the veins do not spread out, but run from the base to the apex with a narrow strip of parenchyma, the leaf is linear or acicular, as in Pines and Firs. These trees are hence called in Germany Nadel-hölzer, or needle trees. When the veins diverge, those in the middle being longest, and the leaf tapering at each end (fig. 110), it becomes lanceolate. If the middle veins exceed the others slightly, and the ends are convex, the leaf is either rounded, elliptical, oval, or oblong (fig. 111). If the veins at the base are longest, and the leaf narrows to the top, it is ovate or egg-shaped (fig. 104), as in Chickweed; if the apex is broadest, the leaf is obovate, or inversely egg-shaped (fig. 112). Leaves are cuneate or wedge-shaped, in Saxifraga; spathulate, or spatula-like, having a broad rounded apex, and tapering down to the stalk, as in the Daisy; subulate, narrow and tapering like an awl; acuminate, or drawn out into a long point, as in Ficus religiosa; mucronate, with a hard stiff point or mucro at the apex. When the parenchyma is deficient at the apex so as to form two rounded lobes, the leaf is obcordate or inversely heart-shaped; when the deficiency is very slight the leaf is called emarginate (fig. 112), as having a portion taken out of the margin; when the apex is merely flattened or slightly depressed the leaf is retuse; and when the apex ends abruptly in a straight margin, as in the Tulip-tree, the leaf is truncate. When the veins at the base of the lamina are prolonged downwards at an obtuse angle with the midrib, and rounded lobes are formed, as in

Dog-violet, the leaf is cordate or heart-shaped; it is kidneyshaped or reniform (fig. 113), when the apex is rounded as in Asarum. When the lobes are prolonged downwards and are acute, the leaf is sagittate (fig. 114); when they proceed at right angles, as in Rumex Acetosella, the leaf is hastate or halbert-shaped. When a simple leaf is divided at the base into two leaf-like appendages, it is called auriculate. When the veins spread out in various planes, and there is a large development of cellular tissue, so as to produce a succulent leaf, such forms occur as conical, prismatical, ensiform or sword-like, acinaciform or scimitar-shaped, and dolabriform or axeshaped. When the development of parenchyma is such that it more than fills up the spaces between the veins, the margins become wavy, crisp, or undulated, as in Rumex crispus and Rheum undulatum. By cultivation the cellular tissue is often much increased, giving rise to the curled leaves of Greens, Savoys, Cresses, Lettuce, &c. In Rushes the shoots which act as leaves are often terete. They are either barren or bear flowers. Their cellular tissue is often stellate, and the shoots sometimes exhibit a peculiar spiral twisting. Amongst parallel veined leaves the margins are usually entire, especially when the veins converge.

Compound leaves are those in which the divisions extend to the midrib or petiole, and the separated portions become each articulated with it, and receive the name of foliola or leaflets. The midrib, or petiole, has thus the appearance of a branch with separate leaves attached to it, but it is considered properly as one leaf, because in its earliest state it arises from the axis as a single piece, and its subsequent divisions in the form of leaflets are all in one plane. The leaflets are either sessile (fig. 115), or have stalks, called petiolules (fig. 116), according as the vascular bundles of the veins spread out or divaricate at once, or remain united for a certain length. Compound leaves have been classified according to the nature of the venation and the development of parenchyma, If we suppose that in a simple feather-veined unicostate leaf the divisions extend to the midrib, and each of the primary veins spreads out or branches, and becomes surrounded with parenchyma, and the leaflets thus formed become articulated to the pe tiole or midrib (fig. 116), the leaf becomes compound and pinnate. If the midrib and primary veins are not covered with parenchyma, while the secondary (or those coming off in a feather-like manner from the primary veins) are, and separate leaflets are thus formed which are articulated with the veins, the leaf is bipinnate. In this case the secondary veins form as it were partial petioles. A farther subdivision, in which the tertiary veins only are covered with parenchyma and have separate leaflets, gives tripinnate or decompound forms, in which case the tertiary veins form the partial petioles. A leaf divided still more is called supradecompound. When a pinnate leaf has one pair of leaflets it is unijugate; with two pairs, it is bijugate; with many pairs, multijugate (fig. 116). When a pinnate leaf ends in a pair of pinnae it is equally or abruptly pinnate (pari-pinnate); when there is a single terminal leaflet (fig. 116), the leaf is unequally pinnate (impari-pinnate); when the leaflets or pinnae are placed alternately on either side of the midrib, and not directly opposite to each other, the leaf is alternately pinnate; and when the pinnae are of different sizes, the leaf is interruptedly pinnate. In the case of a simple multicostate leaf with radiating venation, if we suppose the ribs to be covered with parenchyma, so as to form separate leaflets, each of which is articulated to the petiole, the digitate form of compound leaf is produced; if there are three leaflets the form is ternate or trifoliate; if four, quaternate; if five, quinate; if seven, septenate; and so on. If the three ribs of a ternate leaf subdivide each into three primary veins, which become covered with parenchyma so as to be separate articulated leaflets, the leaf is biternate; and if another threefold division takes place, it is triternate.

The petiole or leaf-stalk is the part which unites the limb or blade of the leaf to the stem (fig. 99, p). It is absent in sessile leaves, and this is also frequently the case when the vagina is present, as in Grasses. It consists of a mass of fibro-vascular bundles with a varying amount of cellular tissue. The vessels are enclosed in an epidermal covering, with few stomata, and are more or less compressed. When the vascular bundles reach the base of the lamina they separate and spread out in various ways, as already described under venation. At the place where the petiole joins the stem there is frequently an articulation, or a constriction with a tendency to disunion, and at the same time there exists a swelling (fig. 117, p), called pulvinus, formed of cellular tissue, the cells of which exhibit the phenomenon of irritability (figs. 117, 118). At the point where the petiole passes into the lamina, or where the midrib joins the leaflets of a compound leaf, there is occasionally a cellular dilatation called struma, and an articulation. In Mimosa pudica a sensitiveness is located in the pulvinus (fig. 117, p), which upon irritation induces a depression of the whole bipinnate leaf; a similar property exists in the struma at the base of the leaflets which fold upwards. In Desmodium gyrans (fig. 118) the pulvinus and struma have a similar power of causing movements of the large terminal leaflet and the two smaller lateral ones. In other cases the petiole is not articulated, but is continuous either with the stem, or with the sheath (vagina). The articulation or joint is by many considered as indicating a compound leaf, and hence the leaf of the Orange is considered as such, although it has an undivided lamina (fig. 119, I). In articulated leaves, the pulvinus may be attached either to the petiole or to the axis, and may fall with the leaf, or remain attached to the stem. When articulated leaves drop, their place is marked by a cicatrix or scar, seen below the bud in fig. 63. In this scar the remains of the vascular bundles c are seen; and its form furnishes characters by which particular kinds of trees may be known when not in leaf. In the case of many Palms and Tree-ferns the scars or cicatrices of the leaves are very conspicuous. In fossil plants important characters are founded on them.

When there is no articulation between the petiole and the stem, as is the case with many Monocotyledons, the leaf is continuous with the axis, and is not deciduous, but withers on the stalk. In many Liliaceous plants the leaves during their decay continue attached to the plants. The petiole varies in length, being usually shρrter than the lamina, but sometimes much longer. In some Palms it is 15 or 20 feet long, and is so firm as to be used for poles or walking-sticks. In general, the petiole is more or less rounded in its form, the upper surface being flattened or grooved. Sometimes it is compressed laterally, as in the Aspen, and to this peculiarity the trembling of the leaves of this tree is attributed. In aquatic plants the leaf-stalk is sometimes distended with air, as in Pontederia and

Trapa, so as to float the leaf. At other times it is winged, and is either leafy, as in the Orange (fig. 119, p}, Lemon, and Dionaea (fig. 46, p), or pitcher-like, as in the Pitcher-plant

(fig. 120, p~) and Sarracenia (fig. 121). Leafy petioles are occasionally united to the axis for some extent, and thus become decurrent. In some Australian Acacias, and in some species of Oxalis and Bupleurum, the petiole is flattened in a vertical direction, the vascular bundles separating immediately after quitting the stem, and running nearly parallel from base to apex. This kind of petiole (fig. 122, p) has been called phyllodium. In these plants the laminae or blades of the leaves are pinnate, bipinnate, or ternate, and are produced at the extremities of the phyllodia in a horizontal direction; but in many instances they are not developed, and the phyllodium serves the purpose of a leaf. Hence some Acacias are called leafless. These phyllodia, by their vertical position and their peculiar form, give a remarkable aspect to vegetation. On the same Acacia there occur leaves with the petiole and lamina perfect; others having the petiole slightly expanded or winged, and the lamina imperfectly developed; and others in which there is no lamina, and the petiole becomes large and broad. Some petioles, in place of ending in a lamina become changed into a delicate filiform body, a tendril or cirrus, so as to enable the plant to climb. In many leaves interposed betwixt the petiole and the stem, or, in sessile leaves, betwixt the lamina and the stem, an expansion of the foliar tissue takes place, so that a sheath (vagina) is formed, which embraces the whole or part of the circumference of the stem (fig. 102, gv). This sheath is comparatively rare in Dicotyledons, but is seen in Umbelliferous plants, where it constitutes the pericladium, and in the Rhubarb order, where it is large and membranous, and has received the name of ochrea or boot. It is much more common amongst Monocotyledons. In Palms it forms a kind of network, to which the name of reticulum has been given. In Sedges the sheath forms a complete investment of the stem, whilst in Grasses (fig. 102) it is split on one side. In the latter plants there is also a membranous outgrowth at right angles to the median plane of the leaf from the point where the sheath passes into the lamina (there being no petiole). To this structure the name of ligule (fig. 102, gl) lias been given. It is of various dimensions, and thus gives a character to the plants.

In leaves in which no sheath is produced we not unfrequently find small foliar organs at the base of the petiole. These have been denominated stipules (fig. 123, s). The stipules are often two in number, and they are important as supplying characters in certain natural orders. Thus they occur in the Pea and Bean family, in Rosaceous plants, and the Cinchona bark family. They are not common in Dicotyledons with opposite leaves. Plants having stipules are called stipulate; those having none are exstipulate. Stipules are formed by some of the vascular bundles diverging as they leave the stem, and becoming covered with parenchyma, so as to resemble true leaves. Like leaves they are large or small, entire or divided, deciduous or persistent, articulated or non-articulated. They are not usually of the same form as the ordinary foliage leaves of the plant, from which they are distinguished by their lateral position at the base of the petiole. In the Pansy (fig. 123) the true leaves I are stalked and crenate, while the stipules 3 are large, sessile, and pinnatifid. In Lathyrus Aphaca, and some other plants, the true pinnate leaves are abortive, the petiole forms a tendril, and the stipules alone are developed, performing the office of leaves. When stipules are attached separately to the stem at the base of the leaf, they are called caulinary, as in Salix aurita. When stipulate leaves are opposite to each other, at the same height on the stem, it occasionally happens that the stipules on the two sides unite wholly or partially, so as to form an interpetiolary or interfoliar stipule, as in Cinchona and in Ipecacuan. In the case of alternate leaves, the stipules at the base of each leaf are sometimes united to the petiole and to each other, so as to form an adnate, adherent, or petiolary stipule, as in the Rose, or an axillary stipule, as in Houttuynia cordata. In other instances the stipules unite together on the side of the stem opposite the leaf, and become synochreate, as in Astragalus. This so-called unionor adhesion of stipules is not an accidental adhesion taking place after they have been developed. In these cases the parts never were separate; from the first they are developed as one portion. In the development of the leaf the stipules frequently play a most important part. They begin to be formed after the origin of the leaves, but grow much more rapidly than the leaves, and in this way they arch over the young leaves and form protective chambers wherein the parts of the leaf may develop. In Ficus, Magnolia, and Potamogeton they are very large and completely envelop the young leaf-bud. The stipules are sometimes so minute as to be scarcely distinguishable without the aid of a lens, and so fugacious as to be visible only in the very young state of the leaf. They may assume a hard and spiny character, as in Robinia Pseudacacia, or maybe cirrose, as in Smilax, where each stipule is represented by a tendril; while in Cucurbitaceae there is only one cirrose stipule. At the base of the leaflets or foliola of a compound leaf, small stipules are occasionally produced, to which some have given the name of stipels.

Variations in the structure and forms of leaves and leaf-. stalks are produced by the increased development of cellular ι tissue, by the abortion or degeneration of parts, by the multiplication or repetition of parts, and by adhesion. When cellular tissue is developed to a great extent, leaves become succulent and occasionally assume a crisp or curled appearance. Such changes take place naturally, but they are often increased by the art of the gardener, and the object of many horticultural operations is to increase the bulk and succulence of leaves. It is in this way that Cabbages and Savoys are rendered more delicate and nutritious. By a deficiency in development of parenchyma and an increase in the fibro-vascular tissue, leaves are liable to become hardened and spinescent. The leaves of Barberry and of some species of Astragalus, and the stipules of the False Acacia (Robinia) are spiny. To the same cause is attributed the spiny margin of the Holly-leaf. In the Gooseberry, the swelling (pulvinus) at the base of the petiole, and below the leaf, assumes a spinose character. Changes in the appearances of leaves are produced by adhesions and foldings of various kinds. When two lobes at the base of a leaf are prolonged beyond the stem and unite (fig. 124), the leaf is perfoliate, the stem appearing to pass through it, as in Bupleurum perfoliatum and Chlora perfoliata; when two leaves unite by their bases they become connate (fig. 125), as in Lonicera Caprifolium; and when leaves adhere to the stem, forming a sort of winged or leafy appendage, they are decurrent, as in Thistles. The formation of peltate and orbicular leaves has been traced to the union of the lobes of a cleft leaf. In the leaf of the Victoria regia the transformation may be traced during germination. The first leaves produced by the young plant are linear, the second are sagittate and hastate, the third are rounded-cordate, and the next are orbicular. The cleft indicating the union of the lobes remains in the large leaves. The parts of the leaf are frequently transformed into tendrils (cirri), with the view of enabling the plants to twine round others for support. In Leguminous plants (the Pea tribe) the pinnae are frequently cirrose. When tendrils occupy the place of leaves, and appear as a continuation of the leaf-stalk, they are called petiolary, as in Lathyrue Λphaca, in which the stipules perform the function of true leaves. In Flagellaria Indica, Gloriosa superba, Anthericum cirrhatum, and Albuca cirrhata, the midrib of the leaf ends in a tendril. In Smilax there are two stipulary tendrils, while in the Cucumber tribe there is a single one at the base of each leaf. In the Passion-flower the lateral leaf-buds, and in the Vine the terminal ones, become tendrils. In the Vanille plant ( Vanilla aromatica) the tendrils are produced opposite the leaves, until the plant gains the top of the trees by which it is supported,—the upper tendrils beiDg then developed as leaves. The midrib is sometimes prolonged in a cup-like or funnel-shaped form; this is occasionally seen in the common Cabbage, and seems to depend on the vascular bundles of the midrib spreading out at their extremity in a radiating manner, and becoming covered with parenchyma in such a way as to form a hollow cavity in the centre.

The vascular bundles and cellular tissue are sometimes developed in such a way as to form a circle, with a hollow in the centre, and thus give rise to what are called fistular or hollow leaves, as in the Onion, and to ascidia oτ pitchers. Pitchers are formed either by petioles or by laminae, and they are composed of one or more leaves. In Sarracenia (tig. 121) and Heliamphora, the pitcher is composed apparently of the petiole of the leaf. In Nepenthes (fig. 120) and perhaps in Cephalotus, while the folding of a winged petiole p forms the pitcher, the lid I, which is united by an articulation, corresponds to the lamina. This kind of ascidium is called calyptrimorphous, and may be considered as formed by a leaf, such as that of the Orange (fig. 119); the lamina I being articulated to the petiole p, which, when folded, forms the pitcher. In Dischidia Rafflesiana, a climbing plant of India, the pitchers are formed by the lamina of the leaf, and have an open orifice into which the rootlets at the upper part of the plant enter. These pitchers would seem, therefore, to contain a supply of fluid for the nourishment of the upper branches of the plant. In Utricularia the leaves form sacs called ampullae. Some suppose that pitchers are not due to folding and adhesion, but that they are produced by a hollowing out of the extremity of the stalk. In some cases the leaves are reduced to mere scales; they are then frequently called cataphyllary leaves. They are produced abundantly upon underground shoots. In parasites (Lathraea, Orobanche), and in plants growing on decaying vegetable matter (Saprophytes), in which no chlorophyll is formed, these scales are the only leaves produced. In Pinus the only leaves produced on the main stem and the lateral shoots are scales, the acicular leaves of the tree growing from axillary shoots. In Cycas whorls of scales alternate with large pinnate leaves. In mauy plants, as already noticed, phyllodia or stipules perform the function of leaves. The production of leaf-buds from leaves has already been noticed. Such leaves are termed proliferous. In Bryophyllum (fig. 67) this is a common occurrence, and it is met with in many plants of the order Gesneraceae. The leaf of Venus’s Fly-trap (Dionaea muscipula, fig. 46), when cut off and placed in damp moss, with a pan of water underneath and a bell-glass for a cover, has produced buds from which young plante were obtained. Some species of Saxifrage and of Ferns also produce buds on their leaves and fronds. In Nymphaea micrantha buds appear at the upper part of the petiole.

Amongst Dicotyledons we have leaves which present the greatest amount of variation in structure and form. The venation is reticulated. They are frequently articulated, exhibit divisions at their margin, and become truly compound. There are, no doubt, instances in which the veins proceed in a parallel manner, but this will be found to occur chiefly in cases where the petiole may be considered as occupying the place of the leaf. Examples of this kind are seen in Acacias (fig. 122). Dicotyledons rarely have a sheath developed, but stipules are very commonly present.

In Monocotyledons the leaves do not present an angular network of vessels, nor do they, as a rule, exhibit divisions on their margin. Exceptions to this occur in some plants, as Tamus and Dioscorea, which have been called Dictyogens by Lindley, on account of their somewhat netted venation; and in Palms, in which, although the leaves are entire at first, they afterwards become split into various lobes. They are rarely stipulate, and very frequently have a sheath at their base. The petiole is often absent, and a sheathing ligule takes its place. The leaves are often continuous with the stem. In some aquatic Monocotyledons the submerged and floating leaves are narrow, like petioles, while those growing erect above the water expand. This is seen in Sagittaria sagittifolia, in which the erect leaves assume an arrow-like shape.

In Acotyledons the leaves vary much, being entire or divided, stalked or sessile, often feather-veined, occasionally with radiating venation, the extremities of the veins being forked. In Ferns the leaf (frond) is usually stalked, with frequently a much-divided lamina, remarkable for the prolonged growth at the apex, and in the young state usually covered with curious flattened hairs (paleae). In Equisetaceae the leaves are sheathing cylinders embracing successive internodes of the stem, and subsequently splitting at the top into few or many teeth. In Lycopodiaceae the leaves vary from mere rudimentary scales, as in Psiloturn, to the flattened acuminate leaf of Selaginella, with only a single fibr∩-vascular bundle, and then to the more complicated form in Isoetes, in which the longest leaves in the order occur, and these have a sheathing basal part and an upper lamina. On the face of the sheathing part is a depression or fovea, in which rests the sporangium, the margin of it rising as a thin membranous outgrowth, the velum. Above the fovea, and separated from it by a saddle-like ridge, lies a smaller depression the foveola, the lower margin of which forms a lip, the labium, and from its bottom is prolonged beyond the foveola an apiculate membranous structure, which is termed the ligule. In Mosses we have the simplest form of leaf, composed of one or more layers of cells, sometimes the central ones being more or less compacted and forming a median vein.

Leaves occupy various positions on the stem and branches, and have received different names according to their situation. Thus leaves arising from the crown of the root, as in the Primrose, are called radical; those on the stem are cauline; on the branches, ramal; on flower-stalks, floral leaves. The first leaves developed are denominated seminal leaves or cotyledons, and those which succeed are primordial. The arrangement of the leaves on the axis and its appendages is called phyllotaxis. In their arrangement leaves follow a definite order. It has been stated already that there are regular nodes or points on the stem at which leaves appear, and that the part of the stem between the nodes is the internode. Each node is capable of giving origin to a leaf. Occasionally several nodes are approximated so as to form as it were one, and then several leaves may be produced at the same height on the stem. When two leaves are thus produced, one on each side of the stem or axis, and at the same level, they are called opposite (fig. 126); when more than two are produced (fig. 127), they are verticillate, and the circle of leaves is then called a verticil or whorl. When the zone of the axis which produces the circle of leaves is transverse from its origin, the whorl is a true one; but when the zone is the result of unequal development or of displacement the whorl is spurious. Again, in each whorl the leaves may be all formed together, and a simultaneous whorl results; or they may be formed one after the other, as in Characeae, when a successive whorl is formed. When leaves are opposite, each successive pair may be placed at right angles to the pair immediately preceding. They are then said to decussate, following thus a law of alternation. The same occurs in the verticillate arrangement, the leaves of each whorl rarely being superposed on those of the whorl next it, as on the branches of Chara, but usually alternating so that each leaf in a whorl occupies the space between two leaves of the whorl next to it. There are considerable irregularities, however, in this respect, and the number of leaves in different whorls is not always uniform, as may be seen in Lysimachia vulgaris. When a single leaf is produced at a node, and the nodes are separated so that each leaf is placed at a different height on the stem, the leaves are alternate (fig. 128). A plane passing through the point of insertion of the leaf in the node, dividing the leaf into two similar halves, is the median plane of the leaf; and when the leaves are arranged alternately on an axis so that their median planes coincide they form a straight row or orthostichy. On every axis there are usually two or more orthostichies. The leaves in such a case are said to be rectiserial. In fig. 129, leaf 1 arises from a node n; leaf 2 is separated by an internode m, and is placed to the right or left; while leaf 3 is situated directly above leaf 1. In this case, then, there are two orthostichies, and the arrangement is said to be distichous. When the fourth leaf is directly above the first, the arrangement is tristichous. The same arrangement continues throughout the branch, so that in the latter case the 7th leaf is above the 4 th, the 10th above the 7th also the 5th above the 2d, the 6th above the 3d, and so on. The size of the angle between the median planes of two consecutive leaves in an alternate arrangement is their divergence; and it is expressed in fractions of the circumference of the axis which is supposed to be a circle. In a regularly-formed straight branch covered with leaves, if a thread is passed from one to the other, turning always in the same direction, a spiral is described, and a certain number of leaves and of complete turns occur before reach ing the leaf directly above that from which the enumeration commenced. If this arrangement is expressed by a fraction, the numerator of which indicates the number of turns, and the denominator the number of internodes in the spiral cycle, the fraction will be found to represent the angle of divergence of the consecutive leaves on the axis. Thus, in fig. 130, α, b, the cycle consists of five leaves, the 6th leaf being placed vertically over the 1st, the 7th Over the 2d, and so on; while the number of turns between the 1st and 6th leaf is two; henee this arrangement is indicated by the fraction ∙∣. In other words, the distance or divergence between the first and second leaf, expressed in parts of a circle, is ⅜ of a circle, or 360° ÷⅜= 144°. In fig. 129, α, b, the spiral is ⅜, i.e., one turn and two leaves; the third leaf being placed vertically over the first, and the divergence between the first and second leaf being one-half the circumference of a circle, 360° ÷⅜=180°. Again, in a tristichous arrangement the number is ⅜, or one turn and three leaves, the angular divergence being 120°.

By this means we have a convenient mode of expressing on paper the exact position of the leaves upon an axis. And in many cases such a mode of expression is of excellent service in enabling us readily to understand the relations of the leaves. The divergences may also be represented diagrammatically on a horizontal projection of the vertical axis, as in fig. 131. Here the outermost circle represents a section of that portion of the axis bearing the lowest leaf, the innermost represents the highest. The broad dark lines represent the leaves, and they are numbered according to their age and position. It will be seen at once that the leaves are arranged in orthostichies marked I.-V., and that these divide the circumference into five equal portions. But the divergence between leaf 1 and leaf 2 is equal to fths of the circumference, and the same is the case between 2 and 3, 3 and 4, &c. The divergence, then, is ⅜, and from this we learn that, starting from any leaf on the axis, we must pass twice round the stem in a spiral through five leaves before reaching one directly over that with which we started. When the leaves or scales are alternate, and run in a single series, they are unijugate; when the leaves are opposite, and there are two series, the arrangement is bijugate; while in the case of whorled leaves the arrangement may be trijugate or quadrijugate. The line which, winding round an axis either to the right or to the left, passes through the points of insertion of all the leaves on the axis is termed the genetic or generating spiral; and that margin of each leaf which is towards the direction from which the spiral proceeds is the kathodic side, the other margin facing the point whither the spiral passes being the anodic side.

In cases where the internodes are very short, and the leaves are closely applied to each other, as in the House-leek, it is difficult to trace the generating spiral. Thus, in fig. 132 there are thirteen leaves which are numbered in their order, and five turns of the spiral marked by circles in the centre

(⅛ indicating the arrangement); but this could not be detected at once. So also in Fir cones (fig. 133), which are composed of scales or modified leaves, the generating spiral cannot be determined easily. But in such cases a series of secondary spirals or parastichies are seen running parallel with each other both right and left, which to a certain extent conceal the genetic spiral. Thus, in fig. 133, it will be found that there are five secondary spirals running towards the right and parallel to each other, the first passing through the scales 1, 6, 11, 16, <fcc.; the second through 9, 14, 19, 24, &c.; the third through 17, 22, 27, 32, 37, <fcc.; the fourth through 30, 35, 40, 45, &c.; the fifth through 43, 48, 53, &c. The number of these secondary spirals indicates the number of scales intervening between every two scales in each of these spirals, the common difference being five. Again, it will be found on examination that there are secondary spirals running to the left, in which the common difference between every two scales is eight, and that this corresponds to the number of secondary spirals, the first of which passes through the scales 1, 9, 17, <kc.; the second through 6, 14, 22, 30, &c.; the third through 3, 11, 19, 27, 35, 43, and so on. Thus it is that, by counting the secondary spirals, all the scales may be numbered, and by this means the generating spiral may be discovered. From the number of secondary spirals the angle of divergence may be easily calculated, the sum of those which wind in both directions giving the denominator of the fraction, while the smaller of the two numbers representing those winding in each direction is the numerator. Thus in the instance last mentioned the angular divergence is τ δ - j . In the cone of the American larch (fig. 134) there is a quincuncial arrangement of scales marked by the fraction ∙∣. There are five vertical ranks, as marked in the tabular numerical view at the side of the cone, which represents the unwound surface of the cone, viz., 2, 7, 12; 4, 9, 14;

1, 6,11; 3,8, 13; 5, 10,15,—the common difference in each row being 5. On looking at the cone we find also parallel oblique ranks, two of which, ascending to the left, are marked by the numbers 1, 3, 5, which, if the diagram is coiled round a cylinder, continue in the numbers 7, 9, 11, 13, 15; and 2, 4, 6, 8, 10, continued into 12, 14. There are thus two lefthanded spirals, with 2 as the common difference in the numbering of the scales. Again, three oblique parallel spirals ascend to the right, marked by the numbers 1, 4, 7, running into 10, 13; 3, 6, 9, 12, going on to 15; and 5, 8, 11, 14; here the common difference in the numbering of the scales is 3, corresponding with the oblique right-handed spirals.

All the constant divergences found in phyllotaxis may be represented as successive convergents of the continued fraction [table]

where a may have the values 1, 2, 3, 4, <fcc. The actual fractions thus resulting are— [table]

The spiral is not always constant throughout the whole length of an axis. The angle of divergence may alter either abruptly or gradually, and the phyllotaxis thus becomes very complicated. This change may be brought about by arrest of development, by increased development of parts, or by a torsion of the axis. The former are exemplified in many Crassulaceae and Aloes. The latter is seen well in the Screw pine (Pandanus). In the bud of the screw pine the leaves are arranged in three orthostichies with the phyllotaxis ⅜, but by torsion the developed leaves become arranged in three strong spiral rows running round the stem. These causes of change in phyllotaxis are also well exemplified in the alteration of an opposite or verticillate arrangement to an alternate, and vice versa; thus the effect of interruption of growth, in causing alternate leaves to become opposite and verticillate, can be distinctly shown in Rhododendron ponticum. Again, parts which are usually opposite or verticillate become alternate by the vigorous development of the axis, as in Hippuris, and also in Lysimachia vulgaris, where on different parts of the same stem there may be seen alternate, opposite, and verticillate leaves. When the interruption to development takes place at the end of a branch, the leaves become fasciculate or clustered, as in the Larch. The primitive or generating spiral may pass either from right to left or from left to right. It sometimes follows a different direction in the brauches from that pursued in the stem. When it follows the same course in the stem and branches, they are homodromous; when the direction differs, they are heterodromous. In different species of the same genus the phyllotaxis frequently varies.

All modifications of leaves follow the same laws of arrangement as true leaves—a fact which is of importance in a morphological point of view. In Dicotyledonous plants the first leaves produced (the cotyledons) are opposite, in some cases verticillate. This arrangement often continues during the life of the plant, but at other times it changes, passing into distichous and spiral forms. Some tribes of plants are distinguished by their opposite or verticillate, others by their alternate, leaves. Labiate plants have decussate leaves, while Boraginaceae have alternate leaves, and Tiliaceae usually have distichous leaves; Cinchonaceae have opposite leaves; Galiaceae, verticillate. Such arrangements as ⅜, f, 1 ⅛, and ⅛ are common in Dicotyledons. The first of these, called a quincunx, is met with in the Apple, Pear, and Cherry (fig. 130); the second, in the Bay, Holly, Plantago media; the third, in the cones of Pinus (Abies) alba (fig. 133); and the fourth in those of the Pinus (Abies) Picea. In Monocotyledonous plants there is only one seed-leaf or cotyledon produced, and hence the arrangement is at first alternate; and it generally continues so more or less, rarely being verticillate. Such arrangements as ∣, ⅜, and -j are common in Monocotyledons, as in Grasses, Sedges, and Lilies. In Acotyledons the leaves assume all kinds of arrangement, being opposite, alternate, and verticillate. It has been found in general that, while the number 5 occurs in the phyllotaxis of Dicotyledons, 3 is common in that of Monocotyledons.

Tn the axil of previously formed leaves leaf-buds arise. These leaf-buds contain the rudiments of a shoot, and consist of leaves covering a growing point. The buds of trees of temperate climates, which lie dormant during the winter, are protected by cataphyllary leaves constituting the tegmenta or peruloe. These scales or protective appendages of the bud consist either of the altered laminae, or of the enlarged petiolary sheath, or of stipules, as in the Fig and Magnolia, or of one or two of these parts combined. These are often of a coarse nature, serving a temporary purpose, and then falling off when the leaf is expanded. They are frequently covered with a resinous matter, as in Balsampoplar and Horse-chestnut, or by a thick downy covering as in the Willow. In plants of warm climates the buds have often no protective appendages, and are then said to be naked.

The arrangement of the leaves in the bud has been denominated vernation, prefoliation, and gemmation. In considering vernation we must take into account both the manner in which each individual leaf is folded and also the arrangement of the leaves in relation to each other. These vary in different plants, but in each species they follow a regular law. The leaves in the bud are either placed simply in apposition, as in the Mistleto, or they are folded or rolled up longitudinally or laterally, giving rise to different kinds of vernation, as delineated in figs. 135 to 144, where the folded or curved lines represent the leaves, the thickened part being the midrib. The leaf taken individually is either folded longitudinally from apex to base, as in the Tulip-tree, and called reclinate or replicate; or rolled up in a circular manner from apex to base, as in Ferns (fig. 135), and called circinate; or folded laterally, conduplicate (fig. 136), as in Oak; or it has several folds like a fan, plicate or plaited (fig. 137), as in Vine and Sycamore, and in leaves with radiating vernation, where the ribs mark the foldings; or it is rolled upon itself, convolute (fig. 138), as in Banana and Apricot; or its edges are rolled inwards, involute (fig. 139), as in Violet; or outwards, revolute (fig. 140), as in Rosemary. The different divisions of a cut leaf may be folded or rolled up separately, as in Ferns, while the entire leaf may have either the same or a different kind of vernation. The leaves have a definite relation to each other in the bud, being either opposite, alternate, or verticillate; and thus different kinds of vernation are produced. Sometimes they are nearly in a circle at the same level, remaining flat, or only slightly convex externally, and placed so as to touch each other by their edges, thus giving rise to valvate vernation. At other times they are at different levels, and are applied over each other, so as to be imbricated, as in Lilac, and in the outer scales of Sycamore (fig. 63); and occasionally the margin of one leaf overlaps that of another, while it in its turn is overlapped by a third, so as to be twisted, spiral, or contortive. When leaves are applied to each other face to face, without being folded or rolled together, they are oppressed. When the leaves are more completely folded they either touch at their extremities and are accumbent or opposite (fig. 141), or are folded inwards by their margin, and become induplicate; or a conduplicate leaf covers another similarly folded, which in turn covers a third, and thus the vernation is equitant (fig. 142), as in Privet; or conduplicate leaves are placed so that the half of the one covers the half of another, and thus they become half-equitant or obvolute (fig. 143), as in Sage. When in the case of convolute leaves one leaf is rolled up within the other, it is supervolute (fig. 144). The scales of a bud sometimes exhibit one kind of vernation, and the leaves another. The same modes of arrangement occur in the flower-buds, as will be afterwards shown.

Leaves expose the fluids of plants to the influence of air and light. The fluids so exposed are elaborated, and thus fitted for the formation of the various vegetable tissues and secretions. For the proper performance of this function the structure of the leaves, and their arrangement on the stem and branches, render them well adapted. The cells in the lower side of a leaf where stomata exist are chiefly concerned in the aeration of the sap, whilst other assimilative processes go on in the upper cells. The elaboration of fluids in the leaves necessarily implies interchange of their constituents with those of the surrounding atmosphere; hence two processes are inevitable—a passing inwards into the leaf of the atmospheric elements by a process of absorption, and an outward current of the components of the plant-juices by a process of exhalation. The absorption of carbonic acid, water, and other fluids is carried on by the leaves, chiefly through their stomata, and most rapidly by the under surface of ordinary leaves in which the cuticle is thinnest, the cellular tissue least condensed, and stomata most abundant; the upper surface of the leaf, which usually presents a polished and dense epidermis with few stomata, taking little part in such a process. An exhalation of both liquids and gases also takes place from the leaves, regulated by the number and the size of stomata as well as by the nature of the epidermis. The process of transpiration of fluids imparts moisture to the atmosphere, and hence the difference between the air of a wooded country and that of a country deprived of forests. Thus leaves have an important influence upon the climate of a country. In darkness little or no transpiration takes place, and in diffuse daylight it is less than in the sun’s rays. The exhalation of gases constitutes the process of respiration. The nature and amount of the gases respired depends both on the circumstances in which the leaves are placed and on the condition of the plant. But normally at all times there seems to be a respiration of carbonic acid, which, under the influence of light, is at once decomposed by the green parts of the plant, the carbon being fixed and the oxygen set free; consequently, in darkness no oxygen is eliminated. Leaves, after performing their functions for a certain time, wither and die. In doing so they frequently change colour, and hence arise the beautiful and varied tints of the autumnal foliage. This change of colour is chiefly occasioned by the diminished circulation in the leaves, and the higher degree of oxidation to which their chlorophyll has been submitted.

Leaves which are articulated with the stem, as in the Walnut and Horse-chestnut, fall and leave a scar, while those which are continuous with it remain attached for some time after they have lost their vitality, as in the Beech. Most of the trees of Great Britain have deciduous leaves, their duration not extending over more than a few months, while in trees of warm climates the leaves often remain for two or more years. In tropical countries, however, many trees lose their leaves in the dry season. The period of defoliation varies in different countries according to the nature of their climate. Trees which are called evergreen, as Pines and Evergreen-oak, are always deprived of a certain number of leaves at intervals, sufficient being left, however, to preserve their green appearance. The cause of the fall of the leaf in cold climates 8eem3 to be deficiency of light and heat in winter, which causes a cessation in the functions of the cells of the leaf; its fluids disappear by evaporation; its cells and vessels become contracted and diminished in their calibre; various inorganic matters accumulate in the textures; the whole leaf becomes dry; its parts lose their adherence; a process of disjunction takes place by a folding inwards of the tissue at the point where the leaf joins the stem or branch which gradually extends until complete separation takes place, and the leaf either falls by its own weight or is detached by tbe wind. In warm climates the dry season gives rise to similar phenomena.

II. ORGANS OF REPRODUCTION.

We now proceed to pass in review the reproductive organs of plants. In Phanerogamous plauts, as already mentioned, these organs are conspicuous, and constitute what is known as the flower; in Cryptogamous plants they are inconspicuous. All Dicotyledonous and Monocotyledonous plants are included in the former; Acotyledonous and Thallogenous plants compose the latter. The structures which go to form these organs are not, however, formations of a new type, but are merely modifications of those structures which we have already considered under the nutritive organs. For example, the various parts of the flower in Phanerogams are really phyHomes, the supporting structures of the flower are caulomes, the spore-bearing sac of many Cryptogams is a trichome; and in this way a morphological equivalency may be traced betwixt the two series of organs. Further, the difference betwixt the reproductive organs of Phanerogams and those of Cryptogams is one more of degree of differentiation than of actual morphological difference. In Phanerogams the flowers or floral axes are produced from flower-buds, just as leaf-shoots arise from leaf-buds. These two kinds of buds have a resemblance to each other as regards the arrangement and the development of their parts; and it sometimes happens, from injury and other causes, that the part of the axis which, in ordinary cases, would produce a leaf-bud, gives origin to a flowerbud. A flower-bud has not in ordinary circumstances any power of extension by the development of its central cellular portion. In this respect it differs from a leaf-bud. In some cases, however, of monstrosity, especially seen in the Rose (fig. 145) and Geum, the central part is prolonged, and bears leaves or flowers. In such cases the flowers, so far as their functional capabilities are concerned, are usua∏y abortive.

Flower-buds, like leafbuds, are produced in the axil of leaves, which are called floral leaves, bracts, or hypsophyllary leaves.

The term bract is properly applied to the leaf from which the primary floral axis, whether simple or branched, arises, while the leaves which arise on the axis between the bract and the outer envelope of the flower are bracteoles or bractlets. Bracts sometimes do not differ from the ordinary leaves, and are then called leafy, as in Veronica hederifolia, Vinca, Anagallis, and Ajuga. Like leaves they are entire or divided. In general as regards their form and appearance, they differ from ordinary leaves, the difference being greater in the upper than in the lower branches of an inflorescence. They are distinguished by their position at the base of the flower or flower-stalk. Their phyllotaxis is similar to that of the leaf. When the flower is sessile the bracts are often applied closely to the calyx, and may thus be confounded with it, as in Malvaceae and species of Dianthus and Eranthis, where they have received the name of epicalyx or calyculus. In some Rosaceous plants an epicalyx is present, due to the formation of stipulary structures by the sepals. In many cases bracts seem to perform the function of protective organs, within or beneath which the young flowers are concealed in their earliest stage of growth.

When bracts become coloured, as in Amherstia nobilis, Euphorbia splendens, Erica elegans, and Salvia splendens, they may be mistaken for parts of the corolla. They are sometimes mere scales or threads, and at other times they are abortive, and remain undeveloped, giving rise to the ebracteated inflorescence of Cruciferae and some Boraginaceae. Sometimes no flower-buds are produced in their axil, and then they are empty. A series of empty coloured bracts terminates the inflorescence of Salvia Horminum. The smaller bracts or bracteoles, which occur among the subdivisions of a branching inflorescence, often produce no flower-buds, and thus anomalies occur in the floral arrangements. Bracts are occasionally persistent, remaining long attached to the base of the peduncles, but more usually they are deciduous, falling off early by an articulation. In some instances they form part of the fruit, becoming incorporated with other organs. Thus, the cones of Firs and the strobili of the Hop are composed of a series of bracts arranged in a spiral manner, and covering fertile flowers; and the scales on the fruit of the Pine-apple are of the same nature. In amenta or catkins the bracts are called sguamae or scales. At the base of the general umbel in Umbelliferous plants, a whorl of bracts often exists, called a general involucre, and at the base of the smaller umbels or umbellules there is a similar leafy whorl called involucel or partial involucre. In some instances, as in Fool’s-parsley, there is no general involucre, but simply an involucel; while, in other cases, as in Fennel, neither involucre nor involucel is developed. In Composite the name involucre is applied to the leaves, scales, or phyllaries, surrounding the head of flowers (fig. 146, t), as in Dandelion, Daisy, Artichoke. This involucre is frequently composed of several rows of leaflets, which are either of the same or of different forms and lengths, and often lie over each other in an imbricated manner. When the bracts are arranged in two rows, and the outer row is perceptibly smaller than the inner, the involucre is sometimes said to be caliculate, as in Senecio. The leaves of the involucre are spiny in Thistles and in Dipsacus, and hooked in Burdock. Such whorled or verticillate bracts may either remain separate (poiyphytious), or may be united by cohesion (gamophyllous), as in many species of Bupleurum, and in Lavatera. In Composite besides the general envelope called the involucre, there are frequently chaffy and setose bracts at the base of each flower, and in Dipsacaceae a membranous tube surrounds each flower. These structures are of the nature of an epicalyx. In the acorn the cupula or cup (fig. 147, α) is formed by a growing upwards of the flowerstalk immediately beneath the flower, upon which scaly or spiny protuberances appear. It is of the nature of bracts. Bracts also compose the husky covering of the Hazel-nut. In the Yew the succulent covering of the seed is by some considered to be formed by the bracts.

When bracts become united together, and overlie each other in several rows, it often happens that the outer ones do not produce flowers, that is, are empty or sterile. In the Artichoke the outer imbricated scales or bracts are in this condition, and it is from the membranous white scales or bracts (paleati) forming the choke attached to the edible receptacle that the flowers are produced. The sterile bracts of the Daisy occasionally produce capitula, and give rise to the Hen-and-Chickens Daisy. In place of developing flowerbuds, bracts may, in certain circumstances, as in proliferous or viviparous plants, produce leaf-buds, and the flower-buds, like the leaf-buds, may be terminal or they may be lateral.

A sheathing bract enclosing one or several flowers is called a spatha or spathe. It is common among ΛIonocotyledons, as Narcissus (figure 148), Snow - flake, Arum, and Palms. In some Palms it is 20 feet long, and encloses 200,000 flowers. It is often associated with that form of inflorescence termed the spadix, and may be coloured, as in Richardia aethiopica, sometimes called the Aethiopian or Trumpet Lily. When the spadix is compound or branching, as in Palms, there are smaller spathes, surrounding separate parts of the inflorescence, to which the name spathéllae has sometimes been given. The spathe protects the flowers in their young state, and often falls off after they are developed, or hangs down in a withered form, as in some Palms, Typha, and Pothos. In Grasses the outer scales of the spikelets have been considered as sterile bracts, and have received the name of glumes (fig. 149, gï) ; and in Cyperaceae bracts enclose the organs of reproduction. Bracts are frequently changed into complete leaves. This change is called phyllody of bracts. It is seen in species of Plantago, especially in the variety of Plantago media, called the Rose-plantain in gardens. In this plant the bracts become leafy, and form a rosette round the flowering axis. Similar changes occur in Plantago major, P. lanceolata, Ajuga reptans, the Dandelion, the Daisy, the Dahlia, and in Umbelliferous plants. The conversion of bracts into stamens has been observed in the case of Abies excelsa. This has been called staminody of bracts. A lengthening of the axis of the female strobilus of Coniferae is not of infrequent occurrence in Cryptomeria japonica, Larix europaea, <tc., and this is usually associated with a leaf-like condition of the bracts, and sometimes even with the development of leaf-bearing shoots in place of the scales.

The arrangement of the flowers on the axis, or the ramification of the floral axie, is called inflorescence or anthotaxis. The primary axis of inflorescence is sometimes called rachis; its branches, whether terminal or lateral, which form the stalks supporting flowers or clusters of flowers, are peduncles, and if small branches are given off by it, they are called pedicels. A flower having a stalk is called pedunculate or pedicellate', one having no stalk is sessile. In describing a branching inflorescence, it is common to speak of the rachis as the primary floral axis, its branches as the secondary floral axes, their divisions as the tertiary floral axes, and so on; thus avoiding any confusion that might arise from the use of the terms rachis, peduncle, and pedicel.

The peduncle τa⅛y be cylindrical, compressed, or grooved; simple, bearing a single flower, as in Primrose; or branched, as in London-pride. It is sometimes succulent, as in the Cashew, in which it forms the large coloured expansion supporting the nut; spiral, as in Cyclamen and Vallisneria; or spiny, as in Alyssum spinosum. In some Rushes there is a green terete, and sometimes spiral, floral axis. Sometimes the peduncle proceeds from radical leaves, that is, from an axis which is so shortened as to bring the leaves close together in the form of a cluster, as in the Primrose, Auricula, Hyacinth, <fcc. In such cases it is termed a scape. The floral axis may be shortened, assuming a flattened, convex, or concave form, and bearing numerous flowers, as in the Artichoke, Daisy, and Fig (fig. 150). In these cases it is called a receptacle or phoranthium or clinanthium. The floral axis sometimes assumes a leaf-like or phylloid appearance, bearing numerous flowers at its margin, as in Xylophylla longifolia and in Ruscus; or it appears as if formed by several peduncles united together, constituting a fasciated axis, as in the Cockscomb, in which the flowers form a peculiar crest at the apex of the flattened peduncles. Adhesions occasionally take place between the peduncle and the bracts or leaves of the plant, as in the Lime-tree (fig. 151), Helwingia, Chailletia, several species of Hibiscus, and Zostera. The adhesion of the peduncles to the stem accounts for the extra-axillary position of flowers, as in many Solanaceae. When this union extends for a considerable length along the stem, several leaves may be interposed between the part where the peduncle becomes free and the leaf whence it originated, and it may be difficult to trace the connection. The peduncle occasionally becomes abortive, and in place of bearing a flower, is transformed into a tendril; at other times it is hollowed at the apex, so as apparently to form the lower part of the outer whorl of floral leaves as in Eschscholtzia. The termination of the peduncle, or the part on which the whorls of the flower are arranged, is called the thalamus or torus. The term receptacle is also sometimes applied to this, whether expanded and bearing several flowers, or narrowed so as to bear one. It may be considered as the growing point of the axis, which usually is arrested by the production of the flowers, but which sometimes becomes enlarged and expanded. Thus, in the Geranium (fig. 152) it is prolonged beyond the flower in the form of a beak (rostrate) ; in the Arum (fig. 153) it is a clubshaped fleshy column; in the Strawberry (fig. 154) it becomes a conical succulent mass, on which the seed-vessels are placed; while in Nelumbium it forms a truncated tabular top-like expansion (turbinate), enveloping the seedvessels., The margins of the receptacle may grow up whilst the centre remains depressed, and thus a concave torus is formed, as in the Rose (fig. 155). In some monstrous flowers of the Rose and Geum it is prolonged as a branch bearing leaves. Peduncles and pedicels sometimes become remarkably elongated, and this elongation of the flower-stalks sometimes alters the general character of the inflorescence. We occasionally observe the heads of flowers of the common White Clover becoming racemose by the lengthening of the flower-stalks.

Before proceeding to an examination of the parts of the flower-bud, we will consider the various modes of arrangement of the flowers upon the axis, or the inflorescence.

We may recognize two distinct types of inflorescence—one in which the flowers always arise as lateral shoots from a primary axi≡, which goes on elongating, and the lateral shoots never exceed in their development the length of the primary axis beyond their point of origin. The flowers are thus always axillary. Exceptions, such as in Cruciferous plants, are due to the non-appearance of the bracts. In the other type the primary axis terminates in a single flower, but lateral axes are given off from the axil of bracts, which again repeat the primary axis, and the development of each lateral axis is stronger than the primary axis beyond its point of origin. The flowers produced in this inflorescence are thus terminal. The first kind of inflorescence is indeterminate, indefinite, or axillary. Here the axis is either elongated, producing flower-buds as it grows, the lower expanding first; or it is shortened and depressed, and the outer flowers expand first. The expansion of the flowers is thus centripetal, that is, from base to apex, or from circumference to centre. This kind of inflorescence is shown in fig. 156, where the leaf from which the cluster of flowers is produced, f, represents the bract or floral leaf. The rachis, or primary axis of the flower is a'; this produces small leaflets b, which bear smaller flower-leaves or bractlets, from which pedunclee or secondary axes spring, each bearing single flowers. In fig. 157 the same kind of inflorescence is shown on a shortened axis, the outer flowers expanding first, and those in the centre last.

The second kind of inflorescence is determinate, definite, or terminal. In this the axis is either elongated and ends in a solitary flower, which thus terminates the axis, and if other flowers are produced, they belong to secondary axes farther from the centre; or the axis is shortened and flattened, producing a number of separate floral axes, the central one expanding first, while the others are developed in succession farther from the centre. The expansion of the flowers is in this case centrifugal, that is, from apex to base, or from centre to circumference. It is illustrated in fig. 158, where a representation is given of a plant of Ranunculus bulbosus; a' is the primary axis swollen at the base in a bulb-like manner b, and with roots proceeding from it. From the leaves which are radical proceeds the axis ending in a solitary terminal flower/'. About the middle of this axis there is a leaf or bract, from which a secondary floral axis a" is produced, ending m a single flower ∕", less advanced than the flower ∕'. This secondary axis bears a leaf also, from which a tertiary floral axis a " is produced, bearing an unexpanded ι solitary flower ∕'". From this tertiary axis a fourth is in progress of formation. Here/' is the termination of the primary axis, and this flower expands first, while the other flowers are developed centrifugally on separate axes.

A third series of inflorescences, termed mixed, may be recognized. In them the primary axis has an arrangement belonging to the opposite type from that on the branches, or vice versa. According to the mode and degree of development of the lateral shoots and also of the bracts, various forms of both inflorescences result.

Amongst indefinite forms the simplest occurs when a lateral shoot produced in the axil of a large single foliage leaf of the plant ends in a single flower, the axis of the plant elongating beyond, as in Veronica hederifolia, Vinca minor, and Lysimachia nemorum. The flower in this case is solitary, and the Inflorescence has been designated solitarιj axillary. The ordinary leaves in this case become floral leaves or bracts, by producing flower-buds in place of leaf-buds. The flowers, being all offshoots of the same axis, are said to be of the same generation or degree, and their number, like that of the leaves of this main axis, is indefinite, .varying with the vigour of the plant. Usually, however, the floral axis, arising from a more or less altered leaf or bract, instead of ending in a solitary flower, is prolonged, and bears numerous leaflets, called bracteoles or bractlets, from which smaller peduncles are produced, and those again in their turn may be branched in a similar way. Thus the flowers are arranged in groups, and frequently very complicated forms of inflorescence result. When the primary peduncle or floral axis, as in fig. 156, a', is elongated, and gives off pedicels a", of nearly equal length ending in single flowers, a raceme or cluster is produced, as in Currant, Hyacinth, and Barberry. If the secondary floral axes give rise to tertiary ones, the raceme is branching, and forms what is by some called a panicle, as in Yucca gloriosa; but it is better to restrict this term to the lax inflorescence of some Grasses and Rushes. If in a raceme the lower flower-stalks are developed more strongly than the upper, and thus all the flowers are nearly on a level, a corymb is formed, which may be simple, as in fig. 159, where the primary axis a divides into secondary axes a", a", which end in single flowers; or branching, where the secondary axes again subdivide. If the peduncles or secondary axes are very short or awanting, so that the flowers are sessile, a spike is produced, as in Plantago and Verbena officinalis (fig. 160). The spike sometimes bears unisexual flowers, usually staminiferous, the whole falling off by an articulation, as in Willow or Hazel (fig. 161), and then it is called an amentum or catkin, hence such trees are called amentiferous; at other times it becomes succulent, bearing numerous flowers, surrounded by a sheathing bract or spathe, and then it constitutes a spadix, which may be simple, as in Arum maculatum (fig. 153), or branching, as in Palms. A spike bearing female flowers only, and covered with scales, is either a strobilus, as in the Hop; or a cone, as in the Fir (fig. 133). It may be mentioned here that many do not consider the cone of Firs an axis ⅛f inflorescence, but regard it as a single flower with the floral leaves placed at different heights. In Grasses there are usually numerous sessile flowers arranged in small spikes, called locustae or spikelets, which are either set closely along a central axis, or produced on secondary axes formed by the branching of the central one; to the latter form the term panicle is applied.

If the primary axis, in place of being elongated, is contracted, it gives rise to other forms of indefinite inflorescence. When the axis is so shortened that the secondary axes arise from a common point, and spread out as radii of nearly equal length, each ending in a single flower, or dividing again in a similar radiating manner, an umbel is produced, as in fig. 162. From the primary floral axis a the secondary axes come off in a radiating or umbrella-like manner, and end in small umbels b, which are called partial umbels or umbellules, to distinguish them from the general umbel formed by the branching of the primary axis. This inflorescence is seen in Hemlock, and other allied plants, which are hence called Umbelliferous. If there are numerous flowers on a flattened, convex, or slightly concave receptacle, having either very short pedicels or none, a capitιdum (head), anthodium, or calathium, is formed, as in Dandelion, Daisy, and other Composite plants (fig. 146), also in Scabiosa (fig. 157) and Dipsacus. In the American Button-bush the heads are globular, in some species of Teazel, elliptical, while in Scabious, and in Com posite plants, as Sunflower, Dandelion, Thistle, Centaury, and Marigold, they are somewhat hemispherical, with a flattened, slightly hollowed, or convex disk. If the margins of such a receptacle be developed upwards, the centre not developing, a concave receptacle is formed, which may partially or completely enclose a number of flowers that are generally unisexual. This gives rise to the peculiar inflorescence of Dorstenia (fig. 163), or to that of the Fig (fig. 150), where the flowers are placed on the inner surface of the hollow receptacle, and are provided with bracteoles. This inflorescence has been called lιypanthodium.

Lastly, we have what are called compound indefinite inflorescences. In these forms the lateral shoots, developed centripetally upon the primary axis, bear numerous bracteoles, from which floral shoots arise which may have a centripetal arrangement similar to that on the mother shoot, or it may be different. Thus we may have a group of racemes, arranged in a racemose manner on a common axis, forming a raceme of racemes or compound raceme, as in Astilbe. In the same way we may have compound umbels, as in Hemlock and most Umbelliferae (fig. 162), a compound spike, as in Rye-grass, a compound spadix, as in some Palms, and a compound capitulum, as in the Hen-and-Chickens Daisy. Again, there may be a raceme of capitula, that is, a group of capitula disposed in a racemose manner, as in Petasites, a raceme of umbels, as in Ivy, and so on, all the forms of inflorescence being indefinite in disposition.

The elongation of secondary flower-stalks sometimes alters the general character of the inflorescence, changing a spike into a raceme, a raceme into a corymb, a capitulum into an umbel, and so forth. The capitulum of flowers in some Compositae, such as Hypochaeris radicata and Senecio vulgaris, by a similar change in the pedicels assumes the form of an umbel. Among Umbelliferae the umbels are sometimes supported on very long 3talks, while the pedicels of the individual flowers are not lengthened. In Eryngium the shortening of the pedicels changes an umbel into a capitulum. The umbellate inflorescence of Pelargonium has been seen changed into a raceme.

The simplest form of the definite type of inflorescence is seen in Anemone nemorosa and in Gentianella (Gentiana acaulis, fig. 164), where the axis terminates in a single flower, no other flowers being produced upon the plant. This is solitary terminal inflorescence. If other flowers were produced, they would arise as lateral shoots from the bracts below the first-formed flower. The general name of cyme is applied to the arrangement of a group of flowers in a definite inflorescence. A cymose inflorescence is an inflorescence where the primary floral axis before terminating in a flower gives off one or more lateral unifloral axes which repeat the process,—the development being only limited by the vigour of the plant. The floral axes are thus centrifugally developed. The cyme, according to its development, has been characterized as biparous or uniparous. In fig. 165 the biparous cyme is represented in the flowering branch of Erythraea Centaurium. Here the primary axis a' ends in a flower/', which has passed into the state of fruit. At its base two leaves are produced, each of which is capable of developing buds. These are flower-buds, and constitute secondary axes a" a", ending in single flowers/"/''', which are thus terminal and solitary; and at the base of these axes a pair of opposite leaves is produced, giving rise to tertiary axes a'" a'" a'", ending in single flowers f"' f"' f ’ ", and so on. The term dichotomous has also been applied to this form of cyme. But these terms are not strictly correct, for here there is no dichotomous branching, although, when the terminal flower ∕ of the generating axis has withered, an apparent dichotomy occurs; but the lateral axes on the several shoots are produced monopodially, and therefore the term is apt to lead to confusion regarding the development of the shoots. The name dichasium has, therefore, been substituted for the old term. In the natural order Caryophyllaceae (Pink family) the dichasial cymose form of inflorescence is very general. In some members of the tribe, as Dianthus barbatus, D. Carthusianorum, <fcc., in which the peduncles are short, and the flowers closely approximated, with a centrifugal expansion, the inflorescence has the form of a contracted dichasium, and receives the name of fascicle (fig. 166). A

similar inflorescence is seen in Xylophylla longifolia. When the axes become very much shortened, the arrangement is more complicated in appearance, and the nature of the inflorescence can only be recognized by the order of opening of the flowers. In Labiate plants, as the Dead-nettle (Lamium), the flowers are produced in the axil of each of the foliage leaves of the plant, and they appear as if arranged in a simple whorl of flowers. But on examination it is found that there is a central flower expanding first, and from its axis two secondary axes spring bearing solitary flowers; the expansion is thus centrifugal. The inflorescence is therefore a contracted dichasium, the flowers being sessile, or nearly so, and the clusters are called verticillasters (fig. 167).

Sometimes, especially towards the summit of a dichasium, owing to the exhaustion of the growing power of the plant, only one of the bracts gives origin to a new axis, the other remaining empty; thus the inflorescence becomes unilateral, and further development is arrested. In addition to the dichasial form there are others where more than two lateral axes are produced from the primary floral axis, each of which in turn produces numerous axes. To this form the terms trichotomous and polytomous cyme have been applied; but these are now usually designated cymose umbels. They are well seen in some species of Euphorbia. Another term, anthela, has been used to distinguish such forms as occur in several species of Luzula and Juncus, where numerous lateral axes arising from the primary axis grow very strongly, and develop in an irregular manner.

In the uniparous cyme a number of floral axes are successively developed one from the other, but the axis of each successive generation, instead of producing a pair of bracts, produces only a single one. The basal portion of the consecutive axes may become much thickened and arranged more or less in a straight line, and thus collectively form an apparent or false axis or sympodium, and the inflorescence thus simulates a raceme. In the true raceme, however, we find only a single axis, producing in succession a series of bracts, from which the floral peduncles arise as lateral shoots, and thus each flower is on the same side of the floral axis as the bract in the axil of which it is developed; but in the uniparous cyme the flower of each of these axes, the basal portions of which unite to form the false axis, is situated on the opposite side of the axis to the bract from which it apparently arises (fig. 168). This bract is not, however, the one from which the axis terminating in the flower arises, but is a bract produced upon it, and gives origin in its axil to a new axis, the basal portion of which, constituting the next part of the false axis, occupies the angle between this bract and its parent axis, —the bract from which the axis really does arise being situated lower down upon the same side of the axis with itself. The uniparous cyme presents two forms, the scor- pioid or cicinal and the helicoid or bostrychoid.

Γn the scorpioid cyme the flowers are arranged alternately in a double row along one side of the false axis (fig. 169) the bracts when developed forming a second double row on the opposite side, as seen in the Henbane; the whole inflorescence usually curves on itself like a scorpion’s tail, hence its name. In fig. 170 is shown a diagrammatic sketch of this arrangement. The false axis, abed, is formed by successive generations of unifloral axes, the flowers being arranged along one side alternately and in a double row; had the bracts been developed they would have formed a similar double row on the opposite side of the false axis; the whole inflorescence is represented as curved on itself. In fig. 169 the same scorpioid form of uniparous cyme is seen in the Forget-me-not, with the double row of flowers on one side of the false axis, but in this case the bracts, which should appear on the opposite side, are not developed, and hence the cyme is not complete. The inflorescences in the family Boraginaceae are usually regarded as true scorpioid cymes. But it has been recently shown that in Myosotis (Forget-me-not) the axis is not a sympodium, but the branching takes place monopodially, therefore it should more correctly be regarded as a raceme with flowers only on one side and curving on itself. In Hyoscyamus (Henbane) and other Atropaceae the branching is dichotomous, so that the inflorescence in this case also ought not to be regarded as a scorpioid cyme.

In the helicoid cyme there is also a false axis formed by the basal portion of the separate axes, but the flowers are not placed in a double row, but in a single row, and form a spiral or helix round the false axis. In Alströmeria, as represented in fig. 168 the axis a' ends in a flower (cut off in the figure) and bears a leaf. From the axil of this leaf, that is, between it and the primary axis a' arises a secondary axis a", ending in a flower/", and producing a leaf about the middle. From the axil of this leaf a tertiary floral axis a'", ending in a flower f"', takes origin. In this case the axes are not arranged in two rows along one side of the false axis, but are placed at regular intervals, so as to form an elongated spiral round it.

Compound definite inflorescences are by no means common, but in Streptocarpus polyanthus, and in several Calceolarias, we probably have examples. Here there are scorpioid cymes of pairs of flowers, each pair consisting of an older and a younger flower.

Forms of inflorescence occur, in which both the definite and indefinite types are represented; these, then, are mixed inflorescences. Thus in Composite plants, such as Hawkweeds (Hieracia) and Ragworts (Senecio, fig. 171), the heads of flowers, taken as a whole, are developed centrifugally, the terminal head first, while the florets, or small flowers on the receptacle, open centripetally, those at the circumference first. So also in Labiatae, such as Dead-nettle (Lamium), the different whorls of inflorescence are developed centripetally, while the florets of the verticillaster are centrifugal. This mixed character presents difficulties in such cases as Labiatae, where the leaves, in place of retaining their ordinary form, become bracts, and thus might lead to the supposition of the whole series of flowers being one inflorescence. In such cases the cymes are described as spiked, racemose, or panicled, according to circumstances. In Saxifraga umbrosa (London-pride) and in the Horse-chestnut we meet with a raceme of scorpioid cymes; in Sea-pink, a capitulum of contracted scorpioid cymes (often called a glomerulus); in Laurustinus, a compound umbel of dichasial cymes; a scorpioid cyme of capitula in Vernonia centriflora. In the catkins of the Birch we have, in reality, spikes of contracted dichasial cymes. In the Bell-flower (Campanula), there is a racemose uniparous cyme. In the Privet (Ligustrum vulgare) there are numerous racemes of dichasia arranged in a racemose manner along an axis; the whole inflorescence thus has an appearance not unlike a bunch of grapes, and has been called a thyrsus. [table]

[table]

’ The flower is the tout ensemble of those organs which are concerned in reproduction. It is constituted by a portion of the floral axis bearing the sexual organs, usually with certain protective envelopes. The parts borne on the floral axis and composing the flower are all metamorphosed leaves, and, though usually very different in appearance from foliage leaves, their morphological resemblance is frequently shown by their developing as foliage leaves. The axis bearing the parts of the flower is usually very much contracted, no internodes being developed, and it frequently expands into a flattened or hollowed expansion termed the thalamus or torus; at other times, though rarely, the internodes are developed and it is elongated. Upon this torus the parts of the flower are arranged in a crowded manner, usually forming a series of verticils, the parts of which alternate; but they are sometimes arranged in a spiral manner, especially if the floral axis be elongated. In a typical flower, as in fig. 172, we recognize four distinct whorls of leaves :—an outer whorl of five parts, the calyx; within it, another whorl of five parts alternating with those of the outer whorl, the corolla; next comes a whorl of parts alternating with the parts of the corolla, the androecium; and in the centre is the gynoecium. Fig 173 is a diagrammatic representation of the arrangement of the parts of such a flower. The flower is supposed to be cut transversely, and the parts of each whorl are distinguished by a different symbol. Of these whorls the two internal, forming the male and female sexual organs, constitute the essential organs of reproduction; the two outer whorls are the protective coverings or floral envelopes. The calyx or outer covering (fig. 172, c) is formed of leaves, called sepals, which are generally of a greenish colour. The corolla p, the next covering, is composed of leaves, called petals, which are often showy, and normally alternate with the sepals. Sometimes, as in many Monocotyledons, the calyx and corolla both display rich colouring, and are apt to be confounded. In such cases the term perianth, or perigone, has been applied to avoid ambiguity. Thus, in the Tulip, Crocus, Lily, Hyacinth, authors speak of the parts of the perianth, in place of calyx and corolla, although in these plants an outer whorl (calyx) may be detected, of three parts, and an inner (corolla), of a similar number, alternating with them. When the parts of the calyx are in appearance like petals they are said to ∖>o, petaloid, as in Liliaceae. In some cases the petals have the appearance of sepals, then they are sepaloid, as in Juncaceae. The term perianth is usually confined to the flowers of Monocotyledons, whatever colour they present, whether green, as in Asparagus, or coloured, as in Tulip. Some use the term perianth as a general one, and restrict the use of perigone to cases where a pistil only is present. In plants, as Nymphsea alba, where a spiral arrangement of the floral leaves occur, it is not easy to say where the calyx ends and the corolla begins, as these two whorls pass insensibly into each other. When both calyx and corolla are present, the plants are dichlamydeous; occasionally one becomes abortive, and then the flower is monochlamy deous (fig. 174), having a calyx only, or apetalous, having no petals. At times both are abortive, and then the flower is achlamydeous, or naked. It is important to remember that if only one perianth-whorl is present it is the calyx. The outermost whorl of the essential organs, collectively termed the androecium, is composed of a series of leaves distinguished as the staminal leaves or stamens. These are the male sexual organs. In their most differentiated form each consists of a stalk, either filiform or foliaceous, the filament (fig. 175, ∕) supporting at its summit a lobed saccate mass, the anther (fig. 175, a), which contains a powdery matter, the pollen (fig. 175, p), which is discharged therefrom. In many cases, however, the staminal leaf resembles more nearly a cataphyllary leaf, bearing the pollen-sacs scattered over its surface, as in Cycadacese. The gynoecium or pistil is the central portion of the flower, and is the female sexual organ. It consists of one or more folded leaves, called carpels, either separate (fig. 172, c) or combined (fig. 174.) The parts distinguished in the pistil are the ovary (fig. 176, o'), which is the lower portion enclosing the ovules destined to become seeds, and the stigma (fig. 176, g), a portion of loose cellular tissue uncovered by epidermis, which is either sessile on the apex of the ovary, as in the Poppy, or is separated from it by a prolonged portion called the style (fig. 176, s). The androecium and gynoecium are not present in all flowers. When both are present the flower is hermaphrodite; and in descriptive botany such a flower is indicated by the symbol ?. When only one of those organs is present the flower is unisexual or diclinous, and is cither male {staminate), J, or female {pistillate), ?. A flower then normally consists of the four whorls of leaves,—calyx, corolla, androecium, and gynoecium,—and when these are all present the flower is complete. These several whorls of the flower arc usually densely crowded upon the thalamus, but in some instances, after apical growth has ceased in the axis, an elongation of portions of the receptacle by intercalary growth occurs, by which changes in the position of the parts may be brought about. Thus in Lychnis an elongation of the axis betwixt the calyx and the corolla takes place, and in this way they are separated by an interval. Again, in Passiflora the stamens are separated from the corolla by an elongated portion of the axis, which has consequently been termed the androphore; and in Passiflora, Fraxinella (fig. 177), Capparidaceae, and some other plants, the ovary is raised upon a distinct stalk termed the gynophore; it is thus separated from the stamens, and is said to be stipitate. The apical growth of the floral axis in the flower soon ceases, and therefore the parts are arranged in whorls; but at times, as in the Ranunculaceae, Magnoliaceae, &c., the growth is of sufficient duration to permit a spiral arrangement of parts. Usually the successive whorls of the flower disposed from below upwards or from without inwards upon the floral axis are of the same number of parts, or are a multiple of the same number of parts, those of one whorl alternating with those of the whorls next it.

When a flower consists of parts arranged in whorls it is said to be cyclic, and if all the whorls have an equal number of parts and are alternate it is encyclic. In contrast to the cyclic flowers are those, as in Magnoliaceae, where the parts are in spirals. When in such a flower the transition from one foliar structure to another, i.e., from calyx to corolla, from corolla to stamens, &c., does not coincide with a definite number of turns of the spiral it is said to be acyclic; if it does so correspond it is termed hemicyclic; the latter term also includes flowers which are cyclic at one portion and spiral at another, as in many Ranunculaceae. In spiral flowers the distinction into whorls is by no means easy, and usually there is a gradual passage from sepaloid through petaloid to staminal parts, as in Nymphaeaceae (figs. 178, 179), although in some plants there is no such distinction, the parts being all petaloid, as in Trollius. We may again note here that some regard the cone in C'oniferae and Cycadaceae as a single flower, and in that case the whorls are much separated and arranged spirally along the floral axis. Normally, the parts of successive whorls alternate; but in some cases we find the parts of one whorl opposite or superposed to those of the next whorl. In some cases, as in the Ampelideae, this seems to be the ordinary mode of development, but in Caryophyllaceae the superposition of the stamens on the sepals in many plants seems to be due to the suppression or abortion of the whorl of petals, and this idea is borne out by the development, in some plants of the order, of the suppressed whorl. In Primulaceae, again, where there is a superposition of stamens and petals the abnormality is due apparently to another cause. The petals are developed after the stamens, and are to be regarded as appendages from them, of the same nature as the appendages to the stamens in Asclepiadaceae, so that morphologically in Primulaceae, according to this view, there are no petals. As a rule, whenever we find the parts of one whorl superposed to those of another we may suspect some abnormality. Frequently, when parts are superposed they become adherent to one another; thus the stamens become adherent to the petals or are epipetalous, or to the sepals and are episepalous.

A flower is said to be symmetrical when each of its whorls consists of an equal number of parts, or when the parts of any one whorl are multiples of that preceding it. Thus, a symmetrical flower may have five sepals, five petals, five stamens, and five carpels, or the number of any of these parts may be ten, twenty, or some multiple of five. Fig. 173 is a diagram of a symmetrical flower, with five parts in each whorl, alternating with each other. In fig. 180 there is a section of a symmetrical flower of Stone-crop, with five sepals, five alternating petals, ten stamens, and five carpels. Here the number of parts in the staminal whorl is double that in the others, and in such a case the additional five parts form a second row alternating with the others. In the staminal whorl especially it is common to find additional rows. In fig. 181 there is a symmetrical flower, with five parts in the three outer rows and ten divisions in the inner. In this case it is th gynoecium which has an additional number of parts. Fig 182 shows a flower of Heath, with four divisions of th calyx and corolla, eight stamens in two rows, and fou divisions of the pistil. In fig. 183 there are three part: in each whorl; and in figs. 184 there are three divi sions of the calyx, corolla, and pistil, and six stamen: in two rows. In all these cases the flower is symmetrical Where, as in the Stone-crop, an extra row of parts is deve loped in any whorl they may be either formed in regula succession within the first row, or they may be interposée to them, i.e., formed between them, or even external to am beneath them. We have examples of the former in man] Caryophyllaccous plants, the latter being well seen iι Crassulaceae, Geraniaceae, <kc. In Monocotyledons it i usual for the staminal whorl to be double, it rarely having more than two rows, whilst amongst Dicotyledons there are often very numerous rows of stamens. The floral envelopes are rarely multiplied Flowers in which the number of parts in each whorl is the same, are isomerous (of equal number); when the number in some of the whorls is different, the flower is anisomerous (of unequal number). The pistilline whorl is very liable to changes. It frequently happens that when it is fully formed, the number of its parts is not in conformity with that of the other whorls. In such circumstances, however, a flower has been called symmetrical, provided the parts of the other whorls are normal, —the permanent state of the pistil not being taken into account in determining symmetry. Thus, fig. 185 shows a pentamerous symmetrical flower, with dimerous pistils. Symmetry, then, in botanical language, has reference to a certain definite numerical relation of parts. A flower in which the parts are arranged in twos, is called dimerous, the symmetry being binary and the arrangement marked thus ^∕∙ This may be considered either as analogous to opposite decussating leaves, or as composed of distichous alternate leaves belonging to the one-half series. When the parts of the floral whorls are three, the flower is trimerous, and the symmetry being ternary or trigonal is marked f∕. This may be looked upon as composed of verticillate leaves, or of tristichous alternate leaves with the angular divergence ⅜. When floral whorls have parts in a series of four, the flower is tetramerous, its symmetry being quaternary, and marked ty. A pentamerous flower, marked ty, has quinary or pentagonal symmetry. The symmetry which is most commonly met with in the vegetable kingdom is trimerous and pentamerous—the former occurring generally among Monocotyledons, the latter among Dicotyledons. Dimerous or tetramerous symmetry occurs also among Dicotyledonous plants, and the numbers 2 and 4 prevail in the reproductive organs of Acotyledons. The various parts of the flower have a certain definite relation to the axis. Thus, in axillary tetramerous flowers (fig. 182), one sepal is next the axis, and is called superior or posterior; another is next the bract, and is inferior or anterior, and the other two are lateral; and certain terms are used to indicate that position. A plane passing through the anterior and posterior sepal and through the floral axis is termed the median plane of the flower; a plane cutting it at right angles, and passing through the lateral sepals, is the lateral plane; whilst the planes which bisect the angles formed by the lateral and median planes are the diagonal planes, and in these flowers the petals which alternate with the sepals are cut by the diagonal planes.

In a pentamerous flower one sepal may be superior, as in the calyx of Rosaceae and Labiatae; or it may be inferior, as in the calyx of Leguminosae (fig. 186),—the reverse, by the law of alternation, being the case with the petals. Thus, in the blossom of the Pea (fig. 187), the odd petal (vexillum) st is superior, while the odd sepal is inferior. In the Figwort order one of the two carpels is posterior and the other anterior, whilst in Convolvulaceae the carpels are arranged laterally. Sometimes the twisting of a part makes a change in the position of other parts, as in Orchids, where the twisting of the ovary changes the position of the labellum. When the different members of each whorl are like in size and shape, the flower is said to be regular; while differences in the size and shape of the parts of a whorl make the flower irregular, as in the Papilionaceous flower, represented in fig. 187. When a flower can be divided by a single median plane into two exactly similar parts, then it is said to be zygomorphic. Such flowers as Papilionaceae, Labiatae, are examples. In contrast with this are poly symmetrical flowers, which can be divided by several planes into several exactly similar portions; such are all regular, symmetrical flowers. When the parts of any whorl are not equal to or some multiple of the others, then the flower is unsymmetrical. This want of symmetry may be brought about in various ways. A consideration of the various unsymmetrical, irregular, and incomplete forms of the flower, and the processes by which they have been brought about, constitutes Teratology. Alteration in the symmetrical arrangement as well as in the completeness and regularity of flowers has been traced to suppression or the non-development of parts, degeneration or imperfect formation, cohesion or union of parts of the same whorl, adhesion or union of the parts of different whorls, multipli· cation of parts, and deduplication (sometimes called chorisis) or splitting of parts.

By suppression or non-appearance of a part at the place where it ought to appear if the structure was normal, the symmetry or completeness of the flower is disturbed. This suppression when confined to the parts of certain verticils makes the flower unsymmetrical. Thus, in the flowers of Staphylea there are five parts of the calyx, five petals, five stamens, and only two carpels; in many Caryophyllaceae, as Polycarpon and Holosteum, while the calyx and corolla are pentamerous, there are only three or four stamens and three carpels; in Impatiens noli-me-tangere the calyx is composed of three parts, while the other verticils have five; in Labiate flowers there are five parts of. the calyx and corolla, and only four stamens; and in Tropaeolum pentaphyllum there are five sepals, two petals, eight stamens, and three carpels. In all these cases the want of symmetry is traced to the suppression of certain parts. In the last-mentioned plant the normal number is five, hence it is said that there are three petals suppressed, as shown by the position of the two remaining ones; there are two rows of stamens, in each of which one is wanting, and there are two carpels suppressed. In many instances the parts which are afterwards suppressed can be seen in the early stages of growth, and occasionally some vestiges of them remain in the fully developed flower. By the suppression of the verticil of the stamens, or of the carpels, flowers become unisexual, or diclinous, and by the suppression of one or both of the floral envelopes, monochlamydeous and achlamydeous flowers are produced. The suppression of parts of the flower may be carried so far that at last a flower consists of only one part of one whorl. In the Euphorbiaceae we have an excellent example of the gradual suppression of parts, where from an apetalous trimerons staminal flower we pass to one where one of the stamens is suppressed, and then to forms where two of them are wanting. We next have flowers in which the calyx is suppressed, and its place occupied by one, two, or three bracts (so that the flower is, properly speaking, achlamydeous), and only one or two stamens are produced. And finally, we find sterile flowers consisting of a single stamen with a bract, and fertile flowers consisting of a single carpel with a bract. There is thus traced a degradation, as it is called, from a flower with three stamens and three divisions of the calyx, to one with a single bract and a single stamen or carpel.

Degeneration, or the transformation of parts, often gives rise either to an apparent want of symmetry or to irregularity in form. In unisexual flowers it is not uncommon to find vestigee of the undeveloped stamens in the form of filiform bodies or scales. To many of these staminal degenerations Linnaeus gave the name of nectaries. In double flowers transformations of the stamens and pistils take place, so that they appear as petals. In C’anna, what are called petals are in reality metamorphosed stamens. In the capitula of Compositae we sometimes find the florets converted into green leaves. The limb of the calyx may appear as a rim, as in some Umbelliferae; or as pappus, in Compositae and Valeriana. In Scrophularia the fifth stamen appears as a scale-like body, called staminodium; in other Scrophulariaceae,as inPentstemon,it assumes the form of a filament, with hairs at its apex in place of an anther.

Cohesion, or the union of parts of the same whorl, and adhesion, or the growing together of parts of different whorls, are very common causes of change both as regards form and symmetry. Thus in Cucurbita the stamens are originally five in number, but subsequently some cohere, so that three stamens only are seen in the mature flower. Again, amongst Malvaceae the stamens cohere by their margins. Cases of what are called gamopetalous corollas, gamosepalous calyces, Ac., must be distinguished carefully from such a process as this ; for in these cases the parts were never free from one another, but developed from the very first as one mass; and the same caution is necessary in many instances of so-called union of stameps, which are really cases of branching. Adhesion is well seen in the gynostemium, of Orchids, where the stamens and ovary adhere; and in flowers where the stamens are perigynous, i.e., are adherent to the calyx. In Capparidaceae the calyx and petals occupy their usual position, but the axis is prolonged in the form of a gynophore, to which the stamens are united. Cohesion and adhesion are rare amongst Monocotyledons.

Multiplication, or an increase of the number of parts, gives rise to changes in plants. We have already alluded to the interposition of new members in a whorl. This takes place chiefly in the staminal whorl, but usually the additional parts produced form a symmetrical whorl with the others. In some instances, however, this is not the case. Thus in the Horse-chestnut there is an interposition of two stamens betwixt the other whorl, and thus seven stamens are formed in the flower, which is unsymmetrical.

Parts of the flower are often increased by a process of deduplication, unlining, dilamination, or chorization, i.e., the splitting of a part so that two or more parts are formed out of what was originally one. This is believed by some to take place in a remarkable degree in the case of appendages to petals. Thus, in Ranunculus, the petal has a scale at its base, which is looked upon as a mere fold of it. This fold may in some cases be more highly developed, as in Caryophyllaceae, and in Crassula rubens (fig. 172, a). Others refer such cases to the formation of outgrowths similar in nature to the ligule in Grasses. In Cruciferous plants a case of chorisis is said to occur. There the staminal whorl consists of four long stamens and two short ones (tetradynamous'). The symmetry in the flower is evidently dimerous, and the abnormality in the androecium, where the four long stamens are opposite the posterior sepals, is said to take place by a splitting of the filaments of two stamens into four; and thus the two long stamens on each side are, by gemination, normally one. This view is supported by cases in which the filaments of the long stamens are more or less united; also by cases in which the shorter filaments exhibit tooth-like processes on both sides, while the longer ones have them only on the outer side. In such cases the two long filaments, if united, would present the same appearance as the shorter ones, and occupy their usual position of alternation with the petals. In some instances, by pelorization, it is found that tetradynamous plants become tetrandrous, with stamens of equal length alternating with the petals. Many cases of what was considered chorisis are in reality due to the development of stipules from the staminal leaf. Thus in Dicentra and Corydalis there are six stamens in two bundles; the central one of each bundle alone is perfect, the lateral ones have each only half an anther, and they are really stipules formed from the staminal leaf. Branching of stamens also produces apparent want of symmetry; thus, in the so called polyadelphous stamens of Hypericaceae there are really only five stamens which give off numerous branches, but the basal portion remaining short, the branches have the appearance of separate stamens, and the flower thus seems unsymmetrical.

Cultivation has a great effect in causing changes in the various parts of plants. Many alterations in form, size, number, and adhesion of parts are due to the art of the horticulturist. The changes in the colour and forms of flowers thus produced are endless. In the Dahlia the florets are rendered quilled, and are made to assume many glowing colours. In Pelargonium the flowers have been rendered larger and more showy; and such is also the case with the Ranunculus, the Auricula, and the Carnation. Some flowers, with spurred petals in their usual state, as Columbine, are changed so that the spurs disappear; and others, as Linaria, in which one petal only is usually spurred, are altered so as to have all the petals spurred, and to present what are called pelorian varieties.

As a convenient method of expressing the arrangement of the parts of the flower, floral formulae have been devised. Several modes of expression are employed. The following is a very simple mode which has been proposed :—The several whorls are represented by the letters S sepals, P petals, St stamens, C carpels, and a figure marked after each indicates the number of parts in that whorl. Thus the formula S 5 P.StC 5 means that the flower is perfect, and has pentamerous symmetry, the whorls being isomerous. Such a flower as that of Sedum (fig. 180) would be represented by the formula S 5 P 5 St s+6 C s , where St 5+6 indicates that the staminal whorl consists of two rows of five parts each. A flower such as the male flower of the Nettle (fig. 188) would be expressed S 4 P 0 St 4 C 0 . It is also possible to indicate, in cases where members of a whorl are absent, which of them are wanting; thus, in the flower of an Orchis we have S 3 P 3 St i+0 C 3 , Here St∙ i+0 indicates that the anterior stamen of the outer row is present, the other two stamens of the row, marked as dots, are wanting, as also all the stamens of the inner row. In Cypripedium, on the other hand, the formula is S 3 P 3 St 0+2 C 3 , for here all the stamens of the outer row and the anterior one of the inner row are absent. When no other mark is appended the whorls are supposed to be alternate; but if it is desired to mark the position of the whorls special symbols are employed. Thus, to express the superposition of one whorl upon another, a line is drawn between them, e.g., the symbol S 5 P 5 ] St 5 C 5 is the formula of the flower of Primulaceae. To indicate the interposition of a row of parts in a whorl a dot is placed between the coefficients, thus S 5 P p St 5 . 5 C 5 indicates that an extra row of five stamens has been interposed between those already in existence. To represent adhesions and cohesions leads to complicated formulae, but in many cases cohesion A A A

can be easily indicated. Thus in the formula S 4 P 4 St 2 C 2 , we have the arrangement of the parts of thefloweι , in Veronica indicated, the sepals and the petals being united and the two carpels also being united into one pistil. The papilionaceous flower, of which fig. 186 is a diagram, may be formulated A

S j P 5 St∕7(4+ 1 )C 1 . We thus learn that there are five sepals united, five free petals, ten stamens in two rows, of which nine are united and one is free, and there is one carpel. When the parts of the flower are arranged spirally on the floral axis, as in Magnoliaceae and other flowers, the formula is prefixed by a curved line thus G—and then the angle of divergence of the members may be marked in addition to their number. Many other points in the arrangement of the flower may be attached to the formula by different symbols, according to the object which one has in view.

. To the flower-bud the name aldbastrus is sometimes given, and its period of opening has been called anthesis, whilst the manner in which the parts are arranged with respect to each other before opening is the aestivation or proefloration. The latter terms are applied to the flowerbud in the same way as vernation is to the leaf-bud, and distinctive names have been given to the different arrangements exhibited, both by the leaves individually and in their relations to each other. As regards each leaf of the flower, it is either spread out, as the sepals in the bud of the Lime-tree, or folded upon itself (conduplicate), as in in ζ e !^ a ^ s some species of Lysimachia, or slightly folded inwards or outwards at the edges, as in the calyx of some species of Clematis, -and of some herbaceous plants, or rolled up at the edges (involute or revolute), or folded transversely, becoming crumpled or corrugated, as in the Poppy. When the parts of a whorl are placed in an exact circle, and are applied to each other by their edges only, without overlapping or being folded, thus resembling the valves of a seed-vessel, the aestivation is valvate, as in the calyx of Guazuma ulmifolia (fig. 189). The edges of each of the parts may be turned either inwards or outwards; in the former case, the aestivation is induplicate, as in the corolla of Guazuma ulmifolia (fig. 190), in the latter case, reduplicate, as in the calyx of Althaea rosea (fig. 191). When the parts of a single whorl are placed in a circle, each of them exhibiting a torsion of its axis, so that by one of its sides it overlaps its neighbour, whilst its side is overlapped in like manner by that standing next to it, the aestivation is tιvisted or contortive, as in the corolla of Althaea rosea (fig. 192). This arrangement is characteristic of the flower-buds of

Malvaceae and Apocynaceae, and it is also seen in Convolvulaceae and Caryophyllaceae. When the flower expands, the traces of twisting often disappear, but sometimes, as in Apocynaceae, they remain. Those forms of aestivation are such as occur in cyclic flowers, and they are included under circular aestivation. But in spiral flowers we have a different arrangement, the leaves of the calyx of Camellia japonica cover each other partially like tiles on a house. This aestivation is imbricate. At other times, as in the petals of Camellia, the parts envelope each other completely, so as to become convolute. This is also seen in a transverse section of the calyx of Magnolia grandiflora, where each of the three leaves embraces that within it. When the parts of a whorl are five, as occurs in many Dicotyledons, and the imbrication is such that there are two parts external, two internal, and a fifth which partially covers one of the internal parts by its margin, and is in its turn partially covered by one of the external parts, the aestivation is quincuncial (fig. 193). This quincunx is common in the corolla of Rosaceae. In fig. 194 a section is given of the bud of Antirrhinum majus, showing the imbricate spiral arrangement. In this case it will be seen that the part marked 5 has, by a slight change in position, become overlapped by 1. This variety of imbricate aestivation has been termed cochlear. In flowers, such as those of the Pea (fig. 187), one of the parts, the vexillum, is often large and folded over the others, giving rise to vexillary aestivation (fig. 195), or the carina may perform a similar office, and then the aestivation is carinal, as in the Judas-tree (Cereis Siliquastrum). The parts of the several verticils often differ in their mode of aestivation. Thus, in Malvaceae the corolla is contortive and the calyx valvate, or reduplicate; in St John’s-wort the calyx is imbricate, and the corolla contortive. In Convolvulaceae, while the corolla is twisted, and has its parts arranged in a circle, the calyx is imbricate, and exhibits a spiral arrangement. In Guazuma the calyx is valvate, and the corolla induplicate. The circular aestivation is generally associated with a regular calyx and corolla, while the spiral aestivations are connected with irregular as well as with regular forms.

I. Protective Organs.

1. Calyx.

The calyx is the external envelope of the flower, and consists of verticillate leaves, called sepals, foliola, or plιylla. These calycine leaves are sometimes separate from each other, at other times they are united to a greater or less extent; in the former case, the calyx is dialysepalous or eleutherosepalous, poly sepalous or polyphyllous, or apo-sepalous; in the latter, gamosepalous or gamoplιyllous, mono-sepalous or monophyllous, or synsepalous. The divisions of the calyx present usually all the characters of leaves, and in some cases of monstrosity they are converted into the ordinary leaves of the plant. Their structure consists of cellular tissue or parenchyma, traversed by vascular bundles, in the form of ribs and veins, containing spiral vessels, which can be unrolled, delicate woody fibres, and other vessels, the whole being enclosed in an epidermal covering, having stomata and often hairs on its outer surface, which corresponds to the under side of the leaf. In the great divisions of the vegetable kingdom the venation of the calyx is similar to that of the leaves,—parallel in Monocotyledons, reticulated in Dicotyledons. The leaves of the calyx are usually entire, but occasionally they are cut in various ways, as in the Rose, and they are sometimes hooked at the margin, as in Rumex uncatus. It is rare to find the leaves of the calyx stalked. They are usually sessile leaves, in which the vaginal portion is only slightly developed, and frequently the laminar part is alone present. Sepals are generally of a more or less oval, elliptical, or oblong form, with their apices either blunt or acute. In their direction they are erect or reflexed (with their apices downwards), spreading outwards {divergent or patulous), or arched inwards {connivent). They are usually of a greenish colour, and are called foliaceous or herbaceous; but sometimes they are coloured, as in the Fuchsia, Tropaeolum, Globe-flower, and Pomegranate, and are then called petaloid. Whatever be its colour, the external envelope of the flower must be considered as the calyx. The nature of the hairs on the calyx gives rise to terms similar to those already mentioned as applied to the surfaces of other parts of plants. The vascular bundles sometimes form a prominent rib, which indicates the middle of the sepal; at other times they form several ribs. The venation is useful as pointing out the number of leaves which constitute a gamosepalous calyx. In a polysepalous calyx the number of the parts is marked by Greek numerals prefixed. Thus, a calyx which has three sepals is trisepalous; one with five sepals is pentasepalous or pentaphyllous. The sepals occasionally are of different forms and sizes. In Aconit» one of them is shaped like a helmet, and has been callee galeate. In a gamosepalous calyx the sepals are united in various ways, sometimes very slightly, and their number is marked by the divisions at the apex. These divisions eitheι are simple projections in the form of acute or obtuse teeth; or they extend down the calyx as fissures about halfway, the calyx being trifid (three-cleft), quinquefid (five-cleft), (fig. 196), &c., according to their number; or they reach to near the base in the form of partitions, the calyx being tripartite, quaaripartite, quinquepartite, &c. The union of the parts may be complete, and the calyx may be quite entire or truncate, as in some Correas, the venation being the chief indication of the different parts. The cohesion is sometimes irregular, some parts uniting to a greater extent than others; thus a two-lipped or labiate calyx is formed, which, when the upper or posterior lip is arched, becomes ringent. The upper lip is often composed of three parts, which are thus posterior or next the axis, while the lower has two, which are anterior. The part formed by the union of the sepals is called the tube of the calyx; the portion where the sepals are free is the limb.

Occasionally, certain parts of the sepals undergo marked enlargement. In the Violet the calycine segments (laciniae) are prolonged downwards beyond their insertions, and in the Indian Cress (Tropaeolum) this prolongation is in the form of a spur (calcar), formed by three sepals; in Delphinium it is formed by one. When one or more sepals are thus enlarged, the calyx is calcarate or spurred. In Pelargonium the spur from one of the sepals is adherent to the flower-stalk. When an epicalyx is present, as in the Mallow order, the flower appears to be provided with a double calyx, and has been denominated caliculate. Degenerations take place in the calyx, so that it becomes dry, scaly, and glumaceous (like the glumes of grasses), as in the Rush tribe; hairy, as in Compositae; or a mere rim, as in some Umbelliferae and Acanthaceae, and in Madder (Rubia tinctorum, fig. 198), when it is called obsolete or marginate. In Diplolaena it is reduced to five scales. In Compositae,

Dipsacaceae, and Valerianaceae, the calyx is attached to the pistil, and its limb is developed in the form of hairs called pappus (fig.197). This pappus is either simple (pilose} or feathery (plumose). In Valeriana the superior calyx is at first an obsolete rim, but as the fruit ripens it is shown to consist of hairs rolled inwards, which expand so as to waft the fruit. The calyx sometimes falls off before the flower expands, as in Poppies, and is caducous (fig. 199); or along with the corolla, as in Ranunculus, and is deciduous; or it remains after flowering, as in Labiatae, Scrophulariaceae, and Boraginaceae; or its base only is persistent, as in Datura Stramonium. In Eschscholtzia and Eucalyptus the sepals remain united at the upper part, and become disarticulated at the base or middle, so as to come off in the form of a lid or funnel. Such a calyx is operculate or calyptrate. The existence or non-existence of an articulation determines the deciduous or persistent nature of the calyx. In the case of Eschscholtzia the axis seems to be prolonged so as to form a sort of tube, from which the calyx separates. In Eucalyptus the calyx consists of leaves, the laminae or petioles of which are articulated like those of the Orange, and the separation between the parts occurs at this articulation. The receptacle bearing the calyx is sometimes uuited to the pistil, and enlarges so as to form a part of the fruit, as in the Apple, Pear, Pomegranate, Gooseberry, <fcα In these fruits the withered calyx is seen at the apex. Sometimes a persistent calyx increases much after flowering, and encloses the fruit without being incorporated with it, becoming accrescent, as in various species of Physalis (fig. 200); at other times it remains in a withered or marcescent form, as in Erica; sometimes it becomes inflated or vesicular, as in Sea Campion (Silene maritima).

2. Corolla.

The corolla is the more or less coloured inner floral envelope, forming the whorl of leaves between the calyx and the stamens. It is generally the most conspicuous whorl. The gay colours and fragrant odours of flowers are resident in it. It is present in the greater number of Dicotyledons. Petals differ more from leaves than sepals do, and are much more nearly allied to the staminal whorl. In some cases, however, they are transformed into leaves, like the calyx, and occasionally leaf-buds are developed in their axil. They are seldom green, although occasionally that colour is met with, as in some Cobaeas, Hoya viridiflora, Gonolobus viridiflorus, and Pentatropis spiralis. As a rule they are highly coloured, the colouring matter being contained in cells, and differing in its nature from the chlorophyll of the leaves. As regards their structure petals consist of cellular tissue, traversed by true spiral vessels and thin-walled tubes. In delicate flowers, as Convolvulus and Anagallis, these vessels are easily seen under the microscope. Petals do not usually present numerous layers of cells like the leaves, neither is the epidermis always distinct, although in some instances it may be detached, especially from the surface next the calyx. The cuticle of the petal of a Pelargonium, when viewed with a ∣ or ∣-inch object glass, shows beautiful hexagons, the boundaries of which are ornamented with several inflected loops in the sides of the cells

On the outer surface of petals, corresponding to the lower side of leaves, stomata are sometimes found. Petals are generally glabrous or smooth; but, in some instances, hairs are produced on their surface. Petaline hairs, though sparse and scattered, present occasionally the same arrangement as those which occur on the leaves; thus, in Bombaceae they are stellate. Coloured hairs are seen on the petals of Menyanthes, and on the segments of the perianth of the Iris. Although petals are usually very thin and delicate in their texture, they occasionally become thick and fleshy, as in Stapelia and Rafllesia; or dry, as in Heaths; or hard and stiff, as in Xylopia. A petal often consists of two portions—the lower narrow, resembling the petiole of a leaf, and called the unguis or claw; the upper broader, like the blade of a leaf, and called the lamina or limb. These parts are seen in the petals of the Wallflower (fig. 201), where c is the claw and I the limb. The claw is often wanting, as in the Crowfoot (fig. 202) and the Poppy, and the petals are then sessile. Petals having a claw are unguiculate. According to the development of veins and the growth of cellular tissue, petals present varieties similar to those already noticed in the case of leaves. Thus the margin is either entire or divided into lobes or teeth. These teeth sometimes form a regular fringe round the margin, and the petal becomes fimbriated, as in the Pink; or laciniated, as in Lychnis Flos-cuculi; or crested, as in Polygala. Sometimes the petal becomes pinnatifid, as in Schizopetalum. The median vein is occasionally prolonged beyond the summit of the petals in the form of a long process, as in Strophanthus hispidus, where it extends for 7 inches; and at other times it ends in a free point or cuspis, and the petal becomes cuspidate; or the prolonged extremity is folded downwards or inflexed, as in Umbelliferae, so that the apex approaches the base. The limb of the petal may be flat or concave, or hollowed like a boat. In Hellebore the petals become folded in a tubular form, resembling a horn (fig. 203); in Aconite (fig. 205) some of the petals resemble a hollow-curved horn, supported on a grooved stalk; while in Columbine (fig. 204), Violet, Snapdragon, and Centranthus, one or all of them are prolonged in the form of a spur, and are calcarate. In Valeriana, Antirrhinum, and Corydalis, the spur is very short, and the corolla or petal is said to be gibbous, or saccate, at the base.

A corolla rarely consists of one petal, and when this occurs, as in Amorpha, it depends on the abortion or nondevelopment of others. Such a corolla is unipetalous, a term quite distinct from monopetalous. A corolla is dipetalous, tripetalous, tetr apetalous, or pentapetalous, according as it has two, three, four, or five separate petals. The general name of polypetalous, dialypetalous, eleuthero-petalous, or apopetalous, is given to corollas having separate petals, while monopetalous, gamopetalous, or sympetalous is applied to those in which the petals are united. This union generally takes place at the base, and extends more or less towards the apex; in Phyteuma the petals are united at their apices also. In some polypetalous corollas, as that of the Vine, the petals are separate at the base, and adhere by the apices. When the petals are equal as regards their development and size, the corolla is regular; when unequal, it is irregular. When a corolla is gamopetalous, it usually happens that the claws are united into a tube, while the upper parts are either free or partially united, so as to form a common limb, the point of union of the two portions being the faux or throat, which often exhibits a distinct constriction or dilatation. The number of parts forming such a corolla can be determined by the divisions, whether existing as teeth, crenations, fissures, or partitions, or if, as rarely happens, the corolla is entire, by the venation The union may be equal among the parts, or some may unite more than others.

Amongst regular polypetalous corollas may be noticed the rosaceous corolla (fig. 206), in which there are five spreading petals, having no claws, and arranged as in the single Rose and Potentilla; the caryophyllaceous corolla, in which there are five petals with long narrow tapering claws, as in many of the Pink tribe; the alsinaceous, where the claw is less narrow, and there are distinct spaces between the petals, as in some species of Chickweed; the cruciform, having four petals, often unguiculate, placed opposite in the form of a cross, as seen in Wallflower, and in other plants called cruciferous. Of irregular polypetalous corollas the most marked is the papilionaceous (fig. 187), in which there are five petals :—one superior (posterior), st, placed next to the axis, usually larger· than the rest, called the vexillum or standard; two lateral, a, the alee or wings; two inferior (anterior), partially or completely covered by the alae, and often united slightly by their lower margins, so as to form a single keel-like piece, car, called carina or keel, which embraces the essential organs. This form of corolla is characteristic of British Leguminous plants. Among the irregular polypetalous corollas might be included the orchideous (fig. 207), although it is really the perianth of a Monocotyledon. This perianth consists of three outer portions equivalent to the calyx, and three inner parts alternating with them, constituting the petals. The latter are often very irregular, some being spurred, others hooded, &c.; and there is always one, called the labellum or lip I, which presents a remarkable development, and.gives rise to many of the anomalous forms exhibited by these flowers.

Regular gamopetalous corollas are sometimes campanulate or bell-shaped, as in Campanula rotundifolia (fig. 208); infundibuliform or funnel-shaped, when the tube is like an inverted cone, and the limb becomes more expanded at the apex, as in Tobacco; hypocrateriform or salver-shaped, when there is a straight tube surmounted by a flat spreading limb, as in Primula (fig. 209); tubular, having a long cylindrical tube, appearing continuous with the limb, as in Spigelia and Comfrey; rotate or wheel-shaped, when the tube is very short, and the limb flat and spreading, as in Myosotis (when the divisions of the rotate corolla are very acute, as in Galium, it is sometimes called stellate or starlike); urceolate or um-shaped, when there is scarcely any limb, and the tube is narrow at both ends, and expanded in the middle, as in Bell-heath (Erica cinerea). Some of these forms may become irregular in consequence of certain parts being more developed than others. Thus, in Veronica, the rotate corolla has one division much smaller than the rest, and in Digitalis there is a slightly irregular campanulate corolla which some have called digitaliform. Of irregular gamopetalous corollas there may be mentioned the labiate or lipped (fig. 210), having two divisions of the limb in the form of what are called labia or lips (the upper one, u, composed usually of two united petals, and the lower, I, of three), separated by a hiatus or gap. In such cases the tube varies in length, and the parts in their union follow the reverse order of what occurs in the calyx, where two sepals are united in the lower lip, and three in the upper. When the upper lip of a labiate corolla is much arched, and the lips separated by a distinct gap, it is called ringent (fig. 210). The labiate corolla characterizes the natural order Labiatae. When the lower lip is pressed against the upper, so as to leave only a chink or rictus between them, the corolla is said to be personate or masklike, as in Frogsmouth, Snapdragon, and some other Scrophulariaceae, and the projecting portion of the lower lip is called the palate. In some corollas the two lips become hollowed out in a remarkable manner, as in Calceolaria, assuming a slipper-like appearance, similar to what occurs in the labellum of some Orchids, as Cypripedium. The calceolate corolla of Calceolaria may be considered as consisting of two slipper-like lips. When a tubular corolla is split in such a way as to form a strap-like process on one side with several tooth-like projections at its apex, it becomes ligulate, or strap-shaped (fig. 211).

This corolla occurs in many Composite plants, as in the florets of Dandelion, Daisy, and Chicory. The number of divisions at the apex indicates the number of united petals, some of which, however, may be abortive. Occasionally some of the petals become more united than others, and then the corolla assumes a bilabiate, or two-lipped form, as seen in the division of Compositae called Labiatiflorae.

In Grasses and Sedges, in place of verticillate leaves forming the flower, there are alternate scales or glumes. The flow ers of Grasses usually occur in spikelets (fig. 212), which consist of one or two glumes a, covering several flowers b. The spikelets are associated in spikes or panicles. In Wheat these spikelets are arranged alternately along a common rachis. Each spikelet consists of two empty glumes a, a, having the form represented in fig. 212, and enclosing flowers composed of scales (paleoe or glumellce), which are placed at different heights in an alternate manner. In the flower of the Oat (fig. 213), after removing the outer pale or glumelloe, the inner one pi is seen with two scales (lodiculoe or squamee) sq, at the base, enclosing the essential organs of reproduction. The paleae of grasses are called by some flowering glumes, while hypogynous scales (lodiculae) within them are considered as the rudimentary perianth. In Wheat (Triticum) there are two empty glumes and two flowering glumes, or, as some say, one flowering glume and one pale. In the Oat (Avena) there are two empty glumes, usually three flowering glumes with awns, and two lodicules representing the perianth. In Sedges (Carices) the male flowers are borne on scales, and so are the female, as shown in fig. 214, in which the scale « is placed on one side. Within the scale the female flower is situated, having a peculiar baglike covering u, termed perigynium.

The parts of the corolla frequently adhere to those of the calyx,. and any change in the latter causes also an alteration in the former. Petals are sometimes suppressed, and sometimes the whole corolla is absent. In Amorpha and Afzelia the corolla is reduced to a single petal, and in some other Leguminous plants it is entirely wanting. In the natural order Ranunculacese, some genera, such as Ranunculus, Globe-flower, and Paeony, have both calyx and rîTii 3 ’ Λ sr kilθ others, such as Clematis, Anemone, and Caltha, have only a coloured calyx. Flowers become double by the multiplication of the parts of the corolline whorl. This arises in general from a metamorphosis of the stamens. Union of separate flowers (synanthos) occasionally occurs, and the adhesion which thus takes place causes various changes in the whorls. Flowers may be united by their peduncles, as seen in some anomalous specimens of Dandelion, Hyacinth, and Centaury; or by their outer parts, such as the calyx. At other times there is a complete fusion, as it were, of all the parts of the flowers, some continuing normal, others being suppressed or abortive.

Certain abnormal appearances occur in the petals of some flowers, which received in former days the name of nectaries. The term nectary was very vaguely applied by Linnaeus to any part of the flower which presented an unusual aspect, as the crown (corona) of Narcissus, the fringes of the Passion-flower, &c. If the name is retained rt ought properly to include only those parts which secrete a honey-like matter, as the glandular depression at the base of the perianth of the Fritillary, or on the petal of Ranunculus, or on the stamens of Rutaceae. The honey secreted by flowers attracts insects, which, by conveying the pollen to the stigma, effect fertilization. What have usually, however, been called nectaries, are mere modifications of somo part of the flower, especially of the corolla and stamens, produced either by degeneration or outgrowth, or by dilamination, chorisis, or deduplication. Of this nature are the scales on the petals in Lychnis, Silene, Cynoglossum, and Ranunculus (fig. 202). Others consider these outgrowths of the petal to be formed in the same way as the ligules of Grasses. Corollas having these scaly appendages are sometimes denominated appendiculate. In Cuscuta and Samolus the scales are alternate with the petals, and may represent altered stamens. The parts formerly called nectaries are mere modifications of the corolla or stamens. Thus the so-called horn-like nectaries under the galeate sepal of Aconite (fig. 205-) are modified petals, so also are the tubular nectaries of Hellebore (fig. 203). The nectaries of Menyanthes and of Iris consist of hairs developed on the petals. Those oi Parnassia and the Passion-flower, Stapelia, Asclepias, and Canna, are fringes, rays, and processes, which are probably modifications of stamens; and some consider the crown of Narcissus as consisting of a membrane similar to that which unites the stamens in Pancratium. It is sometime3 difficult to say whether these nectaries are to be referred to the corolline or to the staminal row. The paraphyses of the Passion-flower, the crown of Narcissus, and the coronet of Stapelia are referred sometimes to the one and sometimes to the other. In general, they may be said to belong to that series with which they are immediately connected. Some have given names indicating the parts of which they are modifications, by prefixing the term para, using such terms as paracorolla and parastemones.

Petals are attached to the axis usually by a narrow base, but occasionally the base is larger than the limb, as in the Orange flower. When this attachment takes place by an articulation, the petals fall off either immediately after expansion (caducous), or after fertilization (^deciduous). A corolla which is continuous with the axis and not articulated to it, as in Campanula and Heaths, may be persistent, and remain in a withered or marcescent state while the fruit is ripening. A gamopetalous corolla falls off in one piece; but sometimes the base of the corolla remains persistent, as in Rhinanthus and Orobanche,.

II. Essential Organs.

These organs are the stamens and the∕>⅛tiζ the latter containing the seeds or germs of young plants, and corresponding to the female, wdιile the former produces a powder necessary for fecundation, and is looked upon as performing the part of the male. The presence of both is required in order that perfect seed may be produced. A flower may have a calyx and corolla, but it will be imperfect if the essential organs are not present. The name of hermaphrodite or bisexual is given to flowers in which both these organs are found; that of unisexual or diclinous to those in which only one of these organs appears,—those bearing stamens only, being staminiferous or male; those having the pistil only, pistilliferous or female. But even in plants with hermaphrodite flowers it is rare that self-fertilization take3 place, and this is provided against by the structure of the parts or by the period of ripening of the organs. For instance, in Primula and Linum some flowers have long stamens and a pistil with a short style, the others having short stamens and a pistil with a long style. The former occur in what are called thrum-eyed Primroses, the latter in those called pin-eyed. Such plants are called dimorphic. Other plants are trimorphic i as species of Lythrum, and proper fertilization is only effected by combination of parts of equal length. In some plants the stamens are perfected before the pistil; they are called proterandrous, as in Ranunculus repens, Silene maritima, Zea Mays. In other plants the pistil is perfected before the stamen3, as in Potentilla argentea, Plantago major, Coix Lachryma, and they are }>roterogynous plants. Plants in which proterandry or proterogyny occurs are called dichogamous. When in the same plant there are unisexual flowers, both male and female, the plant is said to be monoecious or monoicoτιs, as in the Hazel and Castor-oil plant. This is indicated by the symbol Ô - When the male and female flowers of a species are found on separate plants, the term dioecious or dioicous is applied, as in Mercurialis and Hemp, and the symbol J: 5 is used; and when a species has male, female, and hermaphrodite flowers on the same or different plants, as in Parietaria, it is polygamous, for which J φ Ç is the symbol.

1. Male Organs in Phanerogams.

The stamens (stamina) arise from the thalamus or torus within the petals, with -which they alternate, forming one or more verticils or whorls, which collectively constitute the androecium, or the male organs of the plant. Their normal position is below the inner whorl or thé pistil, and when they are so placed (tig. 215, e) upon the thalamus they are hypogynous. Sometimes they become adherent to the petals, or are epipetalous, and the insertion of both is looked upon as similar, so that they are still hypogynous provided they are independent of the calyx and the pistil. Frequently the margins of the thalamus bearing the floral envelopes and stamens elongates, and the gynoecium remains in the centre of the concave receptacle; thus the stamens as it were rise from the calyx, and they surround the ovary, and are perigynous; but when the ovary becomes completely inferior by the growing upwards and inwards of the receptacle, the parts of the flower rising from its summit, the stamens are epigynous (fig. 216). Numerous intermediate forms occur, especially amongst Saxifragaceae, where the parts are half superior or half inferior. In the Orchis tribe, where the stamens become adherent to the pistil so as to form a column, the flowers are said to be gynandrous. The same is the case in Aristolochia (fig. 217). These arrangements of parts are of great importance in classification. The stamens vary in number, from one to many hundreds. Like the other parts of the flower they are modified leaves, resembling leaves in their structure, development, and arrangement. They consist of cellular and vascular tissues. They appear at first in the form of cellular projections, and are arranged in a more or less spiral form. In their general aspect they have a greater resemblance to petals than to the leaves, and there is often seen a gradual transition from petals to stamens, especially in spiral flowers, as Nymphaea alba. When flowers become double by cultivation, the stamens are converted into petals, as in the Paeony, Camellia, Rose, <tc. When there 's or.ly one whorl the stamens are usually equal in number to the sepals or petals, and are arranged opposite to the former, and alternate with the latter. The flower is then isostemonous. When the stamens are not equal in number to the sepals or petals, the flower is anisostemonous. When there is more than one whorl of stamens, then the parts of each successive whorl alternate with those of the whorl preceding it. The staminal row is more liable to multiplication of parts than the outer whorls. A flower with a single row of stamens is aplostemonous. If the stamens are double the sepals or petals as regards number, the flower is diplostemonous; if more than double, polystemonous. The additional rows of stamens may be developed in the usual centripetal (acropetal) order, as in Rhamnaceae; or they may be interposed between the pre-existing ones or be placed outside them, i.e., develop centrifugally (basipetally), as in Geranium and Oxalis. When the stamens are neither equal to, nor a multiple of, the floral envelopes, but are less numerous, the flower is miostemonous. When the stamens are fewer than twenty they are said to be definite, and the flower is oligandrous; when above twenty they are indefinite or polyandrous, and are represented by the symbol ∞. The number of stamens is indicated by the Greek numerals prefixed to the term androus; thus a flower with [table]

The simplest form of stamens is seen in Cycadaceous plants, where cataphyllary leaves (the scales of the cone) are the staminal leaves, bearing the pollen-sacs scattered over their under surface. The stamen usually consists of two parts, a contracted portion, often thread-like, termed the filament (fig. 175, ∕), and a broader portion, usually of two lobes, termed the anther (fig. 175, a), containing a powdery matter, called pollen (p), and supported upon the end of the filament. That portion of the filament in contact with the anther lobes is termed the connective. The exact morphological relations of these parts is not definitely settled. Some consider the filament as representing the petiole, and the anther as being the folded lamina; others again regard the connective as representing the lamina, the rest of the filament being the petiole, and the anther lobes as mere appendages of the lamina. The filament is no more essential to the stamen than the petiole is to the leaf, or the claw to the petal. If the anther is absent the stamen is abortive, and cannot perform its functions. The anther is developed before the filament, and when the latter is not produced, the anther is sessile, or has no stalk, as in the Mistleto.

The filament, when structurally considered, is found to consist of a thin epidermis, on which occasionally stomata and hairs occur, and a layer of cellular tissue enclosing a bundle of spiral vessels, which traverses its whole length, and terminates at the junction between the filament and the anther. The filaments of Callitriche verna are said to have no vessels. The filament is usually, as its name imports, filiform or thread-like, cylindrical, or slightly tapering towards its summit. It is often, however, thickened, compressed, and flattened in various ways, becoming petaloid in Canna, Maranta, Nymphaea alba (fig. 179); subulate or slightly broadened at the base, and drawn out into a point like an awl, as in Butomus umbellatus; clavate, that is, narrow below and broad above, as in Thalictrum. In some instances, as in Tamarix gallica, Peganum Harmala, and Campanula, the base of the filament is much dilated, and ends suddenly in a narrow thread-like portion. In these cases the base may represent the sheath or vagina of the leaf, and may give off stipulary processes in a lateral direction, as in Allium and Alyssum calycinum. The filament varies much in length and in firmness. The length sometimes bears a relation to that of the pistil, and to the position of the flower, whether erect or drooping. The filament is usually of sufficient solidity to support the anther in an erect position; but sometimes, as in Grasses, it is very delicate and capillary, or hair-like, so that the anther is pendulous (fig. 212, s). The filament is generally continuous from one end to the other, but in some cases it is bent or jointed, becoming geniculate; at other times, as in the Pellitory, it is spiral. It is frequently colourless; but, in many instances, it exhibits different colours. In Fuchsia and Poinciana, it is red; in Adamia and Tradescantia virginica, blue; in Oenothera and Ranunculus acris, yellow.

Hairs, scales, teeth, or processes of different kinds are sometimes developed on the filament. In Tradescantia virginica, or Spiderwort, the hairs are beautifully coloured, moniliform or necklace-like, ⅛rd∖, rotation is seen in them (p. 84). Such a filament is bearded or stupose. At the base of the filament glandular or scaly appendages are occasionally produced, either on its internal or its external surface. These may be either parts of a whorl, to be afterwards noticed under the name of the disk, or separate prolongations from the filament itself, which is hence called appendiculate, or sometimes strumose. Filaments are usually articulated to the thalamus or torus, and the stamens fall off after fertilization; but in Campanula and some other plants they are continuous with the torus, and the stamens remain persistent, although in a withered state. Changes are produced in the whorl of stamens by cohesion of the filaments to a greater or less extent, while the anthers remain free; thus, all the filaments of the androecium may unite, forming a tube round the pistil, or a central bundle when the pistil is abortive, the stamens becoming monadelphous, as occurs in plants of the Mallow tribe; or they may be arranged in two bundles, the stamens being diadelphous, as in Polygala, Fumaria, and Pea; in this case the bundles may be equal or unequal. It frequently happens, especially in Papilionaceous flowers, that out of ten stamens nine are united by their filaments, while one (the posterior one) is free (fig. 218). When there are three or more bundles the stamens are triadelphous, as in Hypericum aegyptiacum, or polyadelphous, as in Luhea paniculata and Ricinus communis. In some cases, especially in the Mallow tribe, the stamens cohere, having been originally separate, but in most other cases each bundle is produced by the branching of a single stamen. When there are three stamens in a bundle we may conceive the lateral ones as of a stipulary nature. In Lauraceae there are perfect stamens, each having at the base of the filament two abortive stamens or staminodes, which may be analogous to stipules. Filaments sometimes are adherent to the pistil, forming a columna or column, as in Stylidium, Asclepiadaceae, Rafflesia, Orchidaceae, and Aristolochiaceae (fig. 217). The column is called gynostemium,, and the flowers are denominated gynandrous.

The anther consists of lobes containing minute powdery matter, called pollen, which, when mature, is discharged by a fissure or opening of some sort. The anther-lobes are considered by some as formed by the two halves of the lamina, their back corresponding to the under surface, and their face to the upper surface, united by the midrib, the pollen being the cellular tissue, and the fissure of the anther taking place at the margin, which, however, is often turned towards the face. In this view, the two cavities which are found to exist in each lobe may correspond with the upper and under layer of cells, separated by a septum equivalent to the fibro-vascular layer of the leaf. Others view the anther as formed by each half of the lamina being folded upon itself, so that the outer surface of both face and back corresponds to the lower side of the leaf, and the septum dividing each cavity into two is formed by the united upper surfaces of the folded half. Again, others hold that the connective represents the lamina of the leaf to which the anther lobes are mere appendages. There is a double covering of the anther—the outer, or exothecium, resembles the epidermis, and often presents stomata and projections of different kinds (fig. 219); the inner, or endothecium, is formed by a layer or layers of cellular tissue (fig. 219, cf), the cells of which have a spiral, annular, or reticulated fibre in their interior. This internal lining varies in thickness, generally becoming thinner towards the part where the anther opens, and there disappears entirely. The walls of the cells are frequently absorbed, so that when the anther attains maturity the fibres are alone left, and these by their elasticity assist in discharging the pollen. The anther is developed before the filament, and is always sessile in the first instance, and sometimes continues so. It appears at first as a simple cellular papilla of meristem, upon which an indication of two lobes soon appears. Upon these projections the rudiments of the pollen-sacs are then seen, usually four in number, two on each lobe. In each a differentiation takes place in the layers beneath the epidermis, by which an outer layer of small-celled tissue surrounds an inner portion of large cells. Those central cells are the mothercells of the pollen, whilst the small-celled layer of tissue external to them becomes the endothecium, the exothecium being formed from the epidermal layer.

In the young state there are usually four pollen-sacs or cavities produced, two for each anther-lobe, and when these remain permanently complete, it is a quadrilocular or tetrathecal anther (fig. 220). The four cavities are sometimes placed in apposition, as in Poranthera, and Tetratheca juncea, and at other times two are placed above and two below, as in Persea gratissima. Sometimes, however, only two cavities remain in the anther, by union of the sacs in each lobe, in which case the anther is said to be bilocular or dithecal. Sometimes the anther has a single cavity, and becomes unilocular, or monothecal, or dimidiate, either by the disappearance of the partition between the two lobes, or by the abortion of one of its lobes, as in Styphelia laeta and Althaea officinalis. Occasionally there are numerous cavities in the anther, as in Viscum and Rafflesia. The form of the anther-lobes varies. They are generally of a more or less oval or elliptical form, or they may be globular, as in Mercurialis annua; at other times linear or clavate, curved, flexuose, or sinuose or anfractuose, as in Bryony and Gourd. According to the amount of union of the lobes and the unequal development of different parts of their surface an infinite variety of forms are produced. That part of the anther to which the filament is attached, and which is generally towards the petals, is the back, the opposite being the face. The division between the lobes is marked on the face of the anther by a groove or furrow, and there is usually on the face a suture, indicating the line where the membranous coverings open to discharge the pollen. The suture is often towards one side in consequence of the valves being unequal. The stamens may cohere by their anthers, and become syngenesious or synantherous, as in Composite flowers, and in Lobelia, J asione, Viola, Ac.

The anther-lobes are united to the connective, which may be either continuous with the filament or articulated with it. It consists of parenchyma, in which the spiral vessels of the filament terminate. From the connective a partition or septum extends across each antherine loculus, dividing it either partially or completely. The septum sometimes reaches the suture When the filament is continuous with the connective, and is prolonged so that the anther-lobes appear tobe united to it throughout their whole length, and lie in apposition to it and on both sides of it, the anther is said to be adnate or adherent; when the filament ends at the base of the anther, then the latter is innate or erect. In these cases the anther is to a greater or less degree fixed. When, however, the attachment is very narrow, and an articulation exists, the anthers are movable, and are easily turned by the wind. This is well seen in what are called versatile anthers, as in Tritonia, Grasses, &c. (fig. 212), where the filament is attached only to the middle of the connective; and it may occur also in cases where it is attached to the apex, as in pendulous anthers. The connective may unite tħe ant∩er-lobes completely, or only partially. It is sometimes very short, and is reduced to a mere point, so that the lobes are separate or free. At other times it is prolonged upwards beyond the lobes, assuming various forms, as in Acalypha, and Nerium Oleander; or it is extended backwards and downwards, as in Violet (fig. 221), and Ticorea febrifuga. In Salvia officinalis the connective is attached to the filament in a horizontal manner, so as to separate the two anther-lobes, and then it is called distractile (fig. 222); one of the lobes only being perfect or fertile, containing pollen, the other imperfectly developed and sterile. In Stachys the connective is expanded laterally, so as to unite the bases of the antherine lobes, and bring them into a horizontal line.

The opening of the anthers to discharge their contents is denominated dehiscence. This takes place either by clefts, by hinges, or by pores. When the anther-lobes are erect, the cleft takes place lengthwise along the line of the suture, constituting longitudinal dehiscence (fig. 175). At other times the slit takes place in a horizontal manner, from the connective to the side, as in Alchemilla arvensis (fig. 223) and in Lemna; the dehiscence is then transverse. When the anther-lobes are rendered horizontal by the enlargement of the connective, then what is really longitudinal dehiscence may appear to be transverse. In other cases, when the lobes are united at the base, the fissure in each of them may be continuous, and the two lobes may appear as one. The cleft does not always proceed the whole length of the anther-lobe at once, but often for a time it extends only partially. In other instances the opening is confined to the base or apex, each loculament opening by a single pore, as in Pyrola, Tetratheca juncea, Rhododendron, Vaccinium, and Solanum (fig. 224), where there are two, and Poranthera, where there are four; whilst in the Mistleto the anther has numerous pores for the discharge of the pollen. Another mode of dehiscence is called hinged. In the Barberry (fig. 225) each lobe opens by a valve on the outer side of the suture, separately rolling up from base to apex; in some of the Laurel tribe there are two such valves for each lobe, or four in all. This may be called a combination of transverse and hinged dehiscence. In some Guttiferae, as Hebradendron cambogioides (the Ceylon Gamboge plant), the anther opens by a lid separating from the apex; this is called circumscissile dehiscence.

The anthers dehisce at different periods during the process of flowering; sometimes in the bud, but more commonly when the pistil is fully developed, and the flower is expanded. They either dehisce simultaneously or in succession. In the latter case individual stamens may move in succession towards the pistil and discharge their contents, as in Parnassia palustris, or the outer or the inner stamens may first dehisce, following thus a centripetal or centrifugal order. The anthers are called introrse or anticoe, when they dehisce by the surface next to the centre of the flower; they are extrorse, or posticce, when they dehisce by the outer surface; when they dehisce by the sides, as in. Iris and some grasses, they are laterally dehiscent. Sometimes, from their versatile nature, anthers originally introrse become extrorse, as in the Passionflower and Oxalis.

The usual colour of anthers is yellow, but they present a great variety in this respect. They are red in the Peach, dark purple in the Poppy and Tulip, orange in Eschscholtzia, <fcc. The colour and appearance of the anthers often change after they have discharged their functions.

Stamens occasionally become sterile by the degeneration or non-development of the anthers. Such stamens receive the name of staminodia, or rudimentary stamens. In Scrophularia the fifth stamen appears in the form of a scale; and in many Pentstemons it is reduced to a filament with hairs or a shrivelled membrane at the apex. In other cases, as in double flowers, the stamens are converted into petals; this is also probably the case with such plants as Mesembryanthemum, where there is a multiplication of petals in several rows. Sometimes, as in Ganna, one of the anther-lobes becomes abortive, and a petaloid appendage is produced. Stamens vary in length as regards the corolla. Some are enclosed within the tube of the flower, as in Cinchona, and are called included; others are exserted, or extend beyond the flower, as in Littorella, Plantago, and Exostemma. Sometimes the stamens in the early state of the flower project beyond the petals, and in the progress of growth become included, as in Geranium striatum. Stamens also vary in their relative lengths as respects each other. When there is more than one row or whorl in a flower, those on the outside are sometimes longest, as in many Rosaceae; at other times those in the interior are longest, as in Luhea. When the stamens are in two rows, those opposite the petals are usually shorter than those which alternate with the petals. It sometimes happens that a single stamen is longer than all the rest. A definite relation, as regards number, sometimes exists between the long and the short stamens. Thus, in some flowers the stamens are didynamous, having only four out of five stamens developed, and the two corresponding to the upper part of the flower longer than the two lateral ones. This occurs in Labiatae and Scrophulariaceae (fig. 226). Again; in other cases, there are six stamens, whereof four long ones are arranged in pairs opposite to each other, and alternate with two isolated short ones (fig. 227), giving rise to tetradynamous flowers, as in Cruciferae. Stamens, as regards their direction, may be erect, turned inwards, outwards, or to one side. In the last-mentioned case they are called declinate, as in Amaryllis, Horse-chestnut, and Fraxinella.

The pollen or powdery matter contained in the anther consists of small cells developed in the interior of other cells. These are produced in the large thick-walled mother cells or pollen utricles, formed in the interior of the pollen-sacs of the young anther. These mother-cells are either separated from one another and float in the granular fluid which fills up the cavity of the pollen-sac, or are not so isolated. A division takes place, by which four cells are formed in each, the exact mode of division differing in Dicotyledons and Monocotyledons. These cells are the pollen-cells or grains. They increase in size and acquire a cell-wall, which becomes differentiated into an outer, cuticular layer, or extine, and an inner layer, or intine. Then the walls of the mother-cells are absorbed, and the pollen-grains float freely in the fluid of the pollen-sacs, which gradually disappears, and the mature grains form a powdery mass within the anther. They then either remain united in fours, or multiples of four, as in some Acacias, Periploca graeca, and Inga anomala, or separate into individual grains, which by degrees become mature pollen. Occasionally the membrane of the pollen-cell is not completely absorbed, and traces of it are detected in a viscous matter surrounding the pollen-grains, as in Onagraceae. In Orchidaceous plants the pollen-grains are united into masses, or pollinia (fig. 228), by means of viscid matter.

In Orchids each of the pollen-masses has a prolongation or stalk, called a caudicle, which adheres to a prolongation at the base of the anther, called rostellum, by means of a viscous gland, called retinaculum. The gland is either naked or covered. The term clinandrium is sometimes applied to the part of the column in Orchids where the stamens are situated. In some Orchids, as Cypripedium, the pollen has its ordinary character of separate grains. The number of pollinia varies; thus, in Orchis there are usually two, in Cattleya four, and in Laelia eight. The two pollinia in Orchis Morio contain each about 200 secondary smaller massée. These small masses, when bruised, divide into grains which are united in fours. In Asclepiadaceae the pollinia are usually united in pairs (fig. 229), belonging to two contiguous antherine cavities,— each pollen-mass having a caudicular appendage, ending in a common gland, by means of which they are attached to a process of the stigma. The pollinia are also provided with an appendicular staminal covering (fig. 230).

When mature, the pollen-grain is a cellular body having an external covering, extine, and an internal, intine. Fritzsche states that he has detected, in some cases, other two coverings, which he calls intextine and exintine. They occur between the extine and intine, and are probably formed by foldings of these membranes. In some aquatics, as Zostera marina, Zannichellia pedunculata, Naias minor, <fcc., only one covering exists, and that is said to be the intine. The extine is a firm membrane, which defines the figure of the pollen-grain, and gives colour to it. It is either smooth, or covered with numerous projections (fig. 231), granules, points, minute hairs, or crested reticulations. The colour is generally yellow, and the surface is often covered with a viscid or oily matter. The intine is uniform in different kinds of pollen, thin and transparent, and possesses great power of extension.

Pollen-grains vary from 7r ⅜θ∙ to τ ⅜θ of an inch or less in diameter. Their forms are various. The most common form of grain is ellipsoidal, more or less narrow at the extremities, which are called its poles, in contradistinction to a line equidistant from the extremities, and which is its equator. Pollen-grains are also spherical; cylindrical and curved, as in Tradescantia virginica; polyhedral in Dipsacaceae and Compositae: nearly triangular in Proteaceae and Onagraceae (fig. 232). The surface of the pollen-grain is either uniform and homogeneous, or it is marked by folds dipping in towards the centre, and formed by thinnings of the membrane. In Monocotyledonous plants there is usually a single fold; in Dicotyledons, often three. Two, four, six, and even twelve folds are also met with. There are also rounded portions of the membrane or points of perforation or pores visible in the pollen-grain. These vary in number from one to fifty. In Monocotyledons, as in Grasses, there is often only one, while in Dicotyledons they number from three upwards. When numerous, these points are either scattered irregularly, or in a regular order, frequently forming a circle round the equatorial surface. Sometimes at the place where they exist, the outer membrane, in place of being thin and transparent, is separated in the form of a lid, thus becoming operculate, as in the Passion-flower and Gourd. Grains of pollen have sometimes both folds and pores. Within the pollengrain a granular semifluid matter called f ovilla is contained, along with some oily particles, and occasionally starch. The fovilla contains small spherical granules, sometimes the ¾¾o go ∙ of an inch in diameter, and large ellipsoidal or elongated corpuscles, which exhibit molecular movements under the microscope. The application of moisture affects very markedly the pollen-grain, causing it to swell up by endosmose. If the moisture be long applied the distension becomes so great as to rupture the extine irregularly if it is homogeneous, or to cause projections and final rupture at the folds or pores when they exist. The intine, from its distensibility, is not so liable to rupture; it is often forced through the ruptured extine, or through the pores, in the form of small sac-like projections. The internal membrane ultimately -gives way, and allows the granular fovilla to escape (fig. 233). If the fluid is applied only to one side of the pollen-grain, as when the pollen is applied to the pistil, the distension goes on more slowly, and the intine is prolonged outwards like a hernia, and forms an elongated tube called a pollen-tube (fig. 232). This tube, at its base, is often covered by the ruptured extine, and it contains in its interior fovilla-granules. The number of pollen-tubes which may be produced depends on the number of pores. In some pollinia the number of tubes which are formed is enormous. Thus, it has been calculated that two pollenmasses of Orchis Morio may give out 120,000 tubes.

When the pollen-grains are ripe, the anther dehisces (figs. 175,223), and the pollen is shed. In order that fertilization may be effected the pollen must be conveyed to the female organ of reproduction. This process, termed pollination, is promoted in various ways,—the whole form and structure of the flower having relation to the process. In some plants, as Kalmia and Pellitory (fig. 234), the mere elasticity of the filaments is sufficient to effect this; in other plants pollination is effected by the wind, as in Coniferae, and in such cases enormous quantities of pollen are produced. These plants are anemophilous. But the common agents for pollination are insects. To allure and attract them to visit the flower the odoriferous secretions and gay colours are developed, and the position and complicated structure of the parts of the flower are adapted to the perfect performance of the process. It is rare in hermaphrodite flowers for self-fertilization to occur, and the various forms of dichogamy, dimorphism, and trimorphism are fitted to prevent this.

Under the term dish is included every structure intervening between the stamens and the pistil. It was to such structures that the name of nectary was applied by old authors. It presents great varieties of form, such as a ring, scales, glands, hairs, petaloid appendages, &c., and in the progress of growth it often contains saccharine matter, thus becoming truly nectariferous. The disk is frequently formed by degeneration or transformation of the staminal row. It may consist of processes rising from the torus, alternating with the stamens, and thus representing an abortive whorl; or its parts may be opposito to the stamens, as in Crassula rubens (fig. 172). In some flowers, as Jatropha Curcas, in which the stamens are not developed, their place is occupied by glandular bodies forming the disk. In Gesneraceae and Cruciferae the disk consists of tooth-like scales at the base of the stamens (fig. 227). The parts composing the disk sometimes unite and form a glandular ring, as in the Orange; or they form a dark-red lamina covering the pistil, as in Paeonia Moutan (fig. 235); or a waxy lining of the calyx tube or hollow receptacle, as iu the Bose; or a swelling at the top of the ovary, as in Umbelliferιe, in which the disk is said to be epigynous. The enlarged t0ru3

covering the ovary in Nymphaea and Nelumbium may be regarded as a form of disk.

2. Male Organs in Cryptogams.

Sexual organs have not as yet been demonstrated in all Cryptogamie plants; but in most of them certain structures representing the male organs have been found. These are termed Antheridia or Pollinodia (figs. 236, 250). They are closed sacs of various forms,—rounded, ovate, oblong, clavate, flask-like, <fcc.,—developed from various parts of the plants, and composed either of one or of many cells enclosing a single central cell. In the interior of this organ small cells (fig. 236, c) are formed of varying shape, containing in their interior peculiar bodies, termed antherozoids, spermatozoïde, or phytozoa (fig. 237). These are in most cases filamentous spirally coiled cells, frequently with cilia attached, by means of which they move rapidly through the water. In some cases there are no cilia, as in the Λlgal group of Floridese; and the autherozoids of all Algae differ from those of other Cryptogams in never being filamentous, but short and more or less rounded. The amount of twisting of the antherozoid varies; in some Rhizocarps there are as many as a dozen coils. Antheridia of this kind have been demonstrated in all vascular Cryptogams, but only iu some Thallogens have such structures been seen.

In Characeae, however, the antheridium, which is here termed globule (fig. 240), has a peculiar structure, differing from all other Cryptogams. It is a globular case, formed by eight flat cells, of which four are quadrangular aud four are triangular; and these are folded at the margins so that the edges of contiguous plates dovetail (fig. 241). These are termed shields. Into the cavity of the globule the terminal cell of the stalk supporting it projects. From the centre of each shield projects on the inner surface a large oblong cell or manubrium, which in turn bears at its apex a roundish cell termed the head. From each head-cell six secondary head-cells project, from each of which pass off four slender separate filaments (fig. 242) containing antherozoids.

When mature the antheridium is ruptured, and the mother-cells of the antherozoid escape and discharge the antherozoids into the surrounding medium which is always water—fertilization only being effected in water, by which the antherozoids are carried to the female organ. In some cases, however, among Thallogens the antheridium actually penetrates into the female organ.

Antheridia are produced on various parts of plants. Where they have been found amongst Thallogens, they arise from definite portions of the cellular thallus. In thalloid Hepaticae they may be sunk in the substance of the thallus, or they may be imbedded in the upper surface of a peltate stalked disk rising from the thallus, as in Marchantia (figs. 243,244). In Mosses they may be seated within the same whorl of leaves or perichaetium as the female organs, or they may be separate and surrounded by a perigone; but they are always formed upon the leaf-bearing axis, either terminal or lateral. In Ferns and Equisetum they are produced on the surface—usually the lower surface—of a cellular flattened expansion, termed a prothallus (figs. 245, 250), in close proximity to the female organs. In Characeae they always arise from the termination of one of the whorled leaflets.

In some Lycopodiaceae no antheridia have as yet been found; but in Selaginella, Isoetes, and in the Rhizocarpeae, hollow sacs, termed microsporangia (figs. 246, 248), are produced in the axil of certain leaves of the plant, in which small cells, microspores (fig. 249), are found, each consisting of an outer covering or exospore, and an inner or endospore. Within those microspores (the manner varying iu different genera) smaller cells are formed, which are the mother-cells of the antherozoids. In this development of Lycopodiaceae we have parts which are analogous to those of the stamens in Phanerogams. The microsporangium evidently corresponds to the pollen-sac, and the microspores are the equivalents of the pollengrains,—the forms seen in such Gymnosperms as Cycadaceae being the link connecting them with the highest forms.

3. Female Organs in Phanerogams.

The pistil or gynoecium occupies the centre or axis of the flower, and is surrounded by the stamens and floral envelopes when these are present. It constitutes the innermost whorl, and is the female organ of the plant, which after flowering is changed into the fruit, and contains the seeds. It consists essentially of two parts, a basal portion forming a chamber, the ovary or germen, containing ovules or young seeds attached to a part called the placenta, and an upper portiou, the stigma, a cellular secreting body, which is either seated on the ovary, and is then called sessile, as in the Tulip and Poppy, or is elevated on a stalk called the style, interposed between the ovary and stigma. The style is not necessary for the perfection of the pistil. Like the other organs, the pistil consists of one or more modified leaves, which in this instance are called carpels. When a pistil consists of a single carpel it is simple or monocarpellary (fig. 253), a state usually depending on the nondevelopment of other carpels. When it is composed of several carpels, more or less united, it is compound or polycarpellary. In the first-mentioned case the terms carpel and pistil are synonymous. Each carpel has its own ovary, style (when present), and stigma, and is formed by a folded leaf, the upper surface of which is turned inwards towards the axis, and the lower outwards, while from its margins are developed one or more buds called omdes. That this is the true nature of the pistil may be seen by examining the flower of the doubleflowering Cherry. In it no fruit is produced, and the pistil consists of sessile leaves (fig. 251), the limb of each being green and folded, with a narrow prolongation upwards, as if from the midrib, and ending in a thickened portion. The ovary then represents the limb or lamina of the leaf, and is composed of cellular tissue with fibro-vascular bundles, and an epidermal covering. The cellular tissue, or parenchyma, often becomes much developed, as will be seen particularly when fleshy fruits are considered. The outer epidermis corresponds to the lower side of the leaf, exhibiting stomata, and sometimes hairs; the inner surface represents the upper side of the leaf, being usually very delicate and pale, and forming a layer called sometimes epithelium, which does not exhibit stomata. The vascular bundles correspond with the veins of the leaf, and consist of spiral, annular, and other vessels. Iu Cycas the carpels are ordinary leaves, with ovules upon their margin.

A pistil is usually formed by more than one carpel. The carpels may be arranged like leaves, either at the same or nearly the same height in a verticil, or at different heights in a spiral cycle. When they remain separate and distinct, thus showing at once the composition of the pistil, as in Caltha, Ranunculus, Hellebore, and Spiraea (fig. 252), the term apocarpous is applied. Thus, in Crassula rubens (fig. 172) the pistil consists of five verticillate carpels o, alternating with the stamens e. In the Tulip-tree and Ranunculus (fig. 254) the separate carpels are numerous and are arranged in a spiral cycle upon an elongated axis or receptacle. In the Raspberry the carpels are on a conical receptacle; in the Strawberry, on a swollen succulent one (fig. 154); and in the Rose (fig. 155), on a hollow one. When the fruit consists of several rows of carpels, the innermost have their margins directed to the centre, while the margins of the outer rows are arranged on the back of the inner ones. When the carpels are united, as in the Pear, Arbutus, and Chickweed, the pistil becomes syncarpous. The number of carpels in an apocarpous pistil, or the number of separate styles in a syncarpous one, is indicated in the following way :— [table]

The union in a syncarpous pistil is not always complete; it may take place by the ovaries alone, while the styles and stigmata remain free, the pistil being then gamogastrous (fig. 255), and in this case, when the ovaries form apparently a single body, the organ receives the name of compound ovary; or the union may take place by the ovaries and styles, while the stigmata are disunited; or by the stigmata and the summit of the style only. Various intermediate states exist, such as partial union of the ovaries, a3 in the Rue, where they coalesce at their base; and partial union of the styles, as in Malvaceae. The union is usually most complete at the base; but in Labiatae the styles are united throughout their length, and in Apocynaceae and Asclepiadaceae the stigmata only. When the union is incomplete, the number of the parts of a compound pistil may be determined by the number of styles and stigmata; when complete, the external venation, the grooves on the surface, and the internal divisions of the ovary, indicate the number. The changes which take place in the pistil by adhesion, degeneration, and abortion, are frequently so great as to obscure its composition, and to lead to anomalies.

Ιhe pistil is more liable to changes of this kind than any other part of the flower. The carpels are usually sessile leaves, but sometimes they are petiolate, and then are elevated above the external whorls. This elevation of the pistil may in general, however, be traced to an elongation of the axis itself, in such a way that the carpels, in place of being dispersed over it, arise only from its summit, and the pistil becomes stipitate, or supported on a stalk called a gynophore or thecapkore, as in the Passion-flower and Dictamnus (fig. 177). Sometimes the axis is produced beyond the ovaries, and the styles become united to it, as in Geraniaceae (fig. 152) and Umbelliferae. In this case the prolongation is called a carpophore. A ] monstrosity often occurs in the Rose (fig. 145) by which the axis is prolonged, and bears the carpels f in the form of alternate leaves.

The ovary contains the ovules. These are attached to ' the placenta. This, sometimes called the trophosperm, J consists of a mass of cellular tissue, through which the vessels pass to the ovule. The placenta is usually formed on the edges of the carpellary leaf (fig. 256), and is then said to be of the marginal type. In many cases, however, the placentas are formations from the axis, and are not connected with the carpellary leaves; they are then said to be axi,le. Some restrict the term placenta to the point of attachment of a single ovule, and call the union of placentas, bearing several ovules, placentaries or pistillary cords. In marginal placentation the part of the carpel bearing the placenta is the inner or ventral su faire, corresponding to the margin of the folded carpellary leaf, while the outer or dorsal suture corresponds to the midrib of the carpellary leaf. As the placenta is formed on each margin of the carpel it is essentially double. This is seen in cases where the margins of the carpel do not unite, but remain separate, and consequently two placentas are formed in place of one. When the pistil is formed by one carpel the inner margins unite in the axis, and form usually a common marginal placenta. This placenta may extend along the whole margin of the ovary as far as the base of the style (fig. 256), or it may be confined to the base or apex only. When the pistil is composed of several separate carpels, or, in other words, is apocarpous, there are generally separate placentas at each of their margins. In a syncarpous pistil, on the other hand, the carpels are so united that the edges of each of the contiguous ones, by their union, form a septum, or dissepiment, and the number of these septa consequently indicates the number of carpels in the compound pistil (fig. 257) It is obvious then that each dissepiment is formed by a double wall or two laminae; that the presence of a septum implies the presence of more than one carpel; and that, when carpels are placed side by side, true dissepiments must be vertical, and not horizontal. When the dissepiments extend to the centre or axis, the ovary is divided into cavities, cells, or loculaments, and it may be bilocular, trilocular (fig. 257), quadrilocular, quinquelocular, or multiloσular, according as it is formed by two, three, four, five, or many carpels, each carpel corresponding to a single cell or loculament. In these cases the marginal placentas meet in the axis, and unite so as to form a single central one (figs. 257, 258), and the ovules appear in the central angle of the loculi, as in Canna, Lily (fig. 257). The number of loculaments is equal to that of the dissepiments. When the carpels in a syncarpous pistil do not fold iuwards completely so as to meet in the centre, but only partially, so that the dissepiments appear as projections on the walls of the ovary, then the ovary is unilocular (fig. 260), and the placentas are parietal, as in Viola (fig. 261). In these instances the placentas may be formed at the margin of the united contiguous leaves, so as to appear single, or the margins may not be united, each developing a placenta. Frequently the margins of the carpels, which fold in to the centre, split there into two lamellae, each of which is curved outwards and projects into the loculament, dilating at the end into a placenta. This is well seen in Cucurbitaceae (fig. 262), Pyrola, &c. From this it will be seen that dissepiments are opposite to placentas formed by the union of the margins of two contiguous carpels, but alternate with those formed by the margins of the same carpel. The carpellary leaves may fold inwards very slightly, or they may be applied in a valvate manner, merely touching at their margins, the placentas then being parietal (fig. 259), and appearing as lines or thickenings along the walls. Cases occur, however, in which the placentas are not connected with the walls of the ovary, and form what is called a/ree central placenta (fig. 263).’

This is seen in many of the Caryophyllaceae and Primulaceae (figs. 264, 265). In Caryophyllaceae, however, while the placenta is free in the centre, there are often traces found at the base of the ovary of the remains of septa, as if rupture had taken place, and, in rare instances, ovules are found on the margins of the carpels. But in Primulaceae, Myrsinaceae, and Santalaceae, no vestiges of septa or marginal ovules can be perceived at any period of growth; the placenta is always free, and rises in the centre of the ovary, and the part uncovered by ovules gradually extends into the style. Free central placentation, therefore, has been accounted for in two ways; either by supposing that the placentas in the early state were formed on the margins of carpellary leaves, and that in the progress of development these leaves separated from them, leaving the placentas and ovules free in the centre; or by supposing that the placentas are not marginal but axile formations, produced by an elongation of the axis, the ovules being lateral buds, and the carpels verticillate leaves, united together around the axis. The first of these views would apply well to Caryophyllaceae, the second to Primulaceae. The latter case has also been explained, on the marginal hypothesis, by considering the placentas as formed from the carpels by a process of chorisis, and united together in the centre.

Some consider the axile view of placentation as applicable to all cases, the axis in some· instances remaining free and independent, at other times sending prolongations along the margins of the car ellary leaves, and thus forming the marginal placentas. Occasionally, divisions take place in ovaries which are not formed by the edges of contiguous carpels. These are called spmrious dissepiments. They are often horizontal, and are then called phragmata, as in Cathartocarpus Fistula, where they consist of transverse cellular prolongations from the walls of the ovary, only developed after fertilization, and therefore more properly noticed under fruit. At other times they are vertical, as in Datura, where the ovary, in place of being two-celled, becomes four-celled; in Cruciferae, where the prolongation of the placentas forms a replum or partition; in Astragalus and Thespesia, where the dorsal suture is folded inwards; in Oxytropis, where the ventral suture is folded inwards; and in Diplophractum, where the inner margin of the carpels is inflexed. In Cucurbitaceae divisions are formed in the ovary apparently by peculiar projections inwards from curved parietal placentas. In some cases horizontal dissepiments are supposed to be formed by the union of carpels situated at different heights, so that the base of one becomes united to the apex of another. In such cases the divisions are true dissepiments formed by carpellary leaves. The anomalous divisions in the ovary of the Pomegranate have been thus explained.

The ovary is usually of a more or less spherical or curved form, sometimes smooth and uniform on its surface, at other times hairy and grooved. The grooves usually indicate tho divisions between the caιpels, and correspond to the dissepiments. The dorsal suture may be marked by a slight projection, or by a superficial groove. When the ovary is situated on the centre of the receptacle, free from the other whorls, so that its base is above the insertion of the stamens, it is termed superior, as in Lychnis, Primula, and Geranium (fig. 215). When the margin of the receptacle is prolonged upwards, carrying with it the floral envelopes and staminal leaves, the basal portion of the ovary being formed by the receptacle, and the carpellary leaves alone closing in the apex, the ovary is inferior, as in Pomegranate, Apple, Pear, Gooseberry, and Fuchsia. In some plants, as many Saxifragaceae, there are intermediate forms, in which the term half-inferior is applied to the ovary, whilst the floral whorls are half-superior.

The style proceeds from the summit of the carpel, and may be looked upon as a prolongation of it in an upward direction (fig. 2G7). It is hence called apicilar. It consists not merely of the midrib but of the vascular and cellular tissue of the carpel, and when carefully examined is found to be traversed by a narrow canal, in which there are some loose projecting cells, a continuation of the placenta, constituting what is called conducting tissue, which ends in the stigma. This is particularly abundant when the pistil is ready for fertilization. In some cases, owing to more rapid growth of the dorsal side of the ovary, the style becomes lateral (fig. 266); this may so increase that the style appears to arise from near the base, as in the Strawberry, or from the base, as in Chrysobalanus Icaco, when it is called basilaι. In all these cases the style still indicates the organic apex of the ovary, although it may not be the apparent apex. When in a compound pistil the style of each carpel is thus displaced, it appears as if the ovary were depressed in the centre, and the style rising from the depression in the midst of the carpels seems to come from the torus. Such a style is gynobasic, and is well seen in Boraginaceae and Ochnaceae. The form of the style is usually cylindrical, more or less filiform and simple; sometimes it is grooved on one side, at other times it is flat, thick, angular, compressed, and even petaloid, as in Iris and Canna. In Goodeniaceae it ends in a cup like expansion, enclosing the stigma. It may be smooth and covered with glands and hairs. These hairs occasionally aid in the application of the pollen to the stigma; and are called collecting hairs, as in Goldfussia, in Campanula, where they appear double and retractile, and also in Aster and other Compositae. These hairs, during the upward development of the style, come into contact with the already ripened pollen, and carry it up along with them, ready to be applied by insects to the mature stigma of other flowers. In Vicia and Lobelia the hairs frequently form a tuft below the stigma. The styles of a syncarpous pistil may be either separate or united; when separate, they alternate with the septa. When united completely, it is usual to call the style simple (fig. 267); when the union is partial, then the style is said to be 6z∕Z√, trifid, m/ultifid, according as it is two-cleft, three-cleft, many-cleft; or, to speak more correctly, according to the mode and extent of the union of two, three, or many styles. The style is said to be bipartite, tripartite, or multipartite, when the union of two, three, or many styles only extends a short way above the apex of the ovary. The style of a single carpel, or of each carpel of a compound pistil, may also be divided. Each division of the tricarpellary ovary of Jatropha Curcas has a bifurcate or forked style, and the ovary of Emblica officinalis has three styles, each of which is divided twice in a bifurcate manner, exhibiting thus a dichotomous division. The length of the style is determined by the relation which ought to subsist between the position of the stigma and that of the anthers, so as to allow the proper application of the pollen. In some cases the ovary passes insensibly into the style, as in Digitalis, in other instances there is a marked transition from one to the other. The style may remain persistent, or it may fall off after fertilization is accomplished, and thus be deciduous.

The stigma is the termination of the conducting tissue , of the style, and is usually in direct communication with the placenta. It may, therefore, in most instances, be considered as the placental portion of the carpel, prolonged upwards. In Armeria, and some other plants, this connection with the placenta cannot be traced. The stigma consists of loose cellular tissue, and secretes a viscid matter which detains the pollen and causes it to protrude tubes. This secreting portion is, strictly speaking, the true stigma, but the name is generally applied to all the divisions of the style on which the stigmatic apparatus is situated. The stigma alternates with the dissepiments of a syncarpous pistil, or, in other words, corresponds with the back of the loculaments; but in some cases it would appear that half the stigma of one carpel unites with half that of the contiguous carpel, and thus the stigma is opposite the dissepiments, that is, alternates with the loculaments, as in the Poppy. If the stigma is viewed as essentially a prolongation of the placenta, then there is no necessary alternation between it and the placenta, both being formed by the margins of carpellary leaves, which in the one case are ovuliferous, in the other stigmatiferous. There is often a notch on one side of a stigma (as in some Rosaceae), indicating apparently that it is a double organ like the placenta. To the division of a compound stigma the terms bifid, as in Labiatae, Compositae, trifid, as in Polemonium, &c., are applied, according to the number of the divisions. When the divisions are large, they are called lobes, and when flattened like bands, lamelloe ; so that stigmas may be bilobate, trilobate, bilamellar, tτilamellar, àc.

It has already been stated that the divisions of the stigma mark the number of carpels which are united together. A quinquefid or five-cleft stigma indicates five carpels, as in Campanula, although the other parts are united. In Bignoniaceae, as well as i ι Scrophulariaceae and Acanthaccae, the two-lobed or bilamellar stigma indicates a bilocular ovary. Sometimes, however, as in the case of the styles, the stigma of a single carpel may divide. It is probable that in many instances what is called bifurcation of the style is only the division of the stigma. In Gramineae and Compositae there is a bifid stigma, and only one cavity in the ovary. This, however, may be probably traced to subsequent abortion in the ovary of one of the carpels. Its position may be terminal or lateral; either on one side of the style, uni- lateral, as in Asimina, or on both sides, bilateral, as in Plantago. Occasionally, as in Tasmannia, it is prolonged along the inner surface of the style. In Iris it is situated on a cleft on the back of the petaloid divisions of the style (fig. 268). Some stigmata, as those of the Mimulus, present sensitive flattened laminae, which close when touched. The stigma presents various forms. It may be globular, as in Mirabilis Jalapa; orbicular, as in Arbutus Andrachne; umbrella-like, as in Sarracenia, where, however, the proper stigmatic surface is beneath the angles of the large expansion of the apex of the style; ovoid, as in Fuchsia; hemispherical; polyhedral; radiating, as in the Poppy (fig. 269), where the true stigmatic rays are attached to a sort of peltate or shield-like body, which may represent depressed or flattened styles; cucullate, i.e., covered by a hood, in Calabar Bean, where it is situated on the apex of a declinate style, bearded (hairy) on its concave surface The lobes of a stigma may be flat and pointed, as in Mimulus and Bignonia, fleshy and blunt, smooth or granular, or they may be feathery, as in many Grasses (fig. 213). In Orchidaceae the stigma is situated on the anterior surface of the column formed by the union of the styles and filaments,—the point where it occurs being called gynizus. In Asclepiadacese the stigmas are united to the face of the anthers, and along with them form a solid mass (fig. 229).

Transformations of the pistil are of frequent occurrence, and depend generally on abortion of a certain number of carpels, and on adhesions of various kinds. In the apocarpous pistils of Aconite, Nigella, Larkspur, and Paeony, we find on the same plant pistils composed of two, three, four, five, and six carpels. In some of the Brambles, all the carpels except one have been observed to disappear, thus making the fruit resemble that of the Plum. In the case of Leguminous plants there is usually only a single carpel, although the flower is pentamerous; this state has been traced to abortion of carpels, and the view is confirmed by finding plants in the same natural order with more than one carpel. Pistils of a succulent nature, such as those of the Sloe and Bird-cherry, sometimes assume the form of a pod, like that of the Pea. Occasionally stamens are changed into carpels, and at other times the carpels are transformed into stamens, and bear pollen.

The ovule is the body attached to the placenta, and destined to become the seed. Ovules are most usually produced on the margins of the carpellary leaves, but are also formed over the whole surface of the leaf, as in Cupressus. In other instances they rise from the floral axis itself, either as terminal buds, as in Polygonaceae and Piperaceae, or as lateral buds, as in Primulaceae and Compositae. The ovule is usually contained in an ovary, and all plants in which the ovule is so enclosed are termed angiospermous; but in Conifer» and Cycadaceae it is generally considered as having no proper ovarian covering, and is called naked, these orders being denominated gymnospermous The gymnospermal view is not adopted by all botanists, some maintaining that there is a true ovarian covering In Cycas the altered leaf, upon the margin of which the ovule is produced, and the peltate scales, from which they are pendulous in Zamia, are regarded by all botanists as carpellary leaves. But in the Conifer» great discussion has arisen regarding the morphology of parts in many genera; some considering the scales at the base of the scaly bracts of the cone as a placental process growing from the bract, which is thus a carpellary leaf opened out and bearing a sessile ovule, the whole cone representing a single flower; while others, again, regard the scale as an ovular integument, and the ovule as being destitute of ovary, the outer scales being bracts, and the cone therefore being an inflorescence. The carpellary leaves are sometimes united in such a way as to leave an opening at the apex of the pistil, so that the ovules are exposed or seminude, as in Mignonette. In Leontice thalictroides (Blue Cohosh), species of Ophiopogon, Peliosanthes, and Stateria, the ovary ruptures immediately after flowering, and the ovules are exposed; and in species of Cuphea the placenta ultimately bursts through the ovary and corolla, and becomes erect, bearing the exposed ovules. The ovule is attached to the placenta either directly, when it is called sessile, or by means of a prolongation called a funiculus, umbilical cord, or podosperm (fig. 2 7 Oct, ∕). This cord sometimes becomes much elongated after fertilization. The part by which the ovule is attached to the placenta or cord is its base or hilum, the opposite extremity being its apex. The latter is frequently turned round in such a way as to approach the base. The ovule is sometimes embedded in the placenta, as in Hydnora.

The ovule appears at first as a small cellular projection from the placenta. The cells multiply until they assume a more or less enlarged ovate form, constituting what has been called the nucleus (fig. 270, n), or central cellular mass of the ovule. The nucleus may remain naked, and alone form the ovule, as in Balanophoraceae, Santalaceae, &c.; but in most plants it becomes surrounded by certain coverings or integuments during its development. These appear first in the form of cellular rings at the base of the nucleus, which gradually spread over its surface. In some cases only one covering is formed, especially amongst gamopetalous Dicotyledons, as in Compositae, Campanulaceae, also.in Walnut, &c. But usually besides the single covering (fig. 270α, s) another is developed subsequently (fig. 270α, ρ), which gradually extends over that first formed, and ultimately covers it completely, except at the apex. There are thus two integuments to the nucleus, an outer and an inner, called respectively primine, p, and secundine, s,— the terms having reference to their position as regards the nucleus and not indicating the order of development. The name tercine has been given to the outer layer of cells of the nucleus. The integuments do not completely invest the apex of the nucleus, but an opening termed the foramen or micropyle is left. This foramen in most ovules extends through both coats, the opening in the primine (fig. 270«, ex) being the exostome; that in the secundine (fig. 270«, end) being the endostome. But in many Monocotyledons the exostome has no share in the formation of the micropyle. The micropyle indicates the organic apex of the ovule. The term micropyle is sometimes restricted to the foramen in the perfect seed. The size of the micropyle depends on the development of the nucleus, as well as on the thickness of the integuments. Where the integument is very thick and the nucleus small, the micropyle is a long canal, as in Hippuris; but more usually the nucleus is large, and the integuments reduced to a few layers of cells, and in this case the micropyle is correspondingly reduced. The nucleus alters in the progress of growth so as to be prepared for the development of the embryo in its interior. A single cell of the nucleus near its centre enlarges greatly until it forms a hollow cavity surrounded by the smaller cells of the nucleus. This cavity is the embryo-sac (fig. 274, s), and the protoplasmic contents have been termed the amnios. This embryo-sac increases in size, gradually supplanting the surrounding cellular tissue of the nucleus until it remains surrounded only by a thin layer of it; or it may actually extend at the apex beyond it, as in Phaseolus and Alsine media ; or it may pass into the micropyle, as in Santalum. In Gymnosperms it usually remains deep in the nucleus and surrounded by a thick mass of cellular tissue (fig. 272). In Veronica, Euphrasia, and many Labiatae, the neck of the embryo-sac becomes elongated and swollen, and from it are developed certain vermiform or filamentous appendages, which are probably connected with the nutrition of the embryo. In some cases more than one embryo-sac is formed. This occurs in some gymnospermous plants, as the Yew, and it is also seen in Cruciferae. In the Mistleto several sacs are formed, but it is doubtful whether in this case several ovules have not coalesced. Usually only one becomes fully developed. When the embryo-sac has reached a certain stage of growth, a development of cellular tissue takes place within it by free cell formation. This occurs in all gymnospermous plants, and constitutes what is known as the endosperm (fig. 280, δ). In angiospermous plants the endosperm is not formed until after fertilization, and it is rare to find any special development of cellular tissue within the embryo-sac before fertilization. But in some cases at the base of the embryo-sac a few cells are formed, which have been termed antipodal cells, and are frequently afterwards absorbed, or may be incorporated in the true endosperm when it is formed. The further processes of growth in gymnospermous ovules on the one hand, and angiospermous ovules on the other, differ so much that they must be noted separately.

At the apex of the embryo-sac in gymnosperms, two or more cells of the endosperm enlarge so as to form what have been termed corpuscles (fig. 273). Each of these divides so as to form a large lower terminal or central cell, and an upper neck of smaller cells (stigmatιc cells'), appearing as four when viewed from above, enclosing a canal. From the upper part of the central cell at the bottom of the neck, a small portion is separated—the canal cell. It is from a portion of the central cell that, after fertilization, the embryo is formed. Those parts all have great physiological importance in connection with fertilization. In angiosperms from the protoplasmic mass at the apex of the embryo-sac by free cell formation two (rarely one) elongated ovoid cells are produced. These are the germinal vesicles (fig. 274). In some angiosperms as Crocus, Maize, &c., the vesicles placed side by side are both equally elongated; to a considerable extent their lower ends are rounded, and there it is that the nucleus is found. The upper end projects into the micropyle, and is marked by distinct striae longitudinally. This portion has been distinguished as the filiform apparatus. From the germinal vesicles after impregnation the embryo is formed, only one vesicle developing. This filiform apparatus is considered as corresponding to the canal cell formed from the corpuscula in gymnosperms, and it seems to serve for conducting the impregnating influence to the lower part of the central cell. In most angiosperms the germinal vesicles are placed obliquely over each other—one attached to the apex of the sac, the other lower down, and there is no filiform apparatus. Of these two vesicles the lower one alone forms an embryo, the upper one performing the same function as the filiform apparatus of other angiosperms, and the canal cell of gymnosperms. The point where the integuments are united to the base of the nucleus is called the chalaza (fig. 278, ch). This is often coloured, is of a denser texture than the surrounding tissue, and is traversed by fibro-vascular bundles, which pass from the placenta to nourish the ovule.

When the ovule is so developed that the point of union between the integuments and nucleus (the chalaza) is at the hilum (next the placenta), and the micropyle is at the opposite extremity, there being a short funiculus, the ovule is orthotropal, orthotropous, or atropous. This form is well seen in Polygonaceae (fig. 271), Cistaceae, and most gymnosperms. In such an ovule a straight line drawn from the hilum to the micropyle passes along the axis of the ovule. Where, by more rapid growth on one side than on the other, the nucleus, together with the integuments, is curved upon itself, so that the micropyle approaches the hilum, and ultimately is placed close to it, while the chalaza is at the hilum, the ovule is campylotropal or campylotropous (fig. 275) when the portions on the two sides of the line bisecting the angle of curvature are unequal, or camptotrσpal when they are equal. Curved ovules are found in Leguminosae, Cruciferae, and Caryophyllaceae. Usually the opposite faces of the concavity of curvature of such ovules coalesce, and outwardly no indication of curvature is visible; but in some cases there is no coalition, and the ovule is lecotropal, or horse-shoe shaped. The inverted, anatropal, or anatropous ovule (figs. 276, 277, 278) is the commonest form amongst angiosperms. In this ovule the apex with the micropyle is turned towards the point of attachment of the funiculus to the placenta, the chalaza being situated at the opposite extremity; and the funiculus, which runs along the side usually next the placenta, coalesces with the ovule and constitutes the raplie (r), which often forms a ridge. The anatropal ovule arises from the placenta as a straight or only slightly curved cellular process, and as it grows, gradually becomes inverted, curving from the point of origin of the integuments. As the first integument (secundine) grows round it, the amount of inversion increases, and the funiculus becomes adherent to the side of the nucleus. Then if a primine be formed it covers all the free part of the ovule, but does not form on the side to which, the raphe is adherent. Some anatropal ovules, however, seem to be formed in a different manner, the nucleus arising as a lateral bud from the apex of the funiculus, as in some Compositae. These may be taken as the three types of ovule in the vegetable kingdom; but there are various intermediate forms, such as semi-anatropal, amphitropal, and heterotropal (transverse) ovules, where the funiculus is only, as it were, partially attached along one side, becoming free in the middle.

The position of the ovule relative to the ovary varies. When there is a single ovule, with its axis vertical, it may be attached to the placenta at the base of the ovary (fiasal placenta), and it is then erect, as in Polygonaceae and Composite; or it may be inserted a little above the base, on a parietal placenta, with its apex upwards, and then is ascending, as in Parietaria. It may hang from an apicilar placenta at the summit of the ovary, its apex being directed downwards, and is inverted or pendulous, as in Hippuris vulgaris; or from a parietal placenta near the summit, and then is suspended, as in Daphne Mezereum, Polygalaceae, and Euphorbiaceae. Sometimes a long funiculus arises from a basal placenta, reaches the summit of the ovary, and there bending over suspends the ovule, as in Λrmeria (Sea-pink); at other times the hilum appears to be in the middle, and the ovule becomes horizontal, peltate, or peritropous. When there are two ovules in the same cell, they may be either collateral, that is, placed side by side (fig. 257), or the one may be erect and the other inverted, as in some species of Spiraea and Aesculus; or they may be placed one above another, each directed similarly, as is the case in ovaries containing a moderate or definite number of ovules. Thus, in the ovary of Leguminous plants (fig. 256), the ovules o, are attached to the extended marginal placenta, one above the other, forming usually two parallel rows corresponding to each margin of the carpeL When the ovules are definite (i,e., are uniform, and can be counted), it is usual to find their attachment so constant as to afford good characters for classification. When the ovules are very numerous (indefinite), while at the same time the placenta is not much developed, their position exhibits great variation, some being directed upwards, others downwards, others transversely; and their form is altered by pressure into various polyhedral shapes. In such cases it frequently happens that some of the ovules are arrested in their development and become abortive.

The homology of the ovule is by no means the same in all plants. In such cases as Polygonum and Piperaceae, it represents the termination of the floral axis, and therefore is of the nature of a caulome. Again, in such plants as Primulaceae and Composite, it is produced laterally upon the axis, and therefore represents a leaf, the integuments representing the lamina, and the funiculus the petiole,—the nucleus being an outgrowth from them. In some instances of malformation a transformation into these parts actually takes place. In cases where they are produced on the margin of the carpellary leaves (the usual mode), the ovules represent lobes of a leaf, and in some cases of monstrosity, as in Delphinium datum, they appear as lobes of the carpellary leaf, whilst in Cupressus they are evidently outgrowths of the leaf. Further, the ovules in Orchidaceae must be considered as mere trichomes, as they have no fibro-vascular bundles, and are developed from superficial cells of the placenta.

When the pistil has reached a certain stage in growth it becomes ready for fertilization. Pollination having been effected, and the pollen-grain having reached the stigma in angiosperms, the summit of the nucleus in gymnosperms, it is detained there, and the viscid secretion from the glands of the stigma in the former case, and the moisture from the ovule in the latter, induce the protrusion of the intine as a pollen-tube through the pores or points of perforation of the grain, many or few tubes being formed according to the number of pores. The pollen-tube or tubes pass down the canal (fig. 279), through t⅛e conducting tissue of the style when present, and reach the interior of the ovary in angiosperms, and then pass to the micropyle of the ovule, one pollen-tube going to each ovule. Sometimes the micropyle lies close to the base of the style, and then the pollen-tube enters it at once, but frequently it has to pass some distance into the ovary, being guided in its direction by various contrivances, as hairs, grooves, &c. In gymnosperms the pollen-grain resting on the apex of the nucleus sends out its pollen-tubes, which at once penetrate the nucleus of the ovule (fig. 280). In angiosperms when the pollen-tube reaches the micropyle it passes down into the canal, and this portion of it increases considerably in size. Where, as in Santalum album, Crocus, <tc., there is a filiform apparatus, the pollen-tube comes in contact with it. In most cases, however, it reaches the apex of the embryo-sac, sometimes slightly indenting it, as in Narcissus poeticus and Digitalis purpurea, or even perforating it, as in Canna. The granular protoplasmic matter in the pollen (fovilla) is then transmitted to the embryonal vesicle and fertilization is effected. Consequent upon this, after a longer or shorter period, those changes commence in the embryonal vesicle which result in the formation of the embryo plant, the ovule also undergoing changes which convert it into the seed, and fit it for a protective covering, and a store of nutriment for the embryo. Nor are the effects of fertilization confined to the ovule; they extend to other parts of the plant. The ovary enlarges, and, with the seeds enclosed, constitutes the fruit, frequently incorporated with which are other parts of the flower, as receptacle, calyx, &c. In gymnosperms the pollen-tubes, having penetrated a certain distance down the tissue of the nucleus, are usually arrested in growth for a longer or shorter period, sometimes nearly a year. Subsequently growth recommences; the tube advances to the apex of the embryo-sac, which it pierces, and reaches the mouth of the canal of the corpuscle. It either descends the canal or remains attached at the mouth. The fovilla is transmitted to the central cell, and fertilization is complete, the central cell giving rise to the embryo. In angiosperms usually only one embryonal vesicle is fertilized, and one embryo is produced— monoembryony; but in some plants where many embryonal vesicles are formed, as in Citrus and Scabiosa, several of them may be fertilized, and thus many embryos produced— polyembryony. Usually, however, only one develops, so as to be capable of germination or growth. In gymnosperms it is very common to have polyembryony, although produced in a dιf-erent way, for each central cell of a corpuscle may produce four embryos, and as the central cell of more than one corpuscle may be fertilized a great many embryos may be formed. They do not, however, all come to maturity.

4. Female Organs of Phanerogams after Fertilization, a.—The Fruit.

After fertilization various changes take place in the parts of the flower. Those more immediately concerned in the process, the anther and stigma, rapidly wither and decay, while the filaments and style often remain for some time; the floral envelopes become dry, the petals fall, and the sepals are either deciduous, or remain persistent in an altered form; the ovary becomes enlarged, forming the pericarp; and the ovules are developed as the seeds, containing the embryo-plant. The term fruit is strictly applied to the mature pistil or ovary, with the seeds in its interior. But it often includes other parts of the flower, such as the bracts and floral envelopes. Thus the fruit of the Hazel and Oak consists of the ovary and bracts and calyx combined; that of the Apple, Pear, and Gooseberry, of the ovary and calyx; and that of the Pine-apple, of the ovaries and floral envelopes of several flowers combined. Such fruits are by some distinguished as pseudocarps. In popular language, the fruit includes all those parts which exhibit a striking change as the result of fertilization. In general, the fruit is not ripened unless fertilization has been effected: but cases occur in which the fruit swells, and becomes to all appearance perfect, while no seeds are produced. Thus, there are seedless Oranges, Grapes, and Pine-Apples. When the ovules are unfertilized, it is common to find that the ovary withers and does not come to maturity; but in the case of Bananas, Plantains, and Breadfruit, the non-development of seeds seems to lead to a larger growth, and a greater succulence of fruit.

The fruit, like the ovary, may be formed of a single carpel, or of several. It may have one cell or cavity, being unilocular; or many, multilocular, &c. The number and nature of the divisions depend on the number of carpels, and the extent to which their edges are folded inwards. The appearances presented by the ovary do not always remain permanent in the fruit. Great changes are observed to take place, not merely as regards the increased size of the ovary, its softening and hardening, but also in its internal structure, owing to the suppression, additional formation, or enlargement of parts. Thus, in the Ash (fig. 281) an ovary with two cells, each containing an ovule attached to a central placenta, is changed into a unilocular fruit with one seed; one ovule, I, becoming abortive, while the other, g, gradually enlarging until the septum is pushed to one side, unites with the walls of the cell, and the placenta appears to be parietal. In the Oak and Hazel, an ovary with three cells, and two ovules in each, changes into a one-celled fruit with one seed. In the Coco-nut, a trilocular and triovular ovary is changed into a one-celled, one-seeded fruit. This abortion may depend on the pressure caused by the development of certain ovules, or it may proceed from non fertilization of all the ovules and consequent non-enlarge ment of the carpels. Again, by the growth of the placenta, or the folding inwards of parts of the carpels, divisions occur in the fruit which did not exist in the ovary. In Pretrea zanzibarica, a one-celled ovary is changed into a four-celled fruit by the extension of the placenta. In Cathartocarpus Fistula a one· celled ovary is changed into a fruit, having each of its seeds in a separate cell, in consequence of spurious dissepiments (plιragmatd) being produced in a horizontal manner, from the inner wall of the ovary. In Linum, by the folding inwards of the back of the carpels a five-celled ovary becomes a ten-celled fruit. In Astragalus, the folding inwards of the dorsal suture converts a one-celled ovary into a two-celled fruit; and in Oxytropis the folding of the ventral suture gives rise to a similar change in the fruit. The development of cellular or pulpy matter, and the enlargement of parts not forming whorls of the flower, frequently alter the appearance of the fruit, and render it difficult to discover its formation. In the Gooseberry, Grape, Guava, Tomato, and Pomegranate, the seeds nestle in pulp, formed apparently by the placentas. In the Orange, the pulpy matter surrounding the seeds is formed by succulent cells, which are produced from the inner partitioned lining of the pericarp. In the Strawberry the receptacle becomes succulent, and bears the carpels on its convex surface (fig. 154); in the Rose there is a fleshy hollowtorus or disk, which bears the carpels on its concave surface (fig. 155). In the Juniper the scaly bracts grow up round the seeds and become succulent, and in the Fig (fig. 150) the receptacle becomes succulent and encloses an inflorescence.

The pistil, in its simplest state, consists of a carpel or folded leaf, with ovules at its margin; and the same structure will be found in the fruit, where the pericarp represents the carpellary leaf, and the seeds correspond to the ovules. The pericarp consists usually of three layers, the external, or epicarp (fig. 282, ep∖ corresponding to the lower epidermis of the leaf; the middle, or mesocarp, me, representing the parenchyma of the leaf; and the internal, or endocarp, en, equivalent to the upper epidermis of the leaf, or the epithelium of the ovary. These layers are well seen in such a fruit as the Peach or Plum, where they are separable one from the other; in them the epicarp forms what is commonly called the skin; the mesocarp, much developed, forms the flesh or pulp, and hence has sometimes been called sarcocarp; while the endocarp, hardened by the production of woody cells, forms the stone or putamen, immediately covering the kernel of the seed. The pulpy matter found in the interior of fruits, such as the Gooseberry, Grape, and Cathartocarpus Fistula, is formed from the placentas, and must not be confounded with the sarcocarp. In some fruits, as in the Nut, the three layers become blended together, and are indistinguishable. In Bladder Senna (Colutea arborescens) the pericarp retains its leaf-like appearance, but in most cases it becomes altered both in consistence and in colour. Thus in the Date the epicarp is the outer brownish skin, the pulpy matter is the mesocarp or sarcocarp, and the thin papery-like lining is the endocarp covering the hard seed. In the Medlar the endocarp becomes of a stony hardness. In the Melon the epicarp and endocarp are very thin, while the mesocarp forms the bulk of the fruit, differing in texture and taste in its external and internal parts. The rind of the Orange consists of epicarp and mesocarp, while the endocarp forms partitions in the interior, filled with pulpy cells. The part of the pericarp attached to the peduncle is called its base, and the point where the style or stigma existed is the apex. This latter is not always the apparent apex, as in the case of the ovary; it may be lateral or even basilar. The style sometimes remains in a hardened form, rendering the fruit apiculate; at other times it falls off, leaving only traces of its existence. The presence of the style or stigma serves to distinguish certain singleseeded pericarps from seeds.

When the fruit is mature and the contained seeds ripe, the carpels usually give way either at the ventral or dorsal suture or at both, and so allow the seeds to escape. The fruit in this case is dehiscent. But some fruits are indehiscent, falling to the ground entire, and the seeds eventually reaching the soil by their decay. By dehiscence the pericarp becomes divided into different pieces, or valves, the fruit being univalvτdar, biυalvular, or mτdtivalvular, &c., according as there are one, two, or many valves. The splitting may extend the whole length of the fruit, or it may be only partial, the valves forming teeth at the apex of the fruit, and the dehiscence being apicilar, as in Caryophyllaceae (fig. 283). Sometimes the valves are detached only at certain points, and thus dehiscence takes place by pores at the apex, as in Poppy (fig. 269), or at the base, as in Campanula. Indehiscent fruits are either dry, as the Nut, or fleshy, as the Cherry and Apple. They may be formed of one or several carpels. In the former case they usually contain only a single seed, which may become so incorporated with the pericarp as to appear to be naked. Such fruits are called p>seudospermmιs or false-seeded, and are exemplified in the grain of Wheat. In such cases the presence of the style or stigma determines their true nature.

Dehiscent fruits, when composed of single carpels, may open by the ventral suture only, as in the Paeony, Hellebore, Aquilegia (fig. 300), and Caltha; by the dorsal suture only, as in Magnolias and some Proteaceae, or by both together, as in the Pea (fig. 256) and Bean; in these cases the dehiscence is called sutural. When composed of several united carpels, two types of dehiscence may be recognized —a longitudinal type and a transverse type. In the longitudinal type the separation may take place by the dissepiments throughout their length, so that the fruit is resolved into its original carpels, and each valve will be a carpel, as in Rhododendron, Colchicum, &c. This dehiscence, in consequence of taking place through the lamellae of the septum, is called septicidal (figs. 284, 285).

The valves may separate from their commissure, or central line of union, carrying the placentas with them, or they may leave the latter in the centre, so as to form with the axis a column of a cylindrical, conical, or prismatic shape, which is termed the columella. The union between the edges of the carpels may be persistent, and they may dehisce by the dorsal suture, or through the back of the loculaments, as in the Lily and Iris (figs. 286, 287). In these cases each valve consists of a half of each of two contiguous carpels. The placentas either remain united to the axis, or they separate from it, being attached to the septa on the valves. This dehiscence is loculicidal. When the outer walls of the carpels break off from the septa, leaving them attached to the columella, the dehiscence is said to be septifragal (fig. 288), and where, as in Linum catharticuni and Calluna, the splitting takes place first of all in a septicidal manner, the fruit is described as septicidally septifragal; while in other cases, as in Thorn Apple (Datura Stramonium), where the splitting is at first loculicidal, the dehiscence is loculicidally septifragal. In all those forms the separation of the valves takes place either from above downwards or from below upwards. But when the splitting only extends for a short distance, then dehiscence takes place by pores, which are situated either at the apex, base, or side, depending on the position of the valves. In the Poppy (fig. 269) the opening takes place by numerous pores under the peltate processes bearing the stigmas. In Campanula there are irregular openings towards the middle or base of the pericarp. In Frogsmouth or Snapdragon tho pericarp gives way at certain fixed points, forming two or three orifices, one of which corresponds to the upper carpel, and the other to the lower. These orifices have a ragged appearance at the margins, which has given rise to tho term rupturing, sometimes applied to this mode of dehiscence. In Saxifraga a splitting for a short distance of the ventral sutures of the carpels takes place, so that a single large apical pore is formed. In Caryopliyllaceae (fig. 283) numerous small valvular splittings occur, forming teeth at the apex, and a single apical orifice is formed. In the fruit of Cruciferae, as Wallflower, there is a form of longitudinal dehiscence (fig. 289) in which the valves separate from the base of the fruit, leaving a central replum, or frame, which is a phragma formed by a prolongation from the parietal placentas on opposite sides of the fruit, extending between the ventral sutures of the carpels. In Orchidaceae (fig. 290) the pericarp, when ripe, separates into three valves, in a loculicidal manner, but the midribs of the carpels, to which the placentas are attached, remain adherent to the axis both at the apex and base, and form three arches, bearing the seeds, after the valves have fallen. In fruits with a free central placenta it is sometimes difficult to tell whether the dehiscence is septicidal or loculicidal, inasmuch as there are no dissepiments, and the placentas and seeds form a column in the axis. Their number, as well as alternation or opposition, as compared with the sepals, will aid in determining whether the valves are the entire carpellary leaves, as in septicidal dehiscence, or only half carpels united, as in loculicidal dehiscence. The other type of dehiscence is transverse, the dehiscence in this case being called circumscissile. In this dehiscence the upper part of the united carpels falls off in the form of a lid or operculum, as in Anagallis and in Henbane (Hyoscyamus) (fig. 291), and hence the fruit is often denominated opercidate. In such instances we may either suppose that the fruit or seed-vessel is formed by a number of articulated leaves like those of the Orange, the division taking place where the laminae join the petioles, or that the receptacle is prolonged in the form of a hollow cup, and the lid represents leaves united to it by articulation.

Sometimes the axis is prolonged beyond the base of the carpels, as in the Mallow and Castor-oil plant, the carpels being united to it throughout their length by their faces, and separating from it without opening. In the Umbelliferae the two carpels separate from the lower part of the axis, and remain attached by their apices to a prolongation of it, called a carpophore or podocarp, which splits into two (fig. 292) and suspends them; hence the name cremocarp is applied to this fruit, which divides into two suspended mericarps. The general term schizocarp is applied to all dry fruits, which break, up into two or more one-seeded indehiscent mericarps, as in Hedysarum (fig. 293). In Geraniaceae the axis is prolonged beyond the carpels, forming a carpophore,· to which the styles are attached, and the pericarps separate from below upwards, before dehiscing by their ventral suture (fig. 152). Carpels which separate one from another in this manner are called cocci. They are well seen in Euphorbiaceae, where there are usually three such carpels, and the fruit is designated tricoccous. In many of them, as Hura crepitans, the cocci separate with great force and elasticity, the cells being called dissilient. In many Leguminous plants, such as Ornithopus, Hedysarum (fig. 293),Entada, Coronilla, and the Gum-arabic plant (Acacia arabica), the fruit becomes a schizocarp by the formation of transverse partitions from the folding in of the sides of the pericarp, and distinct separations taking place at these partitions by what has been termed solubility. In Cathartocarpus Fistula the schizocarp exhibits no evident depressions externally. Some look upon these pods as formed by pinnate leaves folded, and the divisions as indicating the points where the differént pairs of pinnae are united.

Fruits may be formed by one flower, or they may be the product of several flowers combined. In the former case they are either apocarpous, of one mature carpel or of several separate free carpels; or sιjncarpous, of several carpels, more or less completely united. These different kinds of fruits may be indehiscent or dehiscent. When the fruit is composed of the ovaries of several flowers united, it is usual to find the bracts and floral envelopes also joined with them, so as to form one mass; hence such fruits are called multiple, confluent, or anthocarpous. The term simple is perhaps properly applied to fruits which are formed by the ovary of a single flower, whether they are composed of one or several carpels, and whether these carpels arc separate or combined. Simple fruits are hence sometimes denominated monogynoecial, as being formed by one gynoecium; while multiple fruits are called polygynoedal, as being formed by many gynoecia.

Simple fruits are either dry or succulent,— the pericarp, in the former instance, remaining more or less foliaceous in its structure, and sometimes being incorporated with the seed, while it is in the latter thick and fleshy or pulpy.

The achoenium is a dry unilocular monospermous indehiscent fruit, the pericarp of which is closely applied to the seed, but separable from it. It may be solitary, forming a single fruit, as in the Dock (fig. 294), and in the Cashew, where it is supported on a fleshy peduncle; or aggregate, as in Ranunculus (figs. 254, 295), where several achaenia are placed on a common elevated receptacle. In the Strawberry the achaenia (fig. 154) are aggregated on a convex succulent receptacle. In the Rose they are supported on a concave receptacle (fig. 155), and in the Fig the succulent receptacle completely encloses the achenes (fig. 150). In Dorstenia (fig. 163) the achenes are situated on a flat or slightly concave receptacle. In the Rose the aggregate achaenia with their covering are sometimes collectively called cynarrhodum. It will thus be remarked that what in common language are called the seeds of the Strawberry, Rose, and Fig, are in reality carpels, which are distinguished from seeds by the presence of styles and stigmas. The styles occasionally remain attached to the achaenia, in the form of feathery appendages, as in Clematis, and they arc called caudate. In Composite, the fruit, which is sometimes called cypsela, is an achaenium (fig. 197), to which the pappose or obsolete calyx remains adherent. Such is also the nature of the fruit in Dipsacaceae. When the pericarp is thin, and appears like a bladder surrounding the seed, the achaenium is termed a utricle, as in Amarantaceae. When the pericarp is extended in the form of a winged appendage, a samara or samaroid achoenium is produced, as in the Ash (fig. 281), Common Sycamore (fig. 296), and Hiraea. In these cases there are usually tw o achaenia united, one of which, however, as in Fraxinus oxyphylla (fig. 281), may be abortive. The wing is either dorsal, that is, it is a prolongation from the median vein (fig. 296, α) or it is margined, that is, formed by the lateral veins. It surrounds the fruit longitudinally in the Elm. When the pericarp becomes so incorporated with the seed as to bo inseparable from it, as in grains of Wheat, Maize, Oats (fig. 297), and other Grasses, then the name caryopsis is given. There are some fruits which consist of two or more achaenia, although originally the carpels were united into a syncarpous pistil, as in Labiatae and Boraginaceae (fig. 298). To this form of schizocarpic fruit, as well as to that of Tropaeolum and Mallow, the name career ulus is given. The cremocarp, or the fruit of (Jmbelliferae (figs. 292 and 299), is composed of two achaenia united by a commissure to a carpophore, from which they are suspended at maturity. It is sometimes denominated diachoenium, from the union of two achaenia, which in this instance receive the name of mericarps or hemicarps.

The nut or glans is a dry one-celled indehiscent fruit with a hardened pericarp, surrounded by bracts at the base, and, when mature, containing only one seed. In the young state the ovary contains two or more ovules, but only one comes to maturity. It is illustrated by the fruits of the Hazel and Chestnut, which are covered by leafy bracts, in the form of a husk, and by the Acorn, in which the bracts and receptacle form a cupula or cup (fig. 147). The parts of the pericarp of the nut are united so as to appear one. In common language the term nut is very vaguely applied both to fruit and seeds.

The drupe is a succulent usually monospermal and unilocular indehiscent fruit, with a pericarp easily distinguishable into epicarp, mesocarp, and endocarp. This term is applied to such fruits as the Cherry (fig. 282), Peach, Plum, Apricot, Mango, Walnut, Nutmeg, and Date. Tbe endocarp is usually hard, forming the stone {putamen) of the fruit, which encloses the kernel or seed. The mesocarp is generally pulpy and succulent, so as to be truly a sarcocarp, as in the Peach, but it is sometimes of a tough texture, as in the Almond, and at other times is more or less fibrous, as in the Coco-nut. In the Almond there, are often two ovules formed, only one of which comes to perfection. In the Walnut, prolongations from the endocarp, which is of two layers, extend into the substance of the seed, and give rise to its characteristic convoluted lobate appearance. This fruit has been sometimes called tryma. In the Raspberry and Bramble several drupes or drupels are aggregated so as to constitute an et<xrio.

The follicle is a dry monocarpellary unilocular polyspermal (many-seeded) fruit, dehiscing by the ventral suture. It is rare to meet with a solitary follicle forming the fruit. There are usually several aggregated together, either in a circular manner on a shortened receptacle, as in Hellebore, Aconite, Delphinium, Aquilegia (figs. 300, 301) Crassulaceae (fig. 172), Butomus, and Asclepiadaceae; or in a spiral manner on an elongated receptacle, as in Magnolia, Banksia, and Liriodendron. Occasionally, follicles dehisce by the dorsal suture, as in Magnolia grandiflora and Banksia.

The legume or pod is a dry monocarpellary unilocular polyspermal fruit, dehiscing both by the ventral and the dorsal suture. It characterises Leguminous plants, as the Bean and Pea (fig. 256). In the Bladder-senna it retains its leaf-like appearance, and forms an inflated legume. In some Leguminosae, as Arachis, Cathartocarpus Fistula, and the Tamarind, the fruit must be considered a legume, although it does not dehisce. The first of these plants produces its fruit underground, and is called earth-nut; the second has a partitioned legume and is schizocarpic; and both the second and third have pulpy matter surrounding the seeds. Some legumes are schizocarpic by the formation of constrictions externally. Such a form is the lomentum or lomentaceous legume of Hedysarum coronarium (fig. 293), Coronilla, Ornithopus, Entada, and of some Acacias. In Medicago the legume is twisted like a snail, and in Caesalpinia coriaria, or Divi-divi, it is vermiform or curved like a worm; in Carmichaelia the valves give way close to the suture, and separate from it, leaving a division. Sometimes the number of seeds is reduced, as in Erythrina monosperma and Geoftroya superba, which are one-seeded, and in Pterocarpus and Dalbergia, which are two-seeded.

The berry (bacca) is a succulent syncarpous polyspermal unilocular indehiscent fruit, with seeds immersed in a pulpy mass, formed by the placentas. The name is usually given to such fruits as the Gooseberry (fig. 302) and Currant, in which the ovary is inferior, and the placentas are parietal, the seeds being ultimately detached from the placenta, and lying loose in the pulp. Others have applied it also to those in which the ovary is superior, as in the Grape, Potato, and Ardisia, and the placentas are central or free central. The latter are frequently separated under the name uva (grape). In general, the name of baccate or berried is applied to all pulpy fruits. In the Pomegranate there is a peculiar baccate many-celled inferior fruit, having a tough rind, enclosing two rows of carpels placed one above the other. The seeds are immersed in pulp, and are attached irregularly to the parietes, base, and centre of the loculi. The fruit has been called balausta, and the tough rind is called malicorium. In the Baobab there is a multilocular syncarpous fruit, in which the seeds are immersed in pulp, to which the name amphisarca is given.

The pepo or peponida, another indehiscent syncarpous fruit, is illustrated by the fruit of the Gourd, Melon (fig. 262), and other Cucurbitaceae. It is formed of three carpels, to which the calyx is superior; the rind is thick and fleshy, partly formed by the calyx; and there are three or more seed-bearing parietal placentas, either surrounding a central cavity, or prolonged inwards into it. The fruit of the Papaw resembles the pepo, but the calyx is not superior.

The hesperidium is the name given to such indehiscent syncarpous fruits as the Orange, Lemon, and Shaddock, in which the epicarp and mesocarp form a separable rind, and the endocarp sends prolongations inwards, forming triangular divisions, to the inner angle of which the seeds are attached, pulpy cells being developed around them. Both pepo and hesperidium may be considered as modifications of the berry.

The pome, seen in the Apple, Pear, Quince, Medlar, and Hawthorn, is a fleshy indehiscent syncarpous fruit with the calyx attached, and has an outer skin or epicarp, a fleshy mesocarp, and a scaly or horny endocarp (the core) enclosing the seeds (fig. 303). Some look upon the so-called epicarp and mesocarp as formed by the prolonged receptacle on the inner surface of which a fleshy lining is developed, while the endocarp represents the true carpels. In this view the endocarp might be regarded as consisting of a number of indehiscent follicles (usually five) surrounded by a succulent receptacle. In the Medlar the endocarp (or what may be called the true pericarp) is of a stony hardness, while the outer succulent covering is open at the summit. The stones of the Medlar are called ρyrenae; some apply the term mιcιdanium to the Medlar. Taking this view of the pome it may be said to resemble in a manner the fruit of the Rose, the cynarrhodum producing achenes, and the pome closed follicles.

The name capsule is applied generally to a 1 I dry syncarpous fruits, which dehisce by valves of whatever kind. It may thus be unilocular or multilocular, monospermous or polyspermous. The true valvular capsule is observed in Colchicum (fig. 284), Lily, and Iris (fig. 286). Υheporose capsule is seen in the Poppy (fig. 269), Antirrhinum majus, and Campanula persicifolia. In Campanula the pores occur at the base of the capsule, and it has been designated a diplotegia. When the capsule opens by a lid, or by circumscissile dehiscence, it is called a pyxidium, as in Anagallis arvensis, Henbane (fig. 291), and Monkey pot (Lecythis). The capsule assumes a screw-like form in Helicteres, and a star-like or stellate form in Illicium anisatum. In certain instances the cells of the capsule separate from each other, and open with elasticity to scatter the seeds. This kind of capsule is met with in the Sandbox tree (Hura crepitans), and other Euphorbiacese, where the cocci, containing each a single seed, burst asunder with force; and in Geraniaceae, where the cocci, each containing, when mature, usually one seed, separate from the carpophore, become curved upwards by their adherent styles, and open by the ventral suture (fig. 152). In the former case the fruit collectively has been called regma.

The siliqua is a dry syncarpous bicarpellary bilocular polyspermal fruit with a replum, dehiscing by two valves from below upwards, the valves separating from the placentas and leaving them united by the replum. The seeds are attached on both sides of the replum, either in one row or in two. When the fruit is long and narrow it is called siliqua (fig. 304); when broad and short, it is called silicula (figs. 305, 306). It occurs in Cruciferous plants, as Wallflower, Cabbage, and Cress. In Glaucium and Eschscholtzia (Papaveraceous plants) the fruit is siliquaeform, the dissepiment or replum being of a spongy nature, and it has been termed a ceratium. In its normal state a siliqua is supposed to consist of four carpels, but two of these are abortive. There are four bundles of vessels in it, one corresponding to each valve which may be called valvular or pericarpial, and others running along the edge called placental. The replum consists of two lamellae. It sometimes exhibits perforations, becoming fenestrate. Rarely its central portion is absorbed, so that the fruit becomes one-celled. It may become lomentaceous, as in Raphanus and Sea-kale, and it may be reduced, as in Woad (Isatis), to a monospermal condition.

It sometimes happens that the ovaries of two flowers unite so as to form a double fruit. This may be seen in many species of Honeysuckle. But the fruits which are now to be considered consist usually of the floral envelopes, as well as the ovaries of several flowers united into one, and are called multiple, confluent, or polygynoecial. The term anthocarpous has also been applied as indicating that the floral envelopes as well as the carpels are concerned in the formation of the fruit.

The sorosis is a succulent multiple fruit formed by the confluence of a spike of flowers, as in the fruit of the Pine-apple (fig. 307), the Bread-fruit, and Jack-fruit. Sometimes a fruit of this kind resembles that formed by a single flower, and a superficial observer might have some difficulty in marking the difference.

The sycon us is an anthocarpous fruit, in which the receptacle completely encloses numerous flowers and becomes succulent. The Fig (fig. 150) is of this nature, and what are called its seeds are the achaenia of the numerous flowers scattered over the succulent hollowed receptacle. In Dorstenia (fig. 163) the axis is less deeply hollowed, and of a harder texture, the fruit exhibiting often very anomalous forms.

The strobilus, or cone, is a fruit-bearing spike, more or less elongated, covered with scales (fig. 134), each of which represents a separate flower, and has often two seeds at its base, the scales being considered as bracts and the seeds naked, and no true ovary present with its style or stigma. This fruit is seen in the cones of Firs, Spruces, Larches, and Cedars, which have received the name of Coniferae, or cone-bearers, on this account. Cone-like fruit is also seen in some Cycadaceae. The scales of the strobilus are sometimes thick and closely united, so as to form a more or less angular and rounded mass, as in the Cypress; while in the Juniper they become fleshy, and are so incorporated as to form a globular fruit like a berry. The dry fruit of the Cypress, and the succulent fruit of the Juniper, have received the name of galbulus. The fruit of the Yew (Taxus baccata) is regarded as a cone reduced to a single naked seed, covered by succulent scales, which unite to form a scarlet fleshy envelope. In the Hop the fruit is called also a strobilus, but in it the scales are thin and membranous, and the seeds are not naked but are contained in pericarps.

The same causes which produce alterations in the other parts of the flower give rise to anomalous appearances in the fruit. The carpels, in place of bearing seeds, are sometimes changed into leaves, with lobes at their margins. Leaves are sometimes produced from the upper part of the fruit, which is then called frondiparous. In the genus Citrus, to whiclι the Orange and Lemon belong, it is very common to meet with a separation of the carpels, so as to produce what are called horned oranges and fingered citrons. In this case a syncarpous fruit has a tendency to become apocarpous. In the Orange we occasionally find a supernumerary row of carpels produced, giving rise to the appearance of small and imperfect oranges enclosed within the original one. The Navel Orange of Pernambuco is of this nature. It sometimes happens that, by the union of flowers, double fruits are produced. Occasionally a double fruit is produced, not by the incorporation of two flowers, but by the abnormal development of a second carpel in the flower. [table]

b. — The Seed.

When the ovule arrives at maturity it constitutes the seed, which is contained in a seed-vessel in the plants called angiospermous ; while in gymnospermous plants, such as Coniferae and Cycadaceae, it is naked, or, in other words, has no true pericarp. It sometimes happens in angiosperms, that the seed-vessel is ruptured at an early period of growth, so that the seeds become more or less exposed during their development; this occurs in Mignonette, where the capsule opens at the apex, and in Cuphea platycentra, where the placenta bursts through the ovary and floral envelopes, and appears as an erect process bearing the young seeds. After impregnation the ovule is greatly changed, in connection with the formation of the embryo. In the embryo-sac of most angiosperms there is a development of cellular tissue, enveloping, when not previously absorbed, the antipodal cells, and more or less filling the embryo sac. In gymnosperms, as already mentioned, the endosperm is formed preparatory to fertilization. The germinal vesicle in angiosperms, the central cell of the corpuscle in cryptogams, enlarges and divides, forming the embryo. The embryo-sac enlarges greatly, displacing gradually the nucleus, which may eventually form merely a thin layer around the sac, or it may completely disappear. The integuments also become much altered, and frequently appendages are developed from them.

The general integumentary covering of the seed is called spermoderm. In it are recognized two parts, an external membrane, called the episperm or testa (fig. 308, te), and an internal membrane, called endopleura or tegmen, e, which however is often incorporated with the testa, and hardly separable from it. The testa may consist of a union of the primine and secundine, or of the primine only, when, as occasionally happens, the secundine is absorbed; the endopleura, of a combination betw een the outer layer of the nucleus (sometimes termed the tercine), and the embryo-sac, or of one of these parts alone. Sometimes the secundine remains distinct in the seed, forming what has been called a mesosperm ; and when it assumes a fleshy character, it has received the name of sarcosperm or sarcoderm. The nature of the testa depends upon that of the pericarp. When the pericarp is dehiscent then the seed-covering is of a strong and tough character; but when the pericarp is indehiscent and encloses the seed for a long period, its outer covering has a proportional softness. The cells of the testa are often coloured, and have projections and appendages of various kinds. Thus in Abrus precatorius and Adenanthera pavonina it is of a bright red colour; in French beaus it is beautifully mottled; in the Almond it is veined; in the Tulip and Primrose it is rough; in the Snapdragon it is marked with depressions; in Cotton and Asclepias (fig. 309) it has hairs attached to it; and in Mahogany, Bignonia, and Firs it is expanded in the form of wing-like appendages (fig. 310). In Collomia, Acantho-dium, and other seeds, it contains spiral cells, from which, when moistened with water, the fibres uncoil in a beautiful manner; so also in Cobaea scandens and Calempelis scaber. In the testa of the seed of Ulmus campestris the cells are compressed, and their sinuous boundaries are traced out by minute rectangular crystals adhering to their walls; and in such plants as the Flax (Linum) the cells are converted into mucilage. These structural peculiarities of the testa in different plants have relation to the scattering of the seed and its germination upon a suitable nidus. But in some plants the pericarps assume structures which subserve the same purpose; this especially occurs in small pericarps enclosing single seeds, as achenes, caryopsides, &c. Thus in Compositae (fig. 197) and Valerian, the pappose limb of the calyx forms a parachute to the pericarp; in Labiatae and some Compositae spiral cells are formed in the epicarp; and the epicarp is prolonged as a wing in Fraxinus (fig. 281) and Acer (fig. 296).

The endopleura (fig. 308, e ) is usually a thin and transparent layer of cells underneath the testa, frequently incorporated with it. It represents the outer layer of the cells of the nucleus, constituting the tercine of the ovule, though frequently the embryo-sac is incorporated with it. Sometimes there is an additional covering to the seed, formed after fertilization, to which the name arillus has been given (fig. 311). This is seen in the Passion-flower, where the covering arises from the placenta or extremity of the funiculus at the base of the ovule and passes upwards towards the apex, leaving the micropyle uncovered. In the Nutmeg and Spindle-tree this additional coat commences at the side of the exostome, and is formed from above downwards, constituting in the former case a laciniated scarlet covering called mace. In such instances it has been called by some an arillode (fig. 312). This arillode, after growing downwards, may be reflected upwards so as to cover the foramen. The fleshy scarlet covering formed around the naked seed in the Yew is by some considered of the nature of an aril. On the testa, at various points, there are produced at times cellular bodies, which are not dependent on fertilization, to which the name of stropihioles, or caruncules, has been given, the seeds being strophiolate or carunculate. These tumours may occur near the base of the seed, as in Polygala, or at the apex, as in Ricinus, where they are swellings of the exostome; or they may occur in the course of the raphe, as in Blood-root and Asarabacca. The funiculi of the ovules frequently attain a great length in the seed, and in some Magnolias, when the fruit dehisces, they appear as long scarlet cords suspending the seeds outside. The hilum or umbilicus of the seed is usually well marked, as a scar of varying size, in the Calabar bean and in some species of Macuna and Dolichos, extending along a large portion of the edge of the seed; and it frequently exhibits marked colours, being black in the Bean, white in many species’ of Phaseolus, &c. The micropyle (fig. 308, m ) of the seed, with its exostome and endostome, may be recognizable by the naked eye, as in the Pea and Bean tribe, Iris, &c., or it may be very minute or microscopic. It indicates the true apex of the seed, and is important as marking the point to which the root of the embryo is directed. At the micropyle in the Bean is observed a small process of integument, which, when the young plant sprouts, is pushed up like a lid, and is called embryotega. The chalaza (fig. 311, ch) is often of a different colour from the rest of the seed. In the Orange (fig. 313) it is of a reddish-brown colour, and is easily recognized at one end of the seed when the integuments are carefully removed. In anatropal seeds the raphe forms a distinct ridge along one side of the seed (fig. 314).

The position of the seed as regards the pericarp resembles that of the ovule in the ovary, and the same terms are applied—erect, ascending, pendulous, suspended, curved, &c. These terms have no reference to the mode in which the fruit is attached to the axis. Thus the seed may be erect while the fruit itself is pendent, in the ordinary meaning of that term. The part of the seed next the axis or the ventral suture is its face , the opposite side being the back. Seeds exhibit great varieties of form. They may be flattened laterally (compressed) or from above downwards (depressed). They may be round, oval, triangular, polygonal, rolled up like a snail, as in Physostemon, or coiled up like a snake, as in Ophiocaryon paradoxum.

The endosperm formed in the embryo-sac of angiosperms after fertilization, and found previous to it in gymnosperms, consists of cells containing protoplasmic substance and starchy or fatty matter, or both, destined for the nutriment of the embryo. It arises by free-cell division from the protoplasm of the embryo-sac,—and may occupy the whole cavity of the embryo-sac, or may be formed only at certain portions of it, at the apex, as in Rhinanthus, at the base, as in Vaccinium, or in the middle, as in Veronica. As the endosperm increases in size along with the embryo-sac and the embryo, the substance of the original nucleus of the ovule is gradually absorbed, and usually at last disappears except the layer constituting the endopleura, and in some cases this goes also. Sometimes, however, as in Musaceae, Cannaceae, Zingiberaceae, no endosperm is formed; the celle of the original nucleus, becoming filled with food-materials for the embryo, are not absorbed, but remain surrounding the embryo-sac with the embryo, and constitute the perisperm,. Again, in other plants, as Nymphaeaceae (fig. 311) and Piperaceae, both endosperm and perisperm are present. It was from observations on cases such as these that old authors, imagining a resemblance betwixt the plant-ovule and the animal ovum, applied the name albumen to the outer nutrient mass or perisperm, and designated the endosperm as vitellus. The term albumen is very generally used as including all the nutrient matter stored up as a separate mass in the ovule, but it would be advisable to discard the name as implying a definite chemical substance. There is a large class of plants in which although at first after fertilization a mass of endosperm is formed, yet, as the embryo increases in size, the nutrient matter from the endospermic cells passes out from them, and is absorbed by the cells of the embryo plant. In the mature seed, in such cases, there is no separate mass of tissue containing nutrient food-material apart from the embryo itself. Such a seed is said to be exalbuminous, as in Compositae, Cruciferae, and most Leguminosae. When either endosperm or perisperm or both are present the seed is said to be albuminous.

The albumen varies much in its nature and consistence, and furnishes important characters. It may be farinaceous or mealy, consisting chiefly of cells filled with starch, as in cereal grains, where it is abundant ; fleshy or cartilaginous, consisting of thicker cells which are still soft, as in the Coco-nut, and which sometimes contain oil, as in the oily albumen of Croton, Ricinus, and Poppy; horny, when the cell-walls are slightly thickened and capable of distension, as in Date and Coffee; and they sometimes become greatly thickened, filling up the testa as a hard mass, as in Vegetable Ivory. The albumen may be uniform throughout, or it may present a mottled appearance, as in the Nutmeg, the seeds of Anonaceae, and some Palms, where it is called ruminated. This mottled appearance is due to a protrusion of a dark lamella of the endopleura between folded protuberances of albumen. The endosperm within the embryo-sac is developed from the protoplasm of that sac, but in many cases as it grows inward the whole protoplasm is not converted into a solid mass, but a cavity is left in the centre which is usually filled with fluid, as in the Coco-nut. In the thickened albumen of this Palm, as well as in that of the Attalea funifera, the Date, and the Doum Palm, the small cavity in the centre and radiating spaces are well seen under the microscope. The albumen is a store of matter laid up for the nourishment of the embryo. The relative size of the embryo and of the endosperm varies much. In Monocotyledons the embryo is usually small, and the endosperm large, and the same is true in the case of Coffee and many other plants amongst Dicotyledons. The opposite is the case in other plants, as in the Labiatae, Plumbaginaceae, &c.

In angiosperms after fertilization the embryonal vesicle undergoes changes by which the embryo plant is eventually formed. The portion of the vesicle nearest the apex of the embryo-sac coalesces with it. The lower portion enlarges and lengthens greatly, and divides by transverse partition until a large terminal cell is formed at the extremity of a suspensor or proembryo (figs. 316, 317), formed of smaller cells. It is this terminal cell which immediately forms the embryo. By longitudinal growth the terminal cell becomes pushed into the endosperm (when present) of the sac, and becomes completely surrounded by it. In the lower part of the central cell of gymnosperms a cell is produced which divides until a proembryo is formed, which elongates and bursts through the central cell. The divisions in this proembryo take place so that several suspensors are formed, penetrating the endosperm. Usually, each of these elongates and bears a large cell at its extremity, which becomes an embryo; thus polyembryony occurs. The terminal cell of the proembryo in both angiosperms and gymnosperms increases in size, and divides repeatedly, until a cellular axis is formed. This is the rudiment of the embryo. Upon this axial mass, one, two, or more rounded cellular protuberances appear, the indications of the first leaves or cotyledons; and at the point where the embryo joins the proembryo is intercalated a mass of cellular tissue forming the hypophysis, from which subsequently the young root or radicle is formed. The axial mass continues to grow, and soon a differentiation is observable into an outer peripheral layer, the dermatogen or primordial epidermis, covering over the central mass, which soon is distinguishable into a plerome or axial mass, from which the fibro-vascular bundles develop, and a portion between it and the dermatogen,—the periblem or primordial cortex. After this has been formed a differentiation. takes place in the hypophysis, by which an upper and a lower layer of cells is formed. The upper divides into two layers, the lower of which forms a continuation of the dermatogen of the axial mass, the upper being continuous with the central portion of the axis. The lowermost cells of the hypophysis form the first layer of the root-cap. Lateral roots are frequently given off from the embryo, especially in Monocotyledons, and these arise as cellular processes from the outer or pericambial layer of the plerome in the hypocotyledonary portion of the axis. The primary root of the embryo is, by reason of its formation, always directed to the micropyle In Monocotyledons and some Dicotyledons the young root rises deep in the tissues of the embryo, which, when it sprouts, form a sheath or coleorhiza around it. The lateral roots have also sheaths.

The embryo then consists of cotyledons (figs. 52, 53, 58, c), or the first leaves of the plant produced upon a cellular axial mass. To that part of this axis immediately beneath the cotyledons the terms hypocotyledonary portion, caulicle, or tigellus (t) have been applied, and continuous backwards with it is the young root or radicle (r), the descending axis, their point of union being the collum, collar, or neck. The terminal growing bud of the axis is called the plumule or gemmule (g), and represents the ascending axis. That extremity of the embryo which produces the first leaves or cotyledons is called the cotyledonary extremity of the embryo, while the opposite is the radicular extremity. The radicular extremity being continuous with the suspensor points towards the micropyle (fig. 318), or the summit of the nucleus, an important fact in practical botany; while the cotyledonary extremity is pointed towards the base of the nucleus or the chalaza. Hence, by ascertaining the position of the micropyle and chalaza, the two extremities of the embryo can in general be discovered. In some rare instances, in consequence of a thickening in the coats of the seed, as in Ricinuβ and some other Euphorbiaceae, there is an alteration in the micropyle, so that the radicle does not point directly to it. It is in many cases difficult to recognize the parts in an embryo; thus in Cuscuta, the embryo appears as an elongated axis without divisions; and in Caryocar butyrosum, the mass of the embryo is made up by the radicular extremity and tigelle, in a groove of which the cotyledonary extremity lies embedded. In some Monocotyledonous embryos, as in Orchidaceae, the parts of the embryo of the ripe seed are not differentiated, and only become so after germination. In parasitic plants which form no chlorophyll, as Orobanche, Monotropa, &c., the embryo remains without differentiation, consisting merely of a mass of cells until the ripening of the seed. When the embryo is surrounded by the endosperm on all sides except its radicular extremity it becomes internal or intrarius; when lying outside the endosperm, and only coming into contact with it at certain points, it is external or extranus. When the embryo follows the direction of the axis of the seed, it is axile or axial (fig. 318), and it may be either external, so as to come into contact with the endosperm only by its cotyledonary apex, or internal. In the latter case the radicular extremity may, as in some Coniferae, become incorporated with the endosperm apparently by means of a thickened suspensor. When the embryo is not in the direction of the axis, it becomes dbaxüe or abaxial; and in this case it may be either straight or curved, internal or external. In campylotropal seeds the embryo is curved, and in place of being embedded in endosperm, is frequently external to it, following the concavity of the seed (fig. 319), and becoming peripherical, with the chalaza situated in the curvature of the embryo, as in Caryophyllaceae

It has been already stated that the radicle of the embryo is directed to the micropyle, and the cotyledons to the chalaza. In some cases, by the growth of the integuments, the former is turned round so as not to correspond with the apex of the nucleus, and then the embryo has the radicle directed to one side, and is called excentric, as is seen in Primulaceae, Plantaginaceae, and many Palms, especially the Date. The position of the embryo in different kinds of seeds varies. In an orthotropal seed the embryo is inverted or antitropal, the radicle pointing to the apex of the seed, or to the part opposite the hilum. Again, in an anatropal seed the embryo is erect or homo-tropal (fig. 318), the radicle being directed to the base of the seed. In some anatropal seeds, as in Castor oil, the exostome is thickened or carunculate, and the endostome does not correspond exactly to it, so that the radicle of the embryo is directed to a point a little removed from the exostome. In curved or campylotropal seeds the embryo is folded so that its radicular and cotyledonary extremities are approximated, and it becomes amphitropal. In this instance the seed may be exalbuminous, and the embryo may be folded on itself; or albuminous, the embryo surrounding more or less completely the endosperm, and being peripherical. According to the mode in which the seed is attached to the pericarp, the radicle may be directed upwards or downwards, or laterally, as regards the ovary. In an orthotropal seed attached to the base of the pericarp it is superior, as also in a suspended anatropal seed. In other anatropal seeds the radical is inferior. When the seed is horizontal as regards the pericarp, the radicle is either centrifugal, when it points to the outer wall of the ovary; or centripetal, when it points to the axis or inner wall of the ovary.

Plants in which there are two cotyledons produced in the embryo are Dicotyledonous. The form of this embryo varies much; and although sometimes resembling in its general aspect that of Monocotyledons, yet it is always distinguished by a division taking place at the cotyledonary extremity (fig. 317), by which two, more or less evident, lobes are formed. The two cotyledons thus formed are opposite to each other (figs. 58 and 320). The cotyledons are not always, however, of the same size. Thus, in a species of Hiraea, one of them is smaller than the other; and in Carapa guianensis there appears to be only one, in consequence of the intimate union which takes place between the two. The union between the cotyledonary leaves may continue after the young plant begins to germinate. Such embryos have been called pseudo-monocotyledonous. The texture of the cotyledons varies. They may be thick, as in the Pea (fig. 58), exhibiting only slight traces of venation, with their flat internal surfaces in contact, and their backs more or less convex; or they may be in the form of thin and delicate laminae, flattened on both sides, and having distinct venation, as in Ricinus, Jatropha, Euonymus, &c. In the former case they are called fleshy, or seminal lobes; in the latter, foliaceous, or seminal leaves. The cotyledons usually form the greater part of the mature embryo, and this is remarkably well seen in such exalbumiuous seeds as the Bean and Pea.

Cotyledons are ’usually entire and sessile. But they occasionally become lobod, as in the Walnut and the Lime, where there are five lobes; or petiolate, as in Geranium molle; or auriculate, as in the Ash. Like leaves in the bud, cotyledons may be either applied directly to each other, or may be folded in various ways. In Geranium the cotyledons a. a twisted and doubled; in Convolvulus they are corrugated; and in the Potato and in Bunias, they are spiral,—the same terms being applied as to the foliage leaves. The radicle and cotyledons may be either straight or curved in various ways. Thus, in some Cruciferous plants, as the Wallflower, the cotyledons are applied by their faces, and the radicle (fig. 321) is folded on their edges, so as to be lateral; the cotyledons are here accumbent. In others, as the Hesperis, the cotyledons (fig. 322) are applied to each other by their faces, and the radicle, r, is folded on their back, so as to be dorsal, and the cotyledons are incumbent. Again, the cotyledons are conduplicate when the radicle is dorsal, and enclosed between their folds. In other divisions the radicle is folded in a spiral manner, and the cotyledons follow the same course.

I In many gymnosperms more than two cotyledons are present, and they are arranged in a whorl. This occurs in Coniferae, especially in the Fir (fig. 323), Spruce, and Larch, in which six, nine, twelve, and even fifteen have been observed. They are linear, and resemble in their form and mode of development the clustered or fasciculated leaves of the Larch. Plants having numerous cotyledons are occasionally denominated pdlycotyledonous. In the gymnospermous genus Welwitschia, there are two cotyledons which last throughout its life (more than a century), and in the course of time they grow to an enormous size, being sometimes six feet long and two or three feet in breadth. They constitute the only leaves of the plant. In species of Strepto∙ carpus the cotyledons are also permanent, and act the part of leaves. One of them is frequently largely developed, while the other is small or abortive.

In those plants in which there is only a single cotyledon (fig. 53) in the embryo, hence called Monocotyledonous, the embryo usually has a cylindrical form more or less rounded at the extremities, or elongated and fusiform, often oblique. The axis is usually very short compared with the cotyledon, which in general encloses the plumule by its lower portion, and exhibits on one side a small slit which indicates the union of the edges of the vaginal or sheathing portion of the leaf (fig. 324). In Grasses, by the enlargement of the embryo in a particular direction, the endosperm is pushed on one side, and thus the embryo comes to lie outside at the base of the endosperm. The lamina of the cotyledon is not developed. Upon the side of the embryo next the endosperm and enveloping it is a large shieldshaped body, termed the scutellum. This is by some authors considered to be an outgrowth from the hypocotyledonary portion of the axis or expansion of the radicular extremity, enveloping more or less the cotyledon and plumule, in some cases, as in Maize, completely investing it; in other cases, as in Rice, merely sending small prolongations over its anterior face at the apex. By others this scutellum is considered as the true cotyledon, and the sheathing structure covering the plumule is regarded as a ligule or axillary stipule. In several other Monocotyledonous plants, as Ruppia and Zostera, this scuteliar structure is well seen, and in these cases its homology, as an expansion of the radicular extremity of the embryo, is clearly discernible; and this is further borne out by such cases as that of Caryocar butyrosum among Dicotyledons, where the radicular swelling occupies most of the embryo (fig. 326). In some Grasses, as Oats and Rice, a projection of cellular tissue is seen upon the side of the embryo opposite to the scutellum, that is, on the anterior side. This has been termed the epiblast. It is very large in Rice. This by some was considered the rudimentary second cotyledon; but this is impossible, as it arises outside, and after the first cotyledon. It is merely an outgrowth of the radicular extremity like the scutellum. The radicular portion of the axis is usually shorter than the cotyledon, and more dense in structure; but in some instances it becomes much larger, giving rise to what has been called a macropodous embryo.

5. Female Organs and Reproduction in Cryptogams.

The female organs of reproduction, like the male organs, have not been demonstrated in all Cryptogams. In all vascular Cryptogams, and in Mosses and Hepaticεe, certain usually flask-shaped bodies, which have been termed archegonia or pistillidia (fig. 327), represent the female organs producing cells or spores, which geπninate and form new plants (fig. 54). These archegonia have the general structure of a large cell, the central cell or oosphere (c) surrounded by a layer of smaller cells. From the apex of this oosphere leads a canal, which is bounded by four rows of small cells and constitutes the neck, and in it is a large cell full of soft mucilaginous matter, which has been formed from the central cell, and is the canal cell. Upon the wall of the oosphere, turned towards the neck, a small portion different from the surrounding part is the receptive spot. Fertilization is effected by the anthérozoïde freed from the antheridium, penetrating the neck of the archegonium, and eventually reaching the receptive spot of the central cell or oosphere; they then enter the oosphere and coalesce with it. The fertilized oosphere is termed an oospore, and it then may escape from the archegonium, and sooner or later germinates; or it germinates within the archegonial cells. These archegonia are, in vascular Cryptogams, produced upon a cellular expansion formed asexually from a spore, and termed the prothallus (flζ. 245), which is of varying size and form. Both archegonia and antheridia may be formed on one prothallus, or only one form of organ may be produced, thus indicating a tendency to diclinisπι.

In Characeae the female organ has a peculiar structure, and is termed a nucule (fig. 240, ?i). This consists of a large central cell, of which the contents a't the apex are clear and hyaline, while the lower part contains much starch and fat. Rising from its base and twisting round it are five long tubes (fig. 328), at the extremity of each of which abovo the apex of the central cell is a single short cell in Chara, while in Nitella it is divided transversely into two. These five or ten short cells form the crown. They unite together so that their apices project as small teeth. Between the crown and the apex of the central cell, which is termed the apical papilla, a cavity is included, bounded at the sides by the five enveloping tubes, which at this point form the of the nucule. The cavity of the neck is constricted in the middle by the projection inwards of the tubes to form a sort of diaphragm, so that the enclosed cavity has a rude hour-glass-like shape, the upper cavity closed above by the crown being connected by a narrow canal with a lower cavity bounded beneath by the apical papilla of the central cell. When fertilization takes place slits are formed betwixt the five tubes of the neck, beneath the crown and above the diaphragm. The antherozoids from the globule enter by them into the upper cavity, pass through the canal into the lower one, and fuse with the apical papilla of the central cell or oosphere. The oospore thus formed becomes detached from the plant, being covered by the thickened inner wall of the tube which invested it. The nucule rises from the base of one of the whorl leaves in Chara, and is therefore above the globule; in Nitella it is produced upon the leafy axis beneath the globule.

In Mosses the archegonia are frequently situated along with the antheridia and paraphyses. They are surrounded by the same whorl of leaves, or perichoetium, when the moss is said to be hermaphrodite, or they occur separately on the same or on different plants, the moss being then monoecious or dioecious. The term perigone is applied to the whorl of leaves around the antheridia. The basal portion of the archegonia which surrounds the oosphere is termed the epigone. In Hepaticae the archegonia (fig. 329) are situated in the substance of the thallus itself, or they may be in various situations, as in Jungermannia (fig. 330); in Marchantia they are produced upon the under surface of a stalked stellate disk (fig. 331).

In many Thallogens no structure analogous to a female organ has been as yet discovered; in some, however, such structures have been met with. Thus amongst Algae large cells, termed oogonia, are found, in which usually one, sometimes many (Fucaceae), oosphères are produced (fig. 332). The antherozoids from the antheridium fertilize these, penetrating the oogonia at a definite point thinner than the surrounding portion; and oospores are formed. In the Florideae, a group of Algae, the organs are different in character. Here the antherozoids are not motile, having no cilia. On discharge from the antheridium, these are washed into a long filiform hollow body, termed the trichogynium. This is supported usually upon a cellular stalk—the tτichophore— at the side of the base of which is a small cellular mass. The antherozoids having entered the trichogynium, fertilization is effected, and results in the formation, from the cellular mass at the base of the trichophore, of a large cell, the cystocarp, in which spores are formed. In the Saprolegnieae the antheridia actually penetrate into the oogonium and discharge their antherozoids. Amongst Fungi it is rare to find sexual organe. In some thecasporous fungi, as Eurotium, a sexual process has been observed. The female organ, here termed ascogonium or carpogonium, is of a spiral elongated character, immediately surrounding which are the antheridia, few in number, which open into the apex of the female organ, and discharge their antherozoids. The result is the formation of cellular tissue round the ascogonium, forming what is termed a perithecium, and within the ascogonium large cells (asci) are produced, in which spores or sporidia are subsequently developed. In Lichens no sexual organs have as yet been clearly demonstrated.

Amongst Fungi and Lichens there are certain bodies to be noted, which may be connected with sexual reproduction, but the function of which is at present unknown. Embedded in the margin of the thallus in Lichens and on various parts of the plant in Fungi, certain hollow urnshaped bodies are found, which have been termed spermagones (conceptacles) (fig. 334). These are usually closed all but a small opening at the apex. From their base rise peg-shaped projections or sterigmata (fig. 335), bearing at their summit elongated, sometimes curved, bodies, termed spermatia, which, when mature, are discharged through the apex of the spermagone. These are by some considered the analogues of the anthérozoïde in other Cryptogams. There are also found other conceptacles, to which the name of pycnidia or pycnides is given. These contain large spores, which have been termed stylospores (fig. 336), but their homology is at present unknown.

Another process of reproduction is seen amongst Cryptogams. This is termed conjugation (fig. 333). It differs from the process of fertilization just described in that it consists of a union of the contents of two similar cells, whilst fertilization essentially consists in the mutual influence of dissimilar cells. It is well seen in many Algae, such as Diatoms, Volvocineae, Conjugatae, also in some Fungi, such as species of Rhizopus and Zyzygites. It may consist in the coming together of two cells,—either moving spores, as in Pandorinia, or motionless, as in Diatoms,—which completely fuse, and a compound spore, or zygospore, results; or prolongations of the wall of two adjacent cells meet, the partition between the two gives way and the contents from the one pass into the other, and a zygospore is formed, as in Zygnema. Zygospores germinate only after a long period of rest.

Amongst Thallogens especially, though it also occurs in other Cryptogams, the asexual mode of reproduction is very common. But the manner of formation and the nature of the spores is diversified.

Amongst Algae two modes of asexual reproduction are seen,—by motile spores and by motionless or resting spores. In the former case the contents of a cell form a new cell, which escapes from the mother-cell, and moves about by means of cilia, which are formed either all round it, as in Vaucheria (fig. 337), or as a tuft at one end, as in Oedogonium. These are termed swarm spores, or zoospores, and frequently are of different sizes, being termed respectively microgonidia and macrogonidia. After swimming about for a time they fix themselves, and develop small rhizoids from one extremity, the other growing up into the plant. The motionless spores are seen in the Florideae, where they are formed in cells, four in each, attached together in a row, or as a tetrahedron, constituting the tetraspore (fig. 338).

Amongst Fungi asexual reproduction is very common. Swarm spores are rarely formed; but the mode of formation of motionless spores is very various. They may be produced in the interior of distinct sacs called thecoe, asci, or cystidia (fig. 339), when they are denominated endospores or ascospores, and the plants are said to be thecasporous or ascosporous. When they are developed on the exterior of sacs called basidia (fig. 340), they are denominated exospores, and the plants are basi- diosporous. When produced in the midst of a gelatinous mass, without any evident differentiation, they are called myxospores, the plants being myxosporous. Both the endospores and the exospores may by division become septate, and form compound spores, each of the secondary spores so formed being termed a merispore. Four or eight basidiospores are usually produced from each basidium. In the asci numerous spores are formed. Other forms of spores seen in Fungi are the conûiïct. These are stalked spores which arise from a mycelium.

In Lichens we find asci which produce spores or sporidia just as in Fungi; and in addition there also exist spermogones with their sterigmata and spermatia as well as pycnides and stylospores. Besides the spores there are organs called soredia in Lichens. These are groups of small round cells or gonidia (fig. 97), surrounded by hyphae, which are capable of independent growth after removal from the thallus. These gonidia are now regarded by many as an Alga, upon which the hypha, which is supposed to be an ascomycetous Fungus, is parasitic.

In Chara a curious mode of asexual reproduction is seen which presents several well-marked modifications, the process always proceeding at the nodes, and consisting in the growth of a new axis from a nodal cell either isolated or united with other cells.

In all vascular Cryptogams, Muscîneae, and many Thallogens, that process of development in which successive different generations are produced finally returning to the original form, usually termed alternation of generations, is well seen. The number of such generations varies.

Amongst Ferns the oospore germinates and produces an asexual generation. The embryo plant formed from this spore is a simple mass of cellular tissue, no cotyledons being formed. This, then, is an Acotyledonous embryo (fig. 54). The same happens in other vascular Cryptogams, which are therefore also Acotyledonous. The embryo so formed has a primary meristem, with an apical cell, by division of which it increases in size,—roots being formed downwards and a leaf-bearing axis upwards, upon which, eventually, the fructification (fig. 341) is formed. This consists in Ferns of the production of either stalked or sessilo spore-cases or sporangia, from the epidermis on the back (Aspidium) cr edges (Pteris) of a leaf (frond), or upon a special branching axis (Osmunda), or a spicate one (Ophioglossum). These, when situated on the frond, arc arranged in definite clusters or sori (fig. 342), either rounded or elongated. With the sporangia are usually associated jointed cells or paraphyses. Frequently the sori are covered by a thin cellular layer forming an indusium or involucre, or the margin of the frond folds over them, forming a false indusium. Within the sporangium spores are produced, and when ripe these escape by the rupture of the sporangium, this in many cases being provided for by the formation, at a definite point, of an annulus or ring (fig. 343) of unequally thickened cells. Hence some Ferns are annulate, others exannulate. The cell-wall of each spore is divisible into an outer layer or exospore, and an inner or endospore. When the spore reaches the soil it germinates, and after a longer or shorter period produces t he sexual generation. In this process the variously marked exospore is ruptured by the enlarging endospore, which divides so as eventually to form a flattened cellular expansion or prothallus (fig. 344) with small cellular rhizoids. Upon this, usually on the under surface, the sexual organs, consisting cf antheridia and archegonia, are produced. A tendency to dioecïsm is observed in prothalli, frequently producing only one kind of sexual organs. By the mutual influence of those sexual organs an oospore is formed, from which a new asexual generation arises. Thus in Ferns the alternation consists of two dissimilar generations,—a sexual or prothallial generation, and an asexual generation.

In Equisetaceae the same alternation of two dissimilar generations occurs. Upon an asexual generation the fructification is formed at the apex of a leaf-bearing axis. Peltate hexagonal stalked scales (fig. 345) are arranged in close whorls round the apex of the axis, forming a pyramidal head (fig. 346, ∕), the surface of the scales being directed tangentially to the axis. Round the margin of tho inner surface of these scales the sporangia are produced projecting towards the axis. When ripe each sporangium opens towards the stalk of the scale, and the spores are shed, each having attached to it four clavate filaments sometimes termed elaters (fig. 3), which are very hygroscopic. By germination of these spores a sexual generation or prothallus is formed.

In Lycopodiaceae a sexual process has not been observed in all genera; but in Selaginella and Isoetes, where it has been observed, it is exceedingly instructive, as it forms a connecting link between the process as seen in Cryptogams and that observed in Angiosperms. The asexual generation produces the fructification which consists in the formation of a single sporangium in the axil of each leaf of a certain part of the axis. In those genera iu which sexual organs have been discovered these sporangia are of two sizes, termed macrosporangia (fig. 347) and microsporangia respectively. The whole fructification frequently forms a spike. In the macrosporangium or oophoridium few, usually four, largo spores (macrospores') are formed, having each an exospore and an endospore. In the microsporangium (fig. 248) numerous small spores or microspores are formed of a similar structure. When the sporangia are shed the capsule decays, and the spores escape. Within the endospore of the macrospore, a development of cellular tissue takes place which eventually completely fills it and constitutes Ülq prothallus, termed sometimes on this account endothallus (fig. 348). As it increases in size the exospore is ruptured triradiately, and the endospore projecting is also ruptured, and thus the prothallus is protruded, in the substance of which archegonia are formed. The microspores, on the other hand, produce within their endospore, sometimes after the formation of a rudimentary prothallus, antherozoid mother-cells, which are discharged by the rupture of the exospore and endospore (fig. 249). By fertilization of the oosphere of the archegonium an oospore is formed, from which the sexual generation arises. The embryo is formed from the lower part of the oospore, the upper half elongates and forms a sτιspensor (fig. 349) of several cells, by which the embryo is pushed downwards into the substance of the prothallium, and there develops into a young plant, rudimentary cotyledons being formed upon the young axis. It will be observed that this process resembles very much the mode of embryogeny in Phanerogams, and, indeed, from this it is but a step to the form found in Gymnosperms and thence to Phanerogams.

In Rhizocarps the sporangia are formed in the interior of ovoid sacs, termed sporocarps (fig. 350), which arc usually supported on stalks, and rise from the leaves near their base, one or many on the same stalk. In each sporocarp, microsporangia and macrosporangia are formed in various positions; and a prothallium (fig. 351) is formed iu the macrospore, as in Lycopodiaceae. From the oospores formed in it the embryos are produced, but without suspensors.

In Characeae there is an alternation of a sexual with an asexual condition. From the oospore a small proembryo is first formed, from a cell of which the sexual plant is developed.

In Mosses the sexual generations are produced as lateral shoots from a filamentous proenibryo or protonema (fig. 352). Upon this leaf-bearing axis either at the apex (acrocarpous) or laterally (pleurocarpous) the sexual organs arise. The asexual generation rises from the oospore and constitutes what is usually termed the fruit (fig. 353) of the moss. It has been designated the sporogonium. The sporogonium is formed by the enlargement of the oospore into an ovoid embryo which presses against the epigone, finally rupturing it, and carries the upper portion upwards with it as the calyptra or cap, the lower portion forming a sheath or vaginula at the base. This calyptra may be either split on one side, dimidiate, or entire, then termed mitriform. The upper part of the sporogonium forms a capsule, um l or theca; the lower portion acte as a supporting stalk or seta, which is often swollen, just beneath the capsule forming the apophysis (fig. 354, a). In Sphagnum the sporogonium is fully developed within the epigonal leaves, and when complete the axis beneath it elongates, forming the pseudopodium, and projects considerably beyond the perichaetium. The walls of the capsule of the sporogonium consist of several layers of cells. It has a central axis or columella, surrounding which aro the spores. When mature the capsule usually dehisces, either by four longitudinal valves, as in Andraea, or most commonly by a lid or operculum (fig. 355), between which and the capsule an annulus or ring of cells is separated. On removal of the operculum the stoma or opening of the capsule is seen. The margin of this is sometimes entire (gymnostomn), more usually fringed with teeth or cellular prolongations, constituting a peristome (fig. 355, p), cither in one row (aploperistomi) or in two rows (diploperistomi). Stretching across the mouth of the capsule is frequently seen a thin cellular membrane, the epiphragm or tympanum, as in Polytrichum. The spores shed from the capsule germinate and give rise to the proembryo which by budding gives off the sexual plants.

In Hepaticae the sexual generation anses either directly from the spore of the asexual generation, or a small proembryo is first formed. The sporogonium is of various forms, and originates differently in different groups from the oospore. It is always formed within a cellular covering or epigone. In Anthoceros the sporogonia are elongated and open by two valves, and there is a central cellular columella. In Jungermannia the sporogonia are globular, and open by four spreading valves. There is no central column, but the inner wall displays cellules, called elaters, fixed by one extremity, and containing one or two spiral fibres which by their elasticity scatter the spores. In Riccia the sporogonia are globular, and there are neither elaters nor columella.

Very little is known of the sexual reproduction of Fungi, and regarding the alternation of sexual and asexual generations in them. But we have abundant instances of alternation amongst the asexual generations, and along with this of that curious phenomenon included under the term heteroecism, which consists in the growth of one generation of a parasitic Fungus upon one host, and the development of another generation upon a different host. By tracing out this phenomenon, many forms, described formerly as distinct species, are now shown to be generations of one and the same Fungus. The asexual spores of Fungi are produced upon certain structures which have been frequently termed the fructification. For instance, in basidiosporous Fungi the collection of basidia forming the hymenium is usually situated on large receptacles. The structure commonly known as a fungus (fig. 356) is a receptacle produced upon a mycelium (my). In its young state it appears as a tubercle upon the mycelium covered by a volva (fig. 356, vol), or wrapper, which it bursts during growth. It consists of a pileus or cap (p), at the extremity of a stalk, the stipe (st). On the under surface of the pileus are the lamelloe (la) or gills, of various forms, which constitute the hymenial (hy) layer, to which the spores are attached. At first the hymenium is covered by a veil or indusium, which during the growth of the stipe is ruptured, the base remaining on the stipe as the annulus (an) or ring. When, as in the Mushroom, the hymenium is exposed, the fungus is said to be gymnocarpous (fig. 356). In other cases, as in Phalloideae, it is covered over by a peridium, and the whole mass so enclosed is the gleba, and the fungus is angiocarpous (fig. 357). In .ascosporous Fungi, as Sphaeria, the asci are usually formed within small roundish receptacles, termed perithecia; and along with the asci are usually numerous abortive filaments or paraphyses. Sometimes these perithecia are situated upon cushions or branching masses of tissue, the stroma. In the myxosporous Fungi the plasmodium, or whole protoplasmic mass of the plant, forms receptacles, in each of which is a capillitium, or network of fine filaments, in the interstices of which are numerous spores.

In Lichens spores are formed in asci, which are arranged along with paraphyses, forming a hymenium in receptacles situated on the thallus, and termed apothecia or lirellae, which are sometimes (fig.' 95) supported on a stalk or podetium. The hyphae, from which the asci arise, form a layer immediately beneath the hymenium, termed the subhymenial layer, the thick mass of hyphae beneath being the hypothecium. The outermost layer of the tissue of the apothecium is termed the excipulum.

Amongst Algae alternation of generations is very common; and here, as in the case of Fungi, it may pass through several generations before reaching the original form. This is well seen in Diatoms, also in Oedogonieae, where three generations are passed through, and in Coleochaete, where there are four dissimilar generations,— sexual and various forms of asexual reproduction alternating. (j. h. B.) [table]

Explanation of the Plates.

Plate I. Papaver Rhceas, Common Red Poppy, belonging to the Nat. Ord. Papaveraceae. The leaves are pinnatifid, the peduncles have spreading hairs, the calyx consists of 2 caducous sepals, enclosing 4 crumpled petals, the stamens are indefinite and hypogynous, and the ovary is surmounted by a radiating stigma.

Fig. 1. Stamens inserted on the thalamus, below the ovary with its sessile stigma. Fig. 2. Capsule opening by pores below the sessile radiating stigmas.

Plate II. Janipha Manihot and Eschscholtzia californica.

Fig. 1. Janipha Manihot, Cassava plant, belonging to the Nat. Ord. Euphorbiaceae. The leaves are digitately-partite, and the flowers are in racemose cymes. Fig. 2. Racemose cyme, with a pentacoccous capsule, which separates in an elastic manner into 5 single-seeded carpels. The cyme bears male as well as female flowers. Fig. 3. Pistil with stigma. Fig. 4. Stamens and fleshy disk. Fig. 5. Seed with strophiole.

Fig. 6. Eschscholtzia californica (nett. size), belonging to the Nat. Ord. Papaveraceae. It has a peculiar caducous calyx like a candle-extinguisher, and a tetrapetalous corolla, with numerous hypogynous stamens. Fig. 7. Hollowed end of the peduncle, with the pistil. Fig. 8. Section of ovary, with numerous seeds attached to 2 parietal placentas. Fig. 9. Ceratium or siliquaeform capsule (ηeιt. size), opening by 2 valves. Fig. 10. Section of the seed, with the Dicotyledonous embryo, which is shown separately in fig. 11.

Plate III. Malva sylvestris, Common Mallow, belonging to the Nat. Ord. Malvaceae. The leaves have 5 lobes, the inflorescence consists of cymose fascicles, and the aestivation is contortive.

Fig. 1. Calyx with a three-leaved epicalyx or involucre. Fig. 2. Obcordate petal, with monadelphous stamens. Fig. 3. Tube of stamens formed by union of the filaments. Fig. 4. Pistil with numerous carpels and styles. Fig. 5. Stamen with reniform anther opening round the margin. Fig. 6. Section of ovary composed of numerous carpels.

Plate IV. Anacardium occidentale, Cashew-nut plant, belonging to the Nat. Ord. Anacardiaceae.

Fig. 1. Branch (somewhat reduced), bearing flowers and fruit. The flowers are in cymes, and the peduncles are enlarged in a pear-like form, bearing the nut (the true fruit) at their apex. Fig. 2. Flower not expanded. Fig. 3. Flower expanded. Fig. 4. Stamen and pistil, with the calyx; one fertile stamen longer than the others. Fig. 5. Stamen separated. Fig. 6. Nut constituting the fruit. Fig. 7. Nut opened longitudinally. Fig. 8. Seed separated from the nut. Fig.' 9. Cotyledons, opened to show the radicle a, and the plumule.

Plate V. Sarothamnus scoparius, Common Broom, belonging to the Nat. Ord. Leguminosae or Fabaceae, Sub-ord. Papilionaceae. The angled branches bear ternate leaves, papilionaceous flowers, and legumes.

Fig. 1. Two-lipped calyx. Fig. 2. Broadly ovate vexillum or standard. Fig. 3. One of the alae or wings of the corolla. Fig. 4. Carina or keel. Fig. 5. Monadelphous stamens. Fig. 6. Hairy ovary with the long style, thickened upwards, and spirally curved. Fig. 7. Legume or pod.

Plate VI. Carica Papaya and Andromeda hypnoides.

Fig. 1. Carica Papaya, the Papaw tree (much. reduced), belonging to the Nat. Ord. Papayaceae. The leaves are palmately-cleft, and the flowers unisexual. Fig. 2. Portion of a racemose cyme of infundibuliform male flowers, with united petals. Fig. 3. Gamopetalous male flower cut open, showing the ten epicorollinc stamens alternately shorter. Fig. 4. Stamen. Fig. 5. Female flowers with a deeply-5-parted corolla.

Fig. 6. Andromeda hypnoides (nat. size), belonging to the Nat. Ord. Ericaceae. Fig. 7. Flower of Andromeda, with 5-parted calyx, and gamopetalous campanulate corolla. Fig. 8. Back view of stamen with its 2-horned anther. Fig. 9. Front view of stamen with bicornute and biporose anther. Fig. 10. Pistil with its ovate acuminate style. Fig. 11. Capsule, 5-celled, with a central 5-lobed placenta.

Plate VII. (Enanthe crocata, Hemlock Watcr-Dropwort, belonging to the Nat. Ord. Umbellifcrae. Leaves compound and sheathing, roots composed of fusiform sessile ι knobs, stem hollow, and flowers in compound umbels, with general and partial involucres.

Fig. 1. Flower composed of 5 petals, with inflexed apicula, 5 stamens, and 2 styles with an epigynous disk. Fig. 2. Lateral view of the cremocarp composed of 2 mericarps or achenes, with blunt convex ribs, sunnounted by the lanceolate teeth of the calyx, and two styles with a disk at their base. Fig. 3. Back view of the cremocarp.

Plate VIII. Valeriana officinalis, Great Wild Valerian, belonging to the Nat. Ord. Valerianaccae, showing the roots, which are officinal, the hollow stem, the alternately∙pinnate leaves, and flowers in corymbose cymes.

Fig. 1. Separate flower, consisting of an adherent calyx, with an obsolete limb, an irregular sympetalous corolla, exserted stamens, and 1 style. Fig. 2. Pistil separated, with ovary, style, and stigma. Fig. 3. Monospeπnal fruit, with the persistent pappose limb of the calyx.

Plate IX. Leontodon Taraxacum, Common Dandelion, belonging to the Nat. Ord. Composite, Sub-ord. Cichoraceae. The leaves are radical and runcinate, the flowers ligulate, and arranged in capitula. The outer phyllaries (involucral leaves) are reflexed.

Fig. 1. Capitulum with reflexed phyllaries and ligulate flowers. Fig. 2. A single flower, with inferior achene (cypsela), stipitate pappus, a ligulate corolla, 5 stamens united by their anthers, and 1 style with 2 stigmas. Fig. 3. Receptacle, become dry and convex, with deflexed phyllaries. Fig. 4. Linear-obovate achene. Plate X. Cuscuta verrucosa, Warty Dodder, belonging to the Nat. Ord. Cuscutaceae, which is often considered as a sub-order of Convolvulaccae.

Fig. 1. Leafless flowering stem of Dodder (pat. size), turning from right to left (contrary to the motion of the sun). Fig. 2. Campanulate corolla, with 5 epicorollinc stamens and alternating scales. Fig. 3. Persistent calyx. Fig. 4. Capsule opening transversely near the base. Fig. 5. Dissepiment and 2 seeds. Fig. 6. Bilocular capsule cut transversely, showing 2 seeds in each loculament. Fig. 7. Roundish compressed seed. Fig. 8. Seed cut longitudinally, showing the perisperm with the spirally-rolled-up embryo. Plate XL Digitalis purpurea, Purple Foxglove, belonging to the Nat. Ord. Scrophulariaceae.

Fig 1. Calyx in 5 deep segments, with the pistil. Fig. 2. Pistil separated. Fig. 3. Didynamous stamens. Fig. 4. Stamens, with the anther-lobes collateral and unopened. Fig. 5. Capsule cut transversely, showing two loculaments.

Plate XII. Myristica moschata, the Nutmeg-tree, belonging to the Nat. Ord. Myristicaceae.

Fig. 1. Male plant (pat. size) with oblong ’ nearly elliptical leaves, and flowers in axillary umbellate cymes. Fig. 2. Perianth of male plant laid open to show the stamens with their united filaments. Fig. 3. Anther, bilocular, and opening longitudinally. Fig. 4. Female flower cut open, showing the single pistil, with short style and 2-lobed stigma. Fig. 5. Young fruit. Fig. 6. Ripe drupaceous fruit in the act of bursting, and showing the mace or arillode. Fig. 7. Section of the ripe fruit. Fig. 8. The nut, consisting of a hard endocarp enclosing the seed. Fig. 9. The oval seed. Fig. 10. The seed cut vertically. Fig. 11. Vertical section of endocarp and seed, showing ruminate perisperm and the embryo. Fig. 12. Embryo, showing 2 large foliaceous plicate cotyledons. Fig. 13. Embryo with radicle and cotyledons. Plate XIII. Euphorbia hypericifolia, Tutsan-leaved Spurge, belonging to the Nat. Ord. Euphorbiaceae.

Fig. 1. Plant (ηat. size) with opposite oblong leaves, and flowers in corymbose cymes. Fig. 2. Involucre with appendages, enclosing male and female flowers. Fig. 3. Involucre cut open, to show the insertion of the numerous achlamydeous male monandrous flowers, and the single naked female flower, all pedicellated. Fig. 4. Involucre deprived of its petaloid appendages, with female flower showing forked styles. Fig. 5. Involucre enclosing the ripe fruit (regma). Fig. 6. Stamen. Fig. 7. Fruit with one of the cocci separated from the columella. Fig. 8. Seed. Fig. 9. Seed cut open, showing fleshy perisperm and inverted embryo.

Plate XIV. Artocarpus incisa, the Bread-fruit tree, belonging to the Nat. Ord. Artocarpaceae.

Fig 1. Branch reduced to -Jrd its natural size, with cuneate-ovate pinnatifid leaves, male flowers in a club-shaped deciduous catkin, female ilow ers in rounded clusters. Fig. 2. Transverse section of the male amentum with numerous flowers. Fig. 3. Male flowers. Fig. 4. Single male flower separated, with a perianth in 2 segments and a single stamen. Fig. 5. Female flowers. Fig. 6. Single female flower separated, with ovary, style, and bilid stigma. Fig. 7. Ovary. Fig. 8. Ovary laid open to show the ovule. Fig. 9. A variety of the ovary with 2 loculaments. Fig. 10. Transverse section of a bilocular ovary.

Plate XV. Cycas circinalis, a species of false Sago-plant, belonging to the Nat. Ord. Cycadaceae.

Fig. 1. Plant reduced to τ 1 2 th its natural size. The leaves are pinnately-divided and form a crown round the stem. Fig. 2. Male amentum (ηa,t. size), consisting of scales bearing on their under side 2-valved anthers. Fig. 3. Upper* side of one of the scales. Fig. 4. Under side of a scale, bearing anthers. Figs. 5, 6, and 7. Anther. Fig. 8. Pollen magnified.

Plate XVI. Arum maculatum, Cuckow-pint or Wake Robin, belonging to the Nat. Ord. Araceae. Fig 1. Flowering stalk, with hastate-sagittate spotted leaves, and spathe enclosing a spadix bearing male and female flowers. Fig. 2. Amylaceous corm, whence arise buds and roots. Fig. 3. Berried fruit, after leaves and spadix have decayed, in an oblong spike, of a bright scarlet colour.