forms-the zoogloea is now known to be a sort of resting condition of the Schizomycetes, the various elements being glued together, as it were, by their enormously swollen and diffluent cell-walls becoming contiguous. The zoogloea is formed by active division of single or of several mother-cells, and the progeny appear to go on secreting the cell-wall substance, which then absorbs many times its volume of water, and remains as a consistent matrix, in which the cells come to rest. The matrix i.e. the swollen cell-walls-in some cases consists mainly of cellulose, in others chiefly of a proteid substance; the matrix in some cases is horny and resistant, in others more like a thick solution of gum. It is intelligible from the mode of formation that foreign bodies may become entangled in the gelatinous matrix, and compound zoogloeae may arise by the apposition of several distinct forms, a common event in macerating troughs (fig. 3, A). Characteristic forms may be assumed by the young zoogloea of different species,-spherical, ovoid, reticular, filamentous, fruiticose, lamellar, &c.,-but these vary considerably as the mass increases or comes in contact with others. Older active along the whole length; (2) there are optimum, minimum and maximum temperatures, other conditions remaining constant, at which growth begins, runs at its best and is soon exhausted, respectively; (3) that the most rapid cell-division and maximum growth do not necessarily accord with the best conditions for the life of the organism; and (4) that any sudden alteration of temperature brings about a check, though a slow rise may accelerate growth (fig. 8). It was also shown that exposure to light, dilution or exhaustion of the food-media, the presence of traces of poisons or metabolic products check growth or even bring it to a standstill; and the death or injury of any single cell in the filamentous series shows its effect on the curve by lengthening the doubling period, because its potential progeny have been put out of play. Hardy has shown that such a destruction of part of the filament may be effected by the attacks of another organism. The growth of an ordinary bacterium consists in uniform elongation of the rodlet until its length is doubled, followed by division by a median septum, then by the simulMeasure ment of taneous doubling in length of each daughter cell, again followed by the median division, and so on (figs. 13, 14). growth. If the cells remain connected the resulting filament repeats these processes of elongation and subsequent division uniformly so long as the conditions are maintained, and very accurate measurements have been obtained on such a form, eg. B.ramosus. If a rodlet in a hanging drop of nutrient gelatine is fixed under the microscope and kept at constant temperature, a curve of growth can be obtained recording the behaviour during many hours or days. The measured lengths are marked off on ordinates erected on an abscissa, along which the times are noted. The curve obtained on joining the former points then brings out a number of facts, foremost among which are (1) that as long as the conditions remain constant the doubling periods-i.e. the times taken by any portion of the filament to double its length-are constant, because each cell is equally 080 60 Temp рр 40 20 0 535 35 330 darkness 2-70 30 A 20 lamp dark 5.45 6.20 77 80 847 9940 105 10 40 853 927 956 10 30 Psy 10p 12p 2 pm. 233 (Fraenkel), constructed from data such as in fig. 4. The abscissae FIG. 8.-Curve of growth of a filament of Bacillus ramosus represent intervals of time, the ordinates the measured lengths of the growing filament. Thus, at 2.33 P.M. the length of the filament was 6; at 5.45, 20 μ; at 8 P.M., 70 and so on. Such curves show differences of steepness according to the temperature (see temp. curve), and to alterations of light (lamp) and darkness. (H. M. W.) A very characteristic method of reproduction is that of sporeformation, and these minute reproductive bodies, which represent a resting stage of the organism, are now known in many Spores. forms. Formerly two kinds of spores were described, arthrospores and endospores. An arthrospore, however, is not a true spore but merely an ordinary vegetative cell which separates and passes into a condition of rest, and such may occur in forms which form endospores, e.g. B. subtilis, as well as in species not known to form endospores. The true spore or endospore begins with the appearance of a minute granule in the protoplasm of a vegetative cell; this granule enlarges and in a few hours has taken to itself all the protoplasm, secreted a thin but very resistive envelope, and is a ripe ovoid spore, smaller than the mother-cell and lying loosely in it (cf. figs. 6, 9, 10, and 11). In the case of the simplest and most minute Schizomycetes Classifica (Micrococcus, &c.) no definite spores have been discovered; any | from the year 1872, when Cohn published his system, which was one of the vegetative micrococci may commence a new series of cell by growth and division. We may call these forms asporous," at any rate provisionally. The spore may be formed in short or long segments, the cellwall of which may undergo change of form to accommodate itself to the contents. As a rule only one spore is formed in a cell, and the process usually takes place in a bacillar segment. In some cases the spore-forming protoplasm gives a blue reaction with iodine solutions. The spores may be developed in cells FIG. 9. B which are actively swarming, the movements not being interfered with by the process (fig. 4, D). The so-called "Köpfchenbacterien" of older writers are simply bacterioid segments with a spore at one end, the mother cell-wall having adapted itself to the outline of the spore (fig. 4, F). The ripe spores of Schizomycetes are spherical, ovoid or long-ovoid in shape and extremely minute (e.g. those of Bacillus subtilis measure 0.0012 mm. long by 0.0006 mm. broad according to Zopf), highly refractive and colourless (or very dark, probably owing to the high index of refraction and minute size). The membrane may be relatively thick, and even exhibit shells or strata. The germination of the spores has now been observed in several forms with care. The spores are capable of germination at once, or they may be kept for months and even years, and are very resistant against desiccation, heat and cold, &c. In a suitable medium and at a proper temperature the germination is completed in a few hours. A, Bacillus anthra- The spore swells and elongates and the cis. (After de Bary.) contents grow forth to a cell like that which Two of the long fila- produced it, in some cases clearly breaking ments (B, fig. 10), in through the membrane, the remains of being developed. The which may be seen attached to the young specimen was culti- germinal rodlet (figs. 5, 9 and 11); in other vated in broth, and cases the surrounding membrane of the spores are drawn a little too small spore swells and dissolves. The germinal they should be of the cell then grows forth into the forms typical same diameter trans- for the particular Schizomycete concerned. versely as the seg- The conditions for spore-formation differ. which spores are ments. B. Bacillus sub- Anaerobic species usually require little tilis. (After de oxygen, but aerobic species a free supply. Bary.) 1, fragments Each species has an optimum temperature of filaments with ripe and many are known to require very special spores; 2-5, successive stages in the food-media. The systematic interference germination of the with these conditions has enabled bacteriospores, the remains logists to induce the development of soof the spore at called asporogenous races, in which the tached to the ger- formation of spores is indefinitely postponed, changes in vigour, virulence and other properties being also involved, in some cases at any rate. The addition of minute traces of acids, poisons, &c., leads to this change in some forms; high temperature has also been used successfully. minal rodlets. The difficult subject of the classification of bacteria dates The difficulties presented by such minute and simple organisms as the Schizomycetes are due partly to the few" characters which they possess and partly to the dangers of error in manipulating them; it is anything but an easy matter either to trace the whole development of a single form or to recognize with certainty any one stage in the development unless the others are known. This being the case, and having regard to the minuteness and ubiquity of these organisms, we should be very careful in accepting evidence as to the continuity or otherwise of any two forms which falls short of direct and uninterrupted observation. The outcome of all these considerations is that, while recognizing that the " genera and species" as defined by Cohn must be recast, we are not warranted in uniting any forms the continuity of which has not been directly extended in 1875; this scheme has in fact dominated the study of bacteria ever since. Zopf in 1885 proposed tos. a scheme based on the acceptance of extreme views of pleomorphism; his system, however, was extraordinarily C (Fraenkel), in the order and at the times given, in a hanging drop FIG. 11.-Stages in the development of spores of Bacillus ramosus culture, under a very high power. The process begins with the formation of brilliant granules (A, B); these increase, and the brilliant substance gradually balls together (C) and forms the spores (D), one in each segment, which soon acquire a membrane and ripen (E). (H. M. W.) to liquefy gelatine, to secrete coloured pigments, to ferment certain media with evolution of carbon dioxide or other gases, or to induce pathological conditions in animals. None of these systems, which are chiefly due to the medical bacteriologists, has maintained its position, owing to the difficulty of applying the characters and to the fact that such properties are physiological and liable to great fluctuations in culture, because a given organism may vary greatly in such respects according to its degree of vitality at the time, its age, the mode of nutrition observed; or, at any rate, the strictest rules should be followed in accepting the evidence adduced to render the union of any forms probable. and the influence of external factors on its growth. Even when used in conjunction with purely morphological characters, these physiological properties are too variable to aid us in the discrimination of species and genera, and are apt to break down at critical periods. Among the more characteristic of these schemes adopted at various times may be mentioned those of Miquel (1891), Eisenberg (1891), and Lehmann and Neumann (1897). Although much progress has been made in determining the value and constancy of morphological characters, we are still in need of a sufficiently comprehensive and easily applied scheme of classification, partly owing to the existence in the literature of imperfectly described forms the life-history of which is not yet known, or the microscopic characters of which have not been examined with sufficient accuracy and thoroughness. The principal attempts at morphological classifications recently brought forward are those of de Toni and Trevisan (1889), Fischer (1897) and Migula (1897). Of these systems, which alone are available in any practical scheme of classification, the two most important and most modern are those of Fischer and Migula. The extended investigations of the former on the number and distribution of cilia (see fig. 1) led him to propose a scheme of classification based on these and other morphological characters, and differing essentially from any preceding one. This scheme may be tabulated as follows: Fischer's scheme. I. ORDER-Haplobacterinae. Vegetative body unicellular; 1. Family-CoCCACEAE. Vegetative cells spheroidal. (b) Sub-family-HOMOCOCCACEAE. Division planes regular 2. Family BACILLACEAE. Vegetative cells cylindric (rodlets), ellipsoid or ovoid, and straight. Division planes always perpendicular to the long axis. (a) Sub-family-BACILLEAE. Sporogenous rodlets cylindric, not altered in shape:-Bacillus (Cohn), non-motile; Bactrinium (Fischer), motile, with one polar flagellum (monotrichous); Bactrillum (Fischer), motile, with a terminal tuft of cilia (lophotrichous); Bactridium (Fischer), motile, with cilia all over the surface (peritrichous). (b) Sub-family-CLOSTRIDIEAE. Sporogenous rodlets, spindle-shaped:-Clostridium (Prazm.), motile (peri trichous). (b) Filaments showing slow pendulous and creeping movements, and with no distinct sheath:Beggiatoa (Trev.), with sulphur particles. The principal objections to this system are the following:-(1) The extraordinary difficulty in obtaining satisfactory preparations showing the cilia, and the discovery that these motile organs are not formed on all substrata, or are only developed during short periods of activity while the organism is young and vigorous, render this character almost nugatory. For instance, B. megatherium and B. subtilis pass in a few hours after commencement of growth from a motile stage with peritrichous cilia, into one of filamentous growth preceded by casting of the cilia. (2) By far the majority of the described species (over 1000) fall into the three genera-Micrococcus (about 400). Bacillus (about 200) and Bactridium Labout 150), so that only a quarter or so of the forms are selected out by the other genera. (3) The monotrichous and lophotrichous condi tions are by no means constant even in the motile stage; thus Pseudomonas rosea (Mig.) may have 1, 2 or 3 cilia at either end, and would be distributed by Fischer's classification between Bac trinium and Bactrillum, according to which state was observed. In Migula's scheme the attempt is made to avoid some of these difficulties, but others are introduced by his otherwise clever devices for dealing with these puzzling little organisms. The question, What is an individual? has given rise to much difficulty, and around it many of the speculations regarding pleomorphism have centred without useful result. If a tree fall apart difficulty on a larger and more complex scale. The fact that every into its constituent cells periodically we should have the same bacterial cell in a species in most cases appears equally capable of performing all the physiological functions of the species has led which cannot be consistent in those cases where a simple or branched most authorities, however, to regard it as the individual-a view filamentous series exhibits differences between free apex and fixed base and so forth. It may be doubted whether the discussion is profitable, though it appears necessary in some cases-e.g. concerning pleomorphy-to adopt some definition of individual.” Myxobacteriaceae.-To the two divisions of bacteria, Haplobacterinae and Trichobacterinae, must now be added a third division, Myxobacterinae. One of the first members of this group, Chondromyces crocatus, was described as long ago as 1857 by Berkeley, but its nature was not understood and it was ascribed to the Hyphomycetes. In 1892, however, Thaxter rediscovered it and showed its bacterial nature, founding for it and some allied forms the group Myxobacteriaceae. Another form, which he described as Myxobacter, was shown later to be the same as Polyangium vitellinum described by Link in 1795, the exact nature of which had hitherto been in doubt. Thaxter's observations and conclusions were (c) Sub-family-PLECTRIDIEAE. Sporogenous rodlets, drum-called in question by some stick-shaped:-Plectridium (Fischer), motile (peri trichous). 3. Family SPIRILLACEAE. Vegetative cells, cylindric but curved more or less spirally. Divisions perpendicular to the long axis:-Vibrio (Müller-Löffler), commashaped, motile, monotrichous; Spirillum (Ehrenb.), more strongly curved in open spirals, motile, lophotrichous; Spirochaete (Ehrenb.), spirally coiled in numerous close turns, motile, but apparently owing to flexile movements, as no cilia are found. II. ORDER-Trichobacterinae. Vegetative body of branched or unbranched cell-filaments, the segments of which separate as swarm-cells (Gonidia). 1. Family-TRICHOBACTERIACEAE. Characters those of the Order. (a) Filaments rigid, non-motile, sheathed:-Crenothrix FIG. 12. A. Myxococcus digelatus,bright red fructification occurring on dung B. Polyangium primigenum, red fructification on dog's dung on C. Chondromyces apiculatus, orange fructification antelope's dung. botanists, but his later observa- D. Young fructification (A, B. after Quehl; C-E,after Thaxter.) From Strasburger's Lehrbuch der Botanik, by permission of Gustav Fischer. inimical to that of the fungi which are so common on this sub- | forms are termed by Fischer Metatrophic, because they require stratum. Function and life of bacteria. The discoveries that some species of nitrifying bacteria and perhaps pigmented forms are capable of carbon-assimilation, that others can fix free nitrogen and that a number of decompositions hitherto unsuspected are accomplished by Schizomycetes, have put the questions of nutrition and fermentation in quite new lights. Apart from numerous fermentation processes such as rotting, the soaking of skins for tanning, the preparation of indigo and of tobacco, hay, ensilage, &c., in all of which bacterial fermentations are concerned, attention may be especially directed to the following evidence of the supreme importance of Schizomycetes in agriculture and daily life. Indeed, nothing marks the attitude of modern bacteriology more clearly than the increasing attention which is being paid to useful fermentations. The vast majority of these organisms are not pathogenic, most are harmless and FIG. 13.-A series of phases of germination of the spore of B. ramosus sown at 8.30 (to the extreme left), showing how the growth can be measured. If we place the base of the filament in each case on a base line in the order of the successive times of observation recorded, and at distances apart proportional to the intervals of time (8.30, 10.0, 10.30, 11.40, and so on) and erect the straightened-out filaments, the proportional length of each of which is here given for each period, a line joining the tips of the filaments gives the curve of growth. (H. M. W.) many are indispensable aids in natural operations important to man. Fischer has proposed that the old division into saprophytes and parasites should be replaced by one which takes into account other peculiarities in the mode of nutrition of bacteria. The nitrifying, nitrogen-fixing, sulphur- and iron-bacteria he regards as monotrophic, i.e. as able to carry on one particular series of fermentations or decompositions only, and since they require no organic food materials, or at least are able to work up nitrogen or carbon from inorganic sources, he regards them as primitive forms in this respect and terms them Prototrophic. They may be looked upon as the nearest existing representatives of the primary forms of life which first obtained the power of working up non-living into living materials, and as playing a correspondingly important rôle in the evolution of life on our globe. The vast majority of bacteria, on the other hand, which are ordinarily termed saprophytes, are saprogenic, i.e. bring organic material to the putrefactive state or saprophilous, i.e. live best in such putrefying materials-or become zymogenic, i.e. their metabolic products may induce blood-poisoning or other toxic effects (facultative parasites) though they are not true parasites. These When we reflect that some hundreds of thousands of tons of urea are daily deposited, which ordinary plants are unable to assimilate until considerable changes have been underNitrogen gone, the question is of importance, What happens in bacteria. the meantime? In effect the urea first becomes carbonate of ammonia by a simple hydrolysis brought about by bacteria, more and more definitely known since Pasteur, van Tieghem and Cohn first described them. Lea and Miquel further proved that the hydrolysis is due to an enzyme-urase -separable with difficulty from the bacteria concerned. Many forms in rivers, soil, manure heaps, &c., are capable of bringing about this change to ammonium carbonate, and much of the loss of volatile ammonia on farms is preventible if the facts are apprehended. The excreta of urea alone thus afford to the soil enormous stores of nitrogen combined in a form which can be rendered available by bacteria, and there are in addition the supplies brought down in rain from the atmosphere, and those due to other living débris. The researches of later years have demonstrated that a still more inexhaustible supply of nitrogen is made available by the nitrogen-fixing bacteria of the soil. There are in all cultivated soils forms of bacteria which are capable of forcing the inert free nitrogen to combine with other elements into compounds assimilable by plants. This was long asserted as probable before Winogradsky showed that the conclusions of M. P. E. Berthelot, A. Laurent and others were right, and that Clostridium pasteurianum, for instance, if protected from access of free oxygen by an envelope of aerobic bacteria or fungi, and provided with the carbohydrates and minerals necessary for its growth, fixes nitrogen in proportion to the amount of sugar consumed. This interesting case of symbiosis is equalled by yet another case. The work of numerous observers has shown that the free nitrogen of the atmosphere is brought into combination in the soil in the nodules filled with bacteria on the roots of Leguminosae, and since these nodules are the morphological expression of a symbiosis between the higher plant and the bacteria, there is evidently here a case similar to the last. As regards the ammonium carbonate accumulating in the soil from the conversion of urea and other sources, we know from Winogradsky's researches that it undergoes oxidation in two stages owing to the activity of the so-called "nitrifying " bacteria (an unfortunate term inasmuch as "nitrification " refers merely to a particular phase of the cycle of changes undergone by nitrogen). It had long been known that under certain conditions large quantities of nitrate (saltpetre) are formed on exposed heaps of manure, &c., and it was supposed that direct oxidation of the ammonia, facilitated by the presence of porous bodies, brought this to pass. But research showed that this process of nitrification is dependent on temperature, aeration and moisture, as is life, and that while nitre-beds can infect one another, the process is stopped by sterilization. R. Warington, J. T. Schloessing, C. A. Müntz and others had proved that nitrification was promoted by some organism, when Winogradsky hit on the happy idea of isolating the organism by using gelatinous silica, and so avoiding the difficulties which Warington had shown to exist with the organism in presence of organic nitrogen, owing to its refusal to nitrify on gelatine or other nitrogenous media. Winogradsky's investigations resulted in the discovery that two kinds of bacteria are concerned in nitrification; one of these, which he terms the Nitroso-bacteria, is only capable of bringing about the oxidation of the ammonia to nitrous acid, and the astonishing result was obtained that this can be done, in the dark, by bacteria to which only pure | the globe generally. The ammonia may be oxidized to nitrites mineral salts-e.g. carbonates sulphates and chlorides of and nitrates, and then pass into the higher plants and be worked ammonium, sodium and magnesium-were added. In other up into proteids, and so be handed on to animals, eventually to words these bacteria can build up organic matter from purely be broken down by bacterial action again to ammonia; or the mineral sources by assimilating carbon from carbon dioxide in nitrates may be degraded to nitrites and even to free nitrogen or the dark and by obtaining their nitrogen from ammonia. The ammonia, which escapes. energy liberated during the oxidation of the nitrogen is regarded as splitting the carbon dioxide molecule, in green plants it is the energy of the solar rays which does this. Since the supply of free oxygen is dependent on the activity of green plants the process is indirectly dependent on energy derived from the sun, but it is none the less an astounding one and outside the limits of our previous generalizations. It has been suggested that urea is formed by polymerization of ammonium carbonate, and formic aldehyde is synthesized from CO, and OH2. The Nitro-bacteria are smaller, finer and quite different from the nitroso-bacteria, and are incapable of attacking and utilizing ammonium carbonate. When the latter have oxidized ammonia to nitrite, however, the former step in and oxidize it still further to nitric acid. It is probable that important consequences of these actions result from the presence of nitrifying bacteria in rotten stone, FIG. 14-Stages in the formation of a colony of a variety of Bacillus (Proteus) vulgaris (Hauser), observed in a hanging drop. At 11 A.M. a rodlet appeared (A); at 4 P.M. it had grown and divided and broken up into eight rodlets (B); C shows further develop ment at 8 P.M., D at 9.30 P.M.-all under a high power. At E, F, and G further stages are drawn, as seen under much lower power. (H. M. W.) decaying bricks, &c., where all the conditions are realized for preparing primitive soil, the breaking up of the mineral constituents being a secondary matter. That "soil" is thus prepared on barren rocks and mountain peaks may be concluded with some certainty. In addition to the bacterial actions which result in the oxidization of ammonia to nitrous acid, and of the latter to nitric acid, the reversal of such processes is also brought about by numerous bacteria in the soil, rivers, &c. Warington showed some time ago that many species are able to reduce nitrates to nitrites, and such reduction is now known to occur very widely in nature. The researches of Gayon and Dupetit, Giltay and Aberson and others have shown, moreover, that bacteria exist which carry such reduction still further, so that ammonia or even free nitrogen may escape. The importance of these results is evident in explaining an old puzzle in agriculture, viz. that it is a wasteful process to put nitrates and manure together on the land. Fresh manure abounds in de-nitrifying bacteria, and these organisms not only reduce the nitrates to nitrites, even setting free nitrogen and ammonia, but their effect extends to the undoing of the work of what nitrifying bacteria may be present also, with great loss. The combined nitrogen of dead organisms, broken down to ammonia by putrefactive bacteria, the ammonia of urea and the results of the fixation of free nitrogen, together with traces of nitrogen salts due to meteoric activity, are thus seen to undergo various vicissitudes in the soil, rivers and surface of osae. That the Leguminosae (a group of plants including peas, beans, vetches, lupins, &c.) play a special part in agriculture was known even to the ancients and was mentioned by Pliny Bacteria (Historia Naturalis, viii.). These plants will not only and grow on poor sandy soil without any addition of nitro- Legumingenous manure, but they actually enrich the soil on which they are grown. Hence leguminous plants are essential in all rotation of crops. By analysis it was shown by Schulz-Lupitz in 1881 that the way in which these plants enrich the soil is by increasing the nitrogen-content. Soil which had been cultivated for many years as pasture was sown with lupins for fifteen years in succession; an analysis then showed that the soil contained more than three times as much nitrogen as at the beginning of the experiment. The only possible source for this increase was the atmospheric nitrogen. It had been, however, an axiom with botanists that the green plants were unable to use the nitrogen of the air. The apparent contradiction was explained by the experiments of H. Hellriegel and Wilfarth in 1888. They showed that, when grown on sterilized sand with the addition of mineral salts, the Leguminosae were no more able to use the atmospheric nitrogen than other plants such as oats and barley. Both kinds of plants required the addition of nitrates to the soil. But if a little water in which arable soil had been shaken up was added to the sand, then the leguminous plants flourished in the absence of nitrates and showed an increase in nitrogenous material. They had clearly made use of the nitrogen of the air. When these plants were examined they had small swellings or nodules on their roots, while those grown in sterile sand without soil-extract had no nodules. Now these peculiar nodules are a normal characteristic of the roots of leguminous plants grown in ordinary soil. The experiments above mentioned made clear for the first time the nature and activity of these nodules. They are clearly the result of infection (if the soil extract was boiled before addition to the sand no nodules were produced), and their presence enabled the plant to absorb the free nitrogen of the air. The work of recent investigators has made clear the whole process. In ordinary arable soil there exist motile rod-like bacteria, Bacterium radicicola. These enter the root-hairs of leguminous plants, and passing down the hair in the form of a long, slimy (zoogloea) thread, penetrate the tissues of the root. As a result the tissues become hypertrophied, producing the well-known nodule. In the cells of the nodule the bacteria multiply and develop, drawing material from their host. Many of the bacteria exhibit curious involution a, b, FIG. 15.-Invasion of leguminous roots by bacteria. cell from the epidermis of root of Pea with "infection thread " (zoogloea) pushing its way through the cellwalls. (After Prazmowski.) forms ("bacteroids"), |