This Bacillus, which was discovered in sections of a Permian coprolite from Central France, has the form of cylindrical rods 12–14µ in length, and 1·3–1·5µ broad, rounded at each end. The rods occur either singly or occasionally, two or three individuals are joined end to end. Fig. 28 B represents a piece of one of Renault and Bertrand’s sections; the small rods are clearly seen lying in various directions in the homogeneous matrix of the coprolite. Each individual is said to be surrounded by an extremely minute empty space ·4µ in width, originally occupied by the Bacillus membrane, the central rod representing the mineralised cell-contents. In this example the petrifying substance was probably derived from the phosphate of calcium of bones which were attacked by Bacteria. I am indebted to Prof. Renault for an opportunity of examining specimens of this and other fossil Bacteria, and in this particular case there is undoubtedly strong evidence in favour of the author’s determination.
Renault has given the name Bacillus Tieghemi to certain minute rods 6–10µ, in length, and 2·2–3·8µ broad, often containing a dark coloured spherical spore-like body 2µ in diameter, which have been found in the tissues of a Coal-Measure plant.
The name Micrococcus Guignardi has been applied to more or less spherical bodies 2·2µ in diameter, also met with in silicified plants.
A portion of one of Renault’s figures is reproduced in Fig. 28 A. The faint and broken lines mark the position of the middle lamellae of parenchymatous cells from the pith of a Calamite. The tissue has been almost completely destroyed, but the more resistant middle lamellae have been partially preserved. The short and broad rods represent what Renault terms Bacillus Tieghemi; the small circle in the middle of some of these being referred to as a spore, and in one specimen shown in the figure, the second rod at right angles to the first is described as a small daughter-Bacillus formed by the germination of the central spore.
The isolated circles in the figure are referred to Micrococcus.
It is unnecessary to give an account of the numerous examples of Micrococci and Bacilli described by Renault from Devonian, Carboniferous, Permian and Jurassic rocks. We may, however, in a few words consider the general question of the existence and possible determination of fossil Bacteria.
In 1877 Prof. Van Tieghem[218] of Paris drew attention to the method of operation and plan of attack of Bacillus amylobacter as a destructive agent in the decay of plant débris in water. He was able to follow the gradual disorganisation of the tissues and the various steps in the ‘butyric fermentation’ effected by this Bacterium. Similarly the same author[219] was able to detect the action of an allied organism in some silicified tissues from the Carboniferous nodules of Grand-Croix, a well-known locality for petrified plants near Saint-Étienne. He recognised also the traces of the Bacillus itself in the partially destroyed plant tissues. The Palaeozoic Bacteria made use of some cellulose-dissolving ferment of which the action is clearly demonstrated in sections of silicified tissues. Many of the phenomena described by Renault and Bertrand as due to similar Bacterial action, afford additional evidence that the gradual disorganisation of vegetable tissues was effected in precisely the same manner as at the present day.
In some cases we have I believe trustworthy examples of the Bacteria themselves, both in coprolites and plant-tissues, but it is more than probable that some of the recorded examples are not of any scientific value. The examination of petrified tissues under the higher powers of a microscope often reveals the existence of numerous spherical particles and rod-like bodies which agree in shape with Micrococci or Bacilli. Minute crystals of mineral substances may occur in the siliceous or calcareous matrix of a petrified plant which simulate minute organic forms. Vogelsang[220] in his important work die Krystalliten has thrown considerable light on the ontogeny of crystals, and the minute globulites and other forms of incipient crystallisation might well be mistaken for Bacterial cells. Granting, however, that we have satisfactory evidence, both direct and indirect, that some forms of Bacteria lived in the decaying tissues of Palaeozoic plants, and in the intestines of reptiles and other animals, we cannot safely proceed to specific diagnoses and determinations[221].
Renault has pointed out that fossil Bacteria may often be more readily detected than living forms owing to the presence of a brown ulmic substance which results from the carbonisation of the protoplasm. He is forced to admit, however, that such diagnostic characters as are obtained by Bacteriologists by means of cultures cannot be utilised when we are dealing with fossil examples! We are told that “Partout où nous avons cherché des Bacteriaceés, nous en avons rencontré.”[222] This indeed is the danger; an extended examination of fossil sections under an immersion-lens must almost inevitably lead to the discovery of minute bodies of a more or less spherical form which might be Micrococci. To measure, and name such bodies as definite species of Micrococci is, I believe, but wasted energy and an attempt to compass the impossible.
Specialists tell us that the accurate determination of species of recent Bacteria is practically hopeless: may we not reasonably conclude that the attempt to specifically diagnose fossil forms is absolutely hopeless? “The imagination of man is naturally sublime, delighted with whatever is remote and extraordinary—”, but it is to be deplored if the fascination of fossil bacteriology is allowed to warp sound scientific sense.
IV. ALGAE.
The presence of chlorophyll is one common characteristic of the numerous plants included in the Algae. The generally adopted classification rests in part on an artificial distinction, namely the prevailing colour of the plant.
It must be definitely admitted, at the outset, that palaeobotany has so far afforded extremely little trustworthy information as to the past history of algae. Were we to measure the importance of the geological history of these plants by the number of recorded fossil species, we should arrive at a totally wrong and misleading estimate. By far the greater number of the supposed fossil algae have no claim to be regarded as authentic records of this class of Thallophytes. It has been justly said that palaeontologists have been in the habit of referring to algae such impressions or markings on rocks as cannot well be included in any other group. “A fossil alga,” has often been the dernier ressort of the doubtful student.
Before discussing our knowledge, or rather lack of knowledge, of fossil algae at greater length, it will be well to briefly consider the manner of occurrence and botanical nature of existing forms. In the sea and in fresh water, as well as in damp places and even in situations subject to periods of drought, algae occur in abundance in all parts of the world. We find them attaining full development and reproducing themselves at a temperature of −1° C. in the Arctic Seas, and again living in enormous numbers in the waters of thermal springs. Around the coast-line of land areas, and on the floor of shallow seas algae exhibit a remarkable wealth of form and luxuriance of growth. As regards habit and structure, there is every gradation from algae in which the whole individual consists of a thin-walled unseptate vesicle, to those in which the thallus attains a length unsurpassed by any other plant, and of which the anatomical features clearly express a well-marked physiological division of labour such as occurs in the highest plants.
The large and leathery seaweeds which flourish in the extreme northern and southern seas are plants which it is reasonable to suppose might well have left traces of their existence in ancient sediments. Sir Joseph Hooker, in his account of the Antarctic flora[223], investigated during Sir James Ross’s voyage in H.M. ships Erebus and Terror, has given an exceedingly interesting description of the gigantic brown seaweeds of southern latitudes. The trunks are described as usually 5–10 feet long, and as thick as a human thigh, dividing towards the summit into numerous pendulous branches which are again broken up into sprays with linear ‘leaves.’ Hooker records how a captain of a brig employed his crew for two bitterly cold days in collecting Lessonia stems which had been washed up on the beach, thinking they were trunks of trees fit for burning. On our own coasts we are familiar with the common Laminaria, the large brown seaweed with long and strap-shaped or digitate fronds which grows on the rocks below low-tide level. The frond passes downwards into a thick and tough stipe firmly attached to the ground by special holdfasts. A transverse section of the stalk of a fairly old plant presents an appearance not unlike that of a section of a woody plant. In the centre there is a well-defined axial region or pith consisting of thick walled, long and narrow tubes pursuing a generally vertical though irregular course, and embedded in a matrix of gelatinous substance derived from the mucilaginous degeneration of the outer portions of the cell-walls. The greater part of such a section consists, however, of regularly disposed rows of cells which have obviously been formed by the activity of a zone of dividing or meristematic elements. The occurrence of distinct concentric rings in this secondary tissue clearly points to some periodicity of growth which is expressed by the alternation of narrow and broader cells. In the Antarctic genus Lessonia, the stem reaches a girth equal to that of a man’s thigh, and in structure it agrees closely with the smaller stem of Laminaria. In these large algal stems, the cells are not lignified as in woody plants, and in longitudinal section they have for the most part the form of somewhat elongated parenchyma, differing widely in appearance from the tracheids or vessels of woody plants. At the periphery of the Laminaria stem, represented in fig. 29, there occur numerous and comparatively large mucilage ducts.
In certain algae of different families the thallus is encrusted with carbonate of lime, and is thus rendered much more resistant. The Diatoms, on the other hand, possess still more durable siliceous tests which are particularly well adapted to resist the solvent action of water and other agents of destruction. It is these calcareous and siliceous forms which supply the greater part of the trustworthy data furnished by fossil algae.
It remains to consider some of the causes to which we may attribute the scarcity of fossil algae, and the possible sources of error which beset any attempt to describe or assign names to impressions and casts simulating algal forms.
In the first place, the delicate nature of algal cells is a serious obstacle to fossilisation. Even in plants in which the woody stems have been preserved by a siliceous or calcareous solution, we frequently find the more delicate cells represented by a mass of crystalline matter without any trace of the cell-walls being preserved. In such plants as algae, where the cell-walls are not lignified, but consist of cellulose or some special form of cellulose, which readily breaks down into a mucilaginous product, the tissues have but a small chance of withstanding the wear and tear of fossilisation.
The danger of relying on external form as a means of recognition is especially patent in the case of those numerous markings or impressions frequently met with on rocks, and which resemble in outline the thallus of recent algae. Among animals, such as certain Polyzoa, the flat branching body of various algae is closely simulated, and in other plants, such as the frondose liverworts, the same thalloid and branched form of body is again met with. Some of the much dissected Aphlebia leaves of ferns (e.g. Rhacophyllum species) bear a striking resemblance to fossil algae; and numerous other examples might be quoted. In palaeobotanical literature we find a host of names, such as Chondrites, Fucoides[224], Caulerpites and others applied to indefinite and indistinct surface markings which happen to resemble in shape certain of the better known genera of recent seaweeds.
The close parallelism in outward form displayed by different genera and families of algae is in itself sufficient argument against the use of recent generic names for fossils of which the algal nature is often more than doubtful. Were external form to be accepted as a trustworthy guide, in the absence of internal structure and reproductive organs, such a genus as Caulerpa[225] would afford material for numerous generic designations. A comparison of the different species of this Siphoneous green alga brings out very clearly the exceedingly protean nature of this interesting genus, and serves as one instance among many of the small taxonomic value which can be attached to external configuration. Caulerpa pusilla Mart. and Her., C. taxifolia (Vahl.), C. plumaris Forsk., C. abies-marina J. Ag., C. ericifolia (Turn.), C. hypnoides (R. Br.), C. cactoides (Turn.), C. scalpelli formis (R. Br.), and others clearly illustrate the almost endless variety of form exhibited by the species of a single genus of algae. We constantly find in the several classes of plants a repetition of the same form either in the whole or in the separate members of the vegetative body, and but a slight acquaintance with plant types should lead us to use the test of external resemblance with the greatest possible caution. To emphasize this danger may seem merely the needless reiteration of a self-evident fact, but there is, perhaps, no source of error which has been more responsible for the creation of numerous worthless species among fossil plants.
There is, however, another category of impressions and casts of common occurrence in sedimentary rocks which requires a brief notice. Very many of the fossil algae described in text-books and palaeobotanical memoirs have been shown to be of animal origin, and to be merely the casts of tracks and burrows. A few examples will best serve to illustrate the identity of many of the fossils referred to algae with animal trails and with impressions produced by inorganic agency.
Dr Nathorst of Stockholm has done more than any other worker to demonstrate the true nature of many of the species of Chondrites, Cruziana, Spirophyton, Eophyton, and numerous other genera. In 1867 there were discovered in certain Cambrian beds of Vestrogothia, long convex and furrowed structures in sandstone rocks which were described as the remains of some comparatively highly organised plant, and described under the generic name Eophyton[226]. By many authors these fossils have been referred to algae, but Nathorst has shown that the frond of an alga trailed along the surface of soft plaster of Paris produces a finely furrowed groove (fig. 30, 2) which would afford a cast similar to that of Eophyton. The same author has also adduced good reasons for believing that the Eophytons of Cambrian rocks may represent the trails made by the tentacles of a Medusa having a habit similar to that of Polydonia frondosa Ag. Impressions of Medusae have been described by Nathorst from the beds in which Eophyton occurs; and the specimens in the Stockholm Museum afford a remarkable instance of the rare preservation of a soft-bodied organism[227]. By allowing various animals to crawl over a soft-prepared surface it is possible to obtain moulds and casts which suggest in a striking manner the branched thallus of an alga. The tracks of the Polychaet, Goniada maculata Örstd.[228], one of the Glyceridae, are always branched and very algal-like in form (fig. 30, 3). Many of the so-called fossil algae are undoubtedly mere tracks or trails of this type. In the fossil-plant gallery of the British Museum there are several specimens of small branched casts, clearly marked as whitish fossils on a dark grey rock of Lower Eocene age from Bognor; these were described by Mantell and Brongniart[229] as an alga, but there is little doubt of their being of the same category as the track shown in fig. 30, 3.
The well-known half-relief casts met with in Cambrian, Silurian and Carboniferous rocks, and known as Cruziana or Bilobites, are probably casts of the tracks of Crustaceans. The impression left by a King-Crab (Limulus) as it walks over a soft surface affords an example of this form of cast. It has been suggested that some of the Bilobites may be the casts of an organism like Balanoglossus[230], a worm-like animal supposed by some to have vertebrate affinities. The resemblance between some of the lower Palaeozoic Bilobites and the external features of a Balanoglossus is very striking, and such a comparison is worth considering in view of the fact that soft-bodied animals have occasionally left distinct impressions on ancient sediments.
The literature on the subject of fossil algae versus inorganic and animal markings is too extensive and too wearisome to consider in a short summary; the student will find a sufficient amount of such controversial writing—with references to more—in the works quoted below[231].
In the Stockholm Museum of Palaeobotany there is an exceedingly interesting collection of plaster casts obtained by Dr Nathorst in his experiments on the manufacture of fossil ‘algae,’ which afford convincing proof of the value and correctness of his general conclusions.
The pressure of the hand on a soft moist surface produces a raised pattern like a branched and delicate thallus. The well-known Oldhamia antiqua Forbes and Oldhamia radiata Forbes[232], from the Cambrian rocks of Ireland may, in part at least, owe their origin to mechanical causes, and we have no sufficient evidence for including them among the select class of true fossil algae. Sollas[233] has shown that the structure known as Oldhamia radiata is not merely superficial but that it extends across the cleavage-planes. Oldhamia is recorded from Lower Palaeozoic rocks in the Pyrenees[234] by Barrois, who agrees with Salter, Göppert and others in classing the fossil among the algae. The photograph accompanying Barrois’ description does not, however, add further evidence in favour of accepting Oldhamia as a genus of fossil algae.
The burrows made by Gryllotalpa vulgaris Latr., the Mole-cricket, have been shown by Zeiller to bear a close resemblance to a branch of a conifer in half-relief (fig. 30, 4), or to such a supposed algal genus as Phymatoderma[235].
In fig. 30, 1, we have what might well be described as a fossil alga. This is merely a cast of a miniature river-system such as one frequently sees cut out by the small rills of water flowing over a gently-sloping sandy beach. A cast figured and described by Newberry as an alga, Dendrophycus triassicus[236], from the Trias of the Connecticut Valley, is practically identical with the rill-marks shown in fig. 30, 1. The cracks produced in drying and contracting sediment may form moulds in which casts are subsequently produced by the deposition of an overlying layer of sand, and such casts have been erroneously referred to algal impressions[237]. Dawson[238] has figured two good examples of Carboniferous rill-marks from Nova Scotia in his paper on Palaeozoic burrows and tracks of invertebrate animals.
The specimen represented in fig. 31 affords an example of a fairly well-known fossil from the Wenlock limestone, originally described by Salter as Chondrites verisimilis Salt, from Dudley[239]. He regarded it as an alga, and the graphitic impression agrees closely in form with the thallus of some small seaweeds. A closer examination of the fossil reveals a curious and characteristic irregular wrinkling on the graphite surface, which suggests an organism of more chitinous and firmer material than that of an alga.
A similar and probably an identical fossil is described and figured by Lapworth[240] in an appendix to a paper by Walter Keeping on the geology of Central Wales, under the name of Odontocaulis Keepingi Lap. and regarded as a dendroid graptolite. In any case we have no satisfactory grounds for including these fossils in the plant-kingdom.
How then are we to recognise the traces of ancient algae? There is no golden rule, and we must admit the difficulty of separating real fossil algae from markings made by animal or mechanical agency. The presence of a carbonaceous film is occasionally a help, but its occurrence is no sure test of plant origin, nor is its absence a fatal objection to an organic origin. While being fully alive to the small value of external resemblance, and to the numerous agents which have been shown to be capable of producing appearances indistinguishable from plant impressions, we must not go too far in a purely negative direction.
An important contribution to the subject of fossil algae has lately appeared by Prof. Rothpletz[241]. He deals more particularly with the much discussed Flysch[242] Fucoids of Tertiary age, and while refusing to accept certain examples as fossil algae, he brings forward weighty arguments in favour of including several other forms among the algae. He is of opinion that most of the main divisions of the algae are represented among the Flysch Fucoids, but considers that the Phaeophyceae are the most numerous.
Rothpletz’s work is chiefly interesting as illustrating the application of microscopic examination and chemical analysis to the determination of fossil algae. Although he makes out a good case in favour of restoring many of the Tertiary fossils to the plant kingdom, the material at his disposal does not admit of satisfactory botanical diagnosis.
No doubt some of the fossils from the Silurian and Cambrian rocks are true algae, and Nathorst has pointed out that such a species as Hall’s Sphenothallus angustifolius[243] may well be an alga. Additional examples might be quoted from Bornemann and other writers, but in view of the attempts which are sometimes made to trace the development of more recent plants to more than doubtful Lower Palaeozoic Algae, one must agree with Nathorst’s opinion,—“Je crois que l’on rend un bien mauvais service à la théorie de l’évolution, en essayant de baser l’arbre généalogique des algues fossiles sur des corps aussi douteux que les Bilobites, Crossochorda, Eophyton, etc.[244]”
There are many carbonaceous impressions on rocks of different ages which it is reasonable to refer to algal origin, and although such are of little or no botanical value, it may be a convenience to refer to them under a definite term. The comprehensive generic name Algites[245] has been suggested as a convenient designation for impressions or casts which are probably those of algae.
Some of the fossils described by Mr Kidston from British Carboniferous rocks as probably algae present an undoubted algal appearance, and might be placed in the genus Algites; but in some cases—e.g. Chondrites plumosa[246] Kidst. from the Calciferous Sandstone of Eskdale, one feels much more doubtful; in this particular instance the impressions suggest the fine roots of a water-plant.
The statement is occasionally made that the numerous fossil algae and the absence of higher plants in the older strata justify the description of the oldest rocks as belonging to the ‘age of algae.’ Such an assertion rests on an unsound basis, and is rather the expression of what might be expected than what has been proved to be the case. The oldest plants with which we are at all closely acquainted are of such a type as to forcibly suggest that in the lowest fossiliferous rocks we are still very far from the sediments of that age which witnessed the dawn of plant life.
Many of the obscure markings on rock surfaces which have been referred to existing genera of algae or described as new genera, are much too doubtful to be included even under such a comprehensive name as Algites. Space does not admit of further reference to determinations of this type which abound in palaeontological literature.
It would be very difficult to produce satisfactory evidence for the algal nature of many of the supposed fossil algae from Cambrian rocks[247]; there has been a special tendency to recognise algal remains in the oldest fossiliferous strata, due in part no doubt to the fallacy that in that period nothing higher than Thallophytes is likely to have existed. The so-called Phycodes referred to by Credner[248] as characteristic of the Cambrian rocks of the Fichtelgebirge (“Phycoden-Schiefer”) is probably of inorganic origin, and comparable to the genus Vexillum of Saporta[249] and other writers, which Solms-Laubach has described as being formed every day in the soft mud of our ponds where local currents are checked by branches and other obstacles[250]. There are several good specimens of Phycodes in the Bergakademie of Berlin and in the Leipzig Museum which, I believe, clearly demonstrate the absence of all satisfactory evidence of an algal origin.
We may next pass to a short description of a few representative types of algae, which may reasonably be classed under definite families, and accepted as evidence possessing some botanical value.
A. DIATOMACEAE (Bacillariaceae).
This family occupies a somewhat isolated position among the algae, and is best considered as a distinct subdivision rather than as a family of the Phaeophyceae or Brown algae, with which it possesses as a common characteristic a brown-colouring matter.
Single-celled plants consisting of a simple protoplasmic body containing a nucleus and brown colouring matter (diatomin) associated with the chlorophyll. The cell-wall is in the form of two halves, known as valves, which fit into one another like the two portions of a pill-box. The cell-wall contains a large amount of silica, and the siliceous cases of the diatoms are commonly spoken of as the valves of the individual, or the frustules. Diatoms exhibit a characteristic creeping movement, and are reproduced by division, also by the development of spores in various forms[251].
The recent members of the family have an exceedingly wide distribution, occurring both in freshwater and in the sea. Owing to the lightness of the frustules, they are frequently carried along in the air, and atmospheric dust falling on ships at sea has been found to contain large numbers of diatoms[252]. The siliceous valves are abundant in guano deposits, and they have been found also in association with volcanic material. Diatomaceous deposits are now being formed in the Yellowstone Park district; “they cover many square miles in the vicinity of active or extinct hot spring vents of the park, and are often three feet, four feet, and sometimes five to six feet thick[253].” The gradual accumulation of the siliceous tests on the floor of a fresh-water lake results in the formation of a sediment consisting in part of pure silica. Such deposits, often spoken of as kieselguhr or diatomite, and used as a polishing material, occur in many parts of Britain, marking the sites of dried-up pools or lakes. At the northern end of the island of Skye there occurs an unusually pure deposit of diatomite overlain by peat and turf, and extending over an area of fifty-eight square miles. Many of the individuals in this deposit were in all probability carried into the lake by running water, while others lived in the lake and after death their tests contributed to the siliceous deposit[254]. The late Dr Ehrenberg published numerous papers on diatomaceous deposits in different parts of the world, and in his great work, Zur Mikrogeologie[255], he gave numerous and beautifully executed illustrations of such siliceous accumulations. In many of the samples he figures one sees fragments of plant tissues, spores of conifers and ferns, associated with the diatom tests. The occurrence of the pollen grains of coniferous trees in lacustrine and marine deposits is not surprising in view of their abundance in Lake Constance and other lakes. It is stated that the pollen of conifers in the Norwegian fiords plays an important part in the nourishment of the Rhizopod Saccamina[256].
In the waters of the ocean diatoms are of frequent occurrence, and very widely distributed. Sir Joseph Hooker records the existence of masses of diatomaceous ooze over a wide area in Antarctic regions[257]. Along the shores of the Victoria Barrier, a perpendicular wall of ice, between one and two hundred feet above sea-level, the soundings were found to be invariably charged with diatom remains, and from the base of the ice-wall there appeared to be in process of formation a bank of these tests stretching north for a distance of 200 miles. The more extended researches conducted during the cruise of the Challenger have clearly proved the enormous accumulations of diatoms now being formed on the ocean-bed[258]. South of latitude 45° S. there is now being built up a vast deposit which may be eventually upraised as a fairly pure siliceous rock. From extreme northern latitudes Nansen has recently recorded the occurrence of these lowly organised plants. He writes,—“I found a whole world of diatoms and other microscopical organisms, both vegetable and animal, living in the fresh-water pools on the Polar drift-ice, and constantly travelling from Siberia to the east coast of Greenland[259].” In warmer latitudes diatoms abound in the surface waters, but there they are associated with numerous other forms of the Plankton vegetation. The waters of the Amazon carry with them into the sea large numbers of fresh-water forms, which are floated out to sea and finally added to the rock-building material which is constantly accumulating on the ocean floor[260]. No definite results have so far been obtained as to the geographical and bathymetrical distribution of marine diatoms.
The enormous number of recent species precludes any attempt to give a description of the better-known forms. It is more important for us to realize how common and widely distributed are the living genera. The hard and almost indestructible valves have been frequently found in a fossil condition, often forming thick and extensive masses of siliceous rock. From diatom-beds now forming in lakes and on the ocean-bed we pass to deposits such as those in Skye and elsewhere, which mark the site of recently dried-up sheets of water, and so to older rocks of Tertiary age formed under similar conditions. Among the many examples of diatomaceous deposits of Tertiary and Cretaceous age mention should be made of those of Berlin, Königsberg, Bilin in Bohemia, and Richmond in Virginia. The diatoms in the beds of Berlin are regarded as fresh-water, and those of Richmond as marine. It has been pointed out by Pfitzer that it is a comparatively easy matter to distinguish between fresh-water and marine forms of diatoms. The diatomaceous rocks of Bilin are known as polishing slates; they attain a thickness of 50 feet. In these, as in many other cases, the deposit has become cemented together as a hard flinty or glassy rock, in which the cementing material was formed by the solution of some of the diatom tests[261]. In many cases in which calcareous and siliceous rocks reveal no direct evidence of organic origin it is probable that they were originally formed by the accumulation of plants of which the structure has been completely obliterated by secondary causes. The genus Gallionella plays an important part in the composition of the Bilin beds. Occasionally impressions of leaves and other organic remains are found associated with the diatoms in the siliceous rocks. In the British Museum (Botanical department) a large block of white powdery rock is exhibited as an example of a diatomaceous deposit of Tertiary age from Australia. It is described as being largely made up of the tests of fresh-water diatoms, such as Navicula, Gomphonema, Cymbella, Synedra, and others.
The abundance of Diatoms in Cretaceous rocks of the Paris basin has recently been recorded by Cayeux[262]; it would seem that these algae had already assumed an important rôle as rock-builders in pre-Tertiary times. Cayeux points out that the silica of these Cretaceous diatomaceous frustules has often been replaced by carbonate of calcium.
In addition to the occurrence of Diatoms in the various diatomaceous deposits, their siliceous tests may occasionally be recognised in argillaceous or other sediments. Shrubsole and Kitton[263] have described several species of Diatoms from the London Clay of Lower Eocene age. In many localities in the London basin the clay obtained from well-sinkings presented the appearance of being dusted with sulphur-like particles of a dark bronze or golden colour which glistened in the sunlight. These yellow bodies have been found to be diatomaceous frustules in which the silica has been replaced by iron pyrites. The genus Coscinodiscus is one of the commonest forms recorded from the London Clay[264].
Without further considering individual examples of diatomaceous rocks we may briefly notice the general facts of the geological history of the family. As Ehrenberg pointed out several years ago, the Tertiary and Cretaceous species of diatoms show a very marked resemblance to living forms. In many cases the species are identical, and the fossil deposits as a whole seem to differ in no special respect from those now being built up.
With the exception of two species of Liassic Diatoms, no trustworthy examples of the Diatomaceae have been found below the Cretaceous series. The oldest known Diatoms were discovered by Rothpletz[265] among the fibres of an Upper Lias sponge from Boll in Württemberg. They occur as small thimble-shaped siliceous tests with coccoliths and foraminifera in the horny skeleton of Phymatoderma, a genus formerly regarded as an alga. Rothpletz describes two species which he includes in the genus Pyxidicula, P. bollensis and P. liasica. This generic name of Ehrenberg is used by Schütt[266] as a subgenus of Stephanopyxis.
Seeing how great a resemblance there is between the recent and Cretaceous species, and how many examples there are of Tertiary diatom deposits, it is not a little surprising that the past history of these plants has not been traced to earlier periods. In 1876 Castracane[267], an Italian diatomist, gave an account of certain species of diatoms said to have been found in a block of coal from Liverpool obtained from the English Coal-Measures. The species were found to be identical with recent forms. It is generally agreed that these specimens cannot have been from the coal itself, but that they must have been living forms which had come to be associated with the coal. The late Prof. Williamson spent many years examining thin sections and other preparations of coal from various parts of the world, but he never found a trace of any fossil diatom. There is no apparent reason why diatoms should not be found in Pre-Cretaceous rocks, and the microscopic investigation of old sediments may well lead to their discovery. Prof. Bertrand of Lille, who has devoted himself for some time past to a detailed microscopical examination of coal, informs me that he has so far failed to discover any trace of Palaeozoic diatomaceous tests.