which seems to show that the cavities must have existed before either dolomite or serpentine were introduced into the substance. Dr Carpenter contends that the disposition of these passages in his decalcified specimens is very regular, and quite unlike any mineral infiltration with which he is acquainted.

and serpentine. But perhaps their most intelligible | system" is not filled with serpentine but with dolomite, sections are those which they present in Bavaria and Bohemia between the valley of the Danube and the headwaters of the Elbe. They are there divided into two well-marked groups-(a) red gneiss, covered by (b) grey gneiss. According to Gümbel the former (called by him the Bojan gneiss) may be traced as a distinct formation associated with granite, but with very few other kinds of crystalline or schistose rocks, while the latter (termed the Hercynian gneiss) consists of gneiss with abundant interstratification of many other schistose rocks, graphitic limestone, and serpentine. The Hercynian gneiss is overlaid by mica-schist, above which comes a vast mass of argillaceous schists and shales. Gümbel some years ago found in the marbles associated with the younger gneiss what he considered to be an organism of the same genus as the Bozoon of Canada, to which reference will immediately be made. He named it Eozoon Bavaricum. More recently a similar substance was obtained in the Archæan cries of Bohemia, and named by Fritsch Eozoon Bohemicum.

AMERICA. In North America Archæan rocks cover a large part of the continent from the Arctic Circle south wards to the great lakes. They appear likewise, as in Europe, along the central parts of prominent mountain chains, as in the Rocky Mountain range and that of the Appalachians. They have been carefully studied in Canada, where the late Sir W. E. Logan, Director of the Geological Survey of the Dominion, estimated their depth at about 30,000 feet, but neither their top nor their base can there be found. He named them the Laurentian system from their abundant development along the shores of the St Lawrence. They have been divided into two series-(1) a lower formation more than 20,000 feet thick, consisting chiefly of granitic, orthoclase gneiss, with bands of quartzrock, schists, iron-ore, and limestone; and (2) an upper formation fully 10,000 feet thick, composed also, for the most part, of gneiss, but marked by the occurrence of bands of Labrador felspar, as well as schist, iron-ore, and limestone. The upper division has been stated to lie unconformably on the lower. Mr Selwyn, however, has recently pointed out that this is almost certainly not the case, but that the limestone-bearing series rests conformably upon a massive granitoid gneiss, to which he would restrict the term Laurentian, classing the limestones in the next or Huronian system (Nat. Hist. Soc. Montreal, Feb. 1879).

In one of the Laurentian limestones of Canada, specimens have been found of a remarkable mixture of calcite and serpentine. These minerals are arranged in alternate layers, the calcite forming the main framework of the substance with the serpentine (sometimes loganite, pyroxene, &c.) disposed in thin, wavy, inconstant layers, as if filling up flattened cavities in the calcareous mass. So different from any ordinary mineral segregation with which he was acquainted did this arrangement appear to Logan, that he was led to regard the substance as probably of organic origin. This opinion was adopted, and the structure of the supposed fossil was worked out in elaborate detail by Dr Dawson of Montreal, who pronounced the organism to be the remains of a massive foraminifer which he called Eozoon, and which he believed must have grown in large thick sheets over the sea-bottom. This opinion was confirmed by Dr W. B. Carpenter, who from a large suite of additional and better preserved specimens, described a system of internal canals having the characters of those in true foraminiferal structures. (See FORAMINIEERA.) Other observers, notably Profesors King and Rowney of Galway and Möbius of Kiel, have opposed the organic nature of Eozoon, and have endeavoured to show that the supposed canals and passages are merely infiltration veinings of serpentine in the calcite. In some cases, however, the "canal

The opinion of the organic nature of Eozoon has been supposed to receive support from the large quantity of graphite found throughout the Archæan rocks of Canada and the northern parts of the United States. This mineral occurs partly in veins, but chiefly disseminated in scales and laminæ in the limestones and as independent layers. Dr Dawson estimates the aggregate amount of it in one band of limestone in the Ottawa district as not less than from 20 to 30 feet, and he thinks it is hardly an exaggeration to say that there is as much carbon in the Laurentian as in equivalent areas of the Carboniferous system. He compares some of the pure bands of graphite to beds of coal, and maintains that no other source for their origin can be imagined than the deoxidation of carbonic acid by living plants. In the largest of three beds of graphite at St John he has found what he considers may be fibrous structure indicative of the existence of land-plants.

Still further evidence in favour of organized existence during Archean time in the North American area has been adduced from the remarkably thick and abundant masses of iron ore associated with the Laurentian rocks of Canada and the United States. Dr Sterry Hunt has called attention to these ores as proving the precipitation of iron by decomposing vegetation during the Laurentian period on a more gigantic scale than at any subsequent geological epoch.1 Some of the beds of magnetic iron ranged up to 200 feet in thickness. Large masses also of hæmatite and titaniferous iron, as well as of iron sulphides, occur in the Canadian Archæan series. These great bands of iron ore run southward, and form an important feature in the economic geology of the Northern States of the Union.

Above the Laurentian rocks in the region of Lake Huron lies a vast mass of slates, conglomerates, limestones, and quartz-rocks, attaining a depth of from 10,000 to 20,000 feet. They are termed Huronian. No fossils have yet been found in them; but they must be much younger than the Laurentian rocks, on which they rest unconformably, and from which they have been in part at least derived.

II. PALEOZOIC.

Under the general term of Primary or Paleozoic are now included all the older sedimentary formations containing organic remains, up to the top of what is termed the Permian system. These rocks consist mainly of sandy and muddy sediment with occasional intercalated zones of limestone. They everywhere bear witness to comparatively shallow water and the proximity of land. Their frequent alternations of sandstone, shale, conglomerate, and other detrital materials, their abundant, rippled, and sun-cracked surfaces marked often with burrows and trails of worms, as well as the prevalent character of their organic remains, show that they must have been deposited in areas of slow subsidence, bordering continental or insular masses of land. As regards the organisms of which they have preserved the casts, the Palæozoic rocks, as far as the present evidence goes, may be grouped into two divisions-an older and a newer the former distinguished more especially by the abundance of its graptolitic, trilobitic, and brachiopodous fauna, and by the absence of vertebrate remains; the latter by the number and variety of its fishes and amphibians, the

Geology of Canada, 1863, p. 573.

and afterwards Hay Cunningham, pointed out that they underlie parts of the schistose rocks of the northern High lands. The discovery by Mr C. W. Peach of Lower Siluriau shells in the overlying limestones showed that the massive red sandstones of western Ross and Sutherland could not Do paralleled with those of the eastern tracts of those counties, but must be of older date than part of the Llandeilo rocks of the Lower Silurian period. Sir R. Murchison classed them as Cambrian-an identification which has much support in the lithological resemblance between these rocks of the north-west Highlands and much of the Lower Cambrian system of Wales.

suite (78 species) of organic remains. The trilobites include the genera Agnostus, Anapolenus, Conocoryphe, Dikelocephalus, Erinnys, Olenus, and Paradoxides. The earliest phyllopods (Hymenocuris) and heteropods (Bellerophon) occur in these beds. The brachiopods include species of Lingulello (L. Darisii), Discina, Obolella, and Orthis. The pteropods are represented by three species of Theca. Several annelides (Cruziana) and polyzoa (Fencstella) likewise occur. According to a careful census by Mr Etheridge, the Lingula flags may be grouped into three zones, each characterized by a peculiar assemblage of organic remains. The fower division contains 37 species, of which 9 are peculiar to it. The middle zone has yielded 5 species, 2 of which (Conocoryphe bucephala and Lingulella Davisii) pass down into the lower division, 1 (Kutorgia cingulata) into the upper, and 2 (Lingula squamosa and Beilerophon Cambrensis) are peculiar. The upper zone has yielded 40 species. Of these 9 pass up into the Tremadoc beds, while 2 In the south-east of Ireland masses of purplish, red, and (Lingulclla lepis and L. Davisii) continue on into the Arenig group. green sliales, slates, grits, quartz-rocks, and schists occupy 4. Tremadoc Slates.-This name was given by Sedgwick to a a considerable area and attain a depth of 14,000 feet withgroup of dark grey slates, about 1000 feet thick, found near Tremadoc in Carnarvonshire, and traceable thence to Dolgelly. out revealing their base, while their top is covered by unTheir importance as a geological formation was not recognized conformable formations (Lower Silurian and Lower Carboniuntil the discovery of a remarkably abundant and varied fauna in ferous). They have yielded Oldhamia, described originally them. They contain the earliest crinoids, star-fishes, lamellias a sertularian zoophyte, but now regarded by many branchs, and cephalopods yet found. The trilobites embrace 14 genera, among which, besides some, as Agnostus, Conocoryphe, and paleontologists as an alga; also numerous burrows and Olenus, found in the Lingula flags, we meet for the first time with trails of annelides (Histioderma Hibernicum, Arenicolites Angelina, Asaphus, Cheirurus, Neseuretus, Niobe, Ogygia, Psilo- didymus, A. sparsus, Haughtonia pocila). No Upper cephalus, &c. The same genera, and in some cases species, of Cambrian forms have been met with in these Irish rocks, brachiopods appear which occur in the Lingula flags, Orthis Carausii and Lingulella Davisii being common forms. Mr Hicks which are therefore placed with the Lower Cambrian, the has described 12 species of lamellibranchs from the Tremadoc beds unconformability at their top. being regarded as equivalent of Ramsay Island and St Davids, belonging to the genera Cteno- to the interval required for the deposition of the intervening donta, Palaarca, Glyptarca, Davidia, and Modiolopsis. The cepha- formations up to the time of the Llandeilo rocks, as in the lopods are represented by Orthoceras sericeum and Cyrtoceras north-west of Scotland. Some portions of the Irish Campræcox; the pteropods by Theca Davidii, T. operculata, and Conularia Homfrayi; the echinoderms by a beautiful star-fish (Palas- brian series have been intensely metamorphosed. Thus ou terina ramsegensis) and by a crinoid (Dendrocrinus Cambrensis).1 the Howth coast they appear as schists and quartz-rocks; Careful analysis of the fossils yielded by the Tremadoc in Wexford they pass into gneiss and granite. In West beds suggests a division of this formation into two zones. Galway Mr Kinahan has described a vast mass of schists, According to a census by Mr Etheridge, the Lower Tremadoc quartz-rocks, and limestones (8000 feet and upwards) passrocks have yielded in all 56 species, of which 9 pass down ing up into schistose, hornblendic, and unaltered rocks coninto the Lingula flags and 10 ascend into the Upper taining Llandeilo fossils, and he agrees with Griffith and Tremadoc zone, 31 being peculiar. The Upper Tremadoc King in regarding these as probably Cambrian. He suggests beds contain, as at present ascertained, 33 species, of which that they are Upper Cambrian, which would imply that 9 are peculiar, and 13 or possibly 15 pass up into the Upper Cambrian rocks pass conformably into the Llandeilu Arenig group. It is at the top of the Upper Tremadoc formation without the occurrence of the thick Arenig rocks strata that the line between the Cambrian and Silurian of Wales. In a difficult country, however, broken by faults systems is here drawn. According to Professor Ramsay, and greatly metamorphosed, an unconformability might there is evidence of a physical break at the top of the easily escape detection. Tremadoc beds of Wales, so that on a large scale the next succeeding or Arenig strata repose unconformably upon everything older than themselves. Mr Etheridge also shows that the paleontological break is nearly complete, only about 7 per cent. of the fossils of the one series passing over into the other. Out of 184 known Arenig species, not more than 13 are common to the Tremadoc beds underneath. Besides these important facts the character of the Arenig fauna strongly distinguishes it from that of the formations below, and further supports the line of division here adopted between the Cambrian and Silurian systems.

In the north-west of Scotland a mass of reddish-brown and chocolate-coloured sandstone and conglomerate (at least 8000 feet thick in the Loch Torridon district) lies unconformably upon the fundamental gneiss in nearly horizontal or gently inclined beds. It rises into picturesque groups of mountains which stand out as striking monuments of denudation, seeing that the truncated ends of their component flat strata can be traced even from a distance forming parallel bars along the slopes and precipices. The denuda tion must have been considerable even in early Silurian times, for the sandstones are unconformably overlaid by quartz-rocks and limestones containing Lower Siluriau fossils. No trace of organic remains of any kind has been found in the red sandstones themselves. They were at one time regarded as Old Red Sandstone, though Macculloch,

Hicks, Quart. Journ. Geol. Soc., xxix. 39.

CONTINENTAL EUROPE.-According to the classification adopted by M. Barrande, the older Paleozoic rocks of Europe suggest an early division of the area of this con tinent into two regions or provinces,-a northern province, embracing the British Islands, and extending through North Germany into Scandinavia, on the one hand, and into Russia on the other, and a central-European province, including Boliemia, France, Spain, Portugal, and Sardinia.

Bohemia.-The classic researches of M. Barrande have given to the oldest fossiliferous rocks of Bohemia an extraj ordinary interest. He has made known the existence there of a remarkable suite of organic remains representative of those which characterize the Cambrian rocks of Britain. At the base of the geological formations of that region lie the Archæan gneisses already described. These are overlaid by vast masses of schists, conglomerates, quartzites! slates, and igneous rocks, which have been more or less metamorphosed, and are singularly barren of organio remains, though some of them have yielded traces of anne lides. They pass up into certain grey and green fissile shales, in which the earliest well-marked fossils occur. organic contents of this zone (Étage C) form what M. Barrande terms his primordial fauna, which contains 40 or more species, of which 27 are trilobites, belonging to the characteristic Cambrian genera-Paradoxides (12), Agnos tus (5), Conocoryphe (4), Ellipsocephalus (2), Hydrocephalus (2), Arionellus (1), Sao (1). Not a single species of any one of these genera, save Agnostus (of which 4 specics

The

appear in the second fauna), has been found by M. Barrande higher than his primordial zone. Among other organisms in this primordial fauna, the brachiopods are represented by 2 species (Orthis and Orbicula), the pteropods by 5 (Theca), and the echinoderms by 5 cystideans.

Scandinavia.-In Norway the vast masses of Archæan gnaiss (Tellemark) are overlaid by schists, red sandstones, and conglomerates. These are termed the Sparagmite formation, and have hitherto proved barren of fossila. They are covered, however, by beds containing Dictyonema Norvegicum, which may represent the primordial zone of Barranda. In Sweden the sparagmite. formation has been more productive of organic remains. It is there represented by a sandy zone not more than 50 or 60 feet thick-a poor equivalent for the great mass of strata in the Cambrian system of Wales. It was, originally termed the Regio Fucoidarum by Angelin, from the fucoids alone found in it. In more recent years, however, its list of organic remains has been considerably increased; 12 species of plants, chiefly facoids, but including some (Eophyton) of higher grade, 9 species of annelides, 4 brachiopods, a pteropod, a bryozman, a coral, a crinoid, and a sponge have been obtained. Above the strata containing these organisms comes a zone which has yielded 77 species of primordial trilobites, including the genera Agnostus (19 species), coryphe (13), Olenus (21), Paradoxides (9).

The important system of rocks next to be described was first investigated by the late Sir R. Murchison in Wales and the bordering counties of England. He found it to be charCono-acteristically developed over the tract once inhabited by the Silures, an ancient British tribe, and he thence chose the name of Silurian as a convenient designation. It there passes down conformably into the Tremadoc slates at the top of the Cambrian series, and is covered conformably by the base of the Old Red Sandstone.

NORTH AMERICA.-Rocks corresponding in position and in the general character of their organic contents with the Cambrian formations of Britain have been recognized in different parts of the United States and Canada. They appear in Newfoundland, whence, ranging by Nova Scotia and New Brunswick, they enter Canada, the northern parts of New York, Vermont, and eastern Massachusetts. They rise again along the Appalachian ridge, in Wisconsin, Minnesota, Missouri, Arkansas, Texas, and Georgia. Westwards from the great valley of the Mississippi, where they have been found in many places, they reappear from under the Secondary and younger Palæozoic rocks of the Rocky Mountains. They have been divided by American geologists into two formations-(1) Acadian, a mass (2000 feet) of grey and dark shales and some sandstones; and (2) Potsdain (or Georgian), which attains in Newfoundland á A. Lower Silurian. 3. Bala and Caradoc group.. depth of 5600 feet, but thins away westward and southward till in the valley of the St Lawrence, where it was studied by Logan and his associates of the Geological Survey of Canada, it is only from 300 to 600 feet thick.

GREAT BRITAIN.-In the typical area where Murchison's discoveries were first made he found the Silurian rocks divisible into two great and well-marked series, which he termed Lower and Upper. This classification has been found to hold good over a large part of the world. The subjoined table shows the present arrangement and nomenclature of the various subdivisions of the Silurian system. 6. Wenlock group.......... 7. Ludlow group 5. Upper Llandovery group 4. Lower Llandovery group..

Among the organie remains of the North American Cambrian rocks fucoid casts appear in many of the sandstones, but no traces of higher vegetation. The Acadian formation has yielded primordial trilobites of the genera Paradoxides, Conocoryphe, Agnostus, and some others; brachiopods of the genera Lingulella, Discina, Obolella, and Orthis; and several kinds of annelide-tracks. The Potsdam rocks contain a few sponges, the earliest forms of graptolite, some brachiopods, including, besides the, genera in the Acadian beds, Obolus, Camarella, and Orthisina; some pteropods (Hyolites or Theca); two species of Orthoceras; annelide tracks; trilobites of the genera Conocoryphe, Agnostus, Dikelocephalus, Olenellus, Ptychaspis, Chariocephalus, Aglaspis, and Illanurus.

M. Barrande has called attention to the remarkable uniformity of character in the organic remains of his primordial zone over the continents of Europe and America. He pub lished in 1871 the subjoined table, to show how close is the parallelism between the proportions in which the different classes of the animal kingdom are represented.1

1 Trilobites, Prague, 1871, p. 196. Since the publication of this table the progress of research has increased the number of species from some localities; but the general facies of the primordial fauna has not been materially affected thereby.

R. Upper Silurian.

2. Llandeilo group......

Feet.

1,950

1,600

1,500

1,000

6,000

2,500

1. Arenig or Stiper Stone group................... 4,000

Approximate average thickness -18,550

A. Lower Silurian.

1. Arenig or Stiper Stone Group.-These rocks consist of dark slates, shales, flags, and bands of sandstone, They are abundantly developed in the Arenig mountain, where, as originally described by Sedgwick, they contain masses of associated porphyry. Throughout that district they have been deposited at a time when streams of lava and showers of volcanic ashes were thrown out in great quantity from submarine vents. They contain an abundant suite of organic remains (184 species), of which only 13 species are, common to the Tremadoc beds below. Trilobites occur of the genera Eglina, Agnostus, Ampyx, Barrandea, Calymene, Cheirarus, Illanopsis, Illanus, Ogygia, Phacops, and Trinucleus. Three species of ptero pods (Conularia, Theca), 18 species of brachiopods (Lingula, Lingulella, Obolella, Discina, Siphonotreta, Orthis), 8 lamellibranchs, 3 gasteropods, and 5 cephalopods have been found, but the most abundant organisms are the graptolites, of which the Arenig rocks of St David's, in Pembrokeshire, have yielded 48 species, which belong to 20 genera, including Didymograptus, Tetragraptus, Diplo graptus, Dendrograptus, and Callograptus. Altogether Hicks, Quart. Journ. Geol. Soc.. xxxi. 167: Hopkinson and Lap. worth, ibid., p. 635.