The Arenig Beds succeed the Tremadoc slates at St. David's in South Wales, and in North Wales they also overlie the Tremadoc slates between Towyn and the neighbourhoods of Dolgelli, Ffestiniog, Tremadoc, and Criccieth in Caernarvonshire, north of which they also occur in part of the country between Caernarvon and Bangor. They were first distinguished by Professor Sedgwick, and named Arenig slates, and afterwards termed lower Llandeilo beds by Sir Roderick Murchison, who had previously included them as part of the Llandeilo flags in his descriptions and sections of the Lower Silurian rocks that lie west of the Stiper Stones near Shelve, in Shropshire.

In the large district of Merionethshire the Arenig slates appear at the base of the great volcanic series of felspathic lavas and ashes, of which the mountains of Cader Idris, Aran Mowddwy, Arenig, and the Moelwyns form distinguished features in the landscape. They are in these districts never more than about 800 feet in thickness, and the Arenig beds of Merionethshire, at their base invariably consist of beds of grit, sometimes conglomeratic. The higher strata of this sub-formation are generally slaty. For reasons that will afterwards appear, I believe that the Arenig strata, on a

[70 Arenig Slates.]

large scale, rest unconformably on the underlying rocks of North Wales.

In Cumberland the Arenig slates form the mountains of Skiddaw and Saddleback, and from the borders of the Old Red Sandstone, a few miles further east, they stretch right across the country westward to Egremont and northward to Sunderland, south of which town, near Cockermouth, they are directly overlaid by the Carboniferous Limestone. In that country they have usually been called the Skiddaw slates. In Scotland the Durness strata belong to the same rocks.

In Britain the fossils that belong to this part of the Silurian series are not very numerous, taken as a whole, though some groups are remarkably abundant. As far as observation has gone, Hydrozoa of the sub-class Graptolitidæ first appear in these strata, including some 20 genera and 48 species. The greatest number of species belong to the genus Didymograptus, of which 20 species have been named, after which come Tetragraptus, Diplograptus, Dichograptus, and Dendrograptus.

Eighteen genera and 47 species of Trilobites occur in the same rocks, including Agnostus (A. hirundo, &c.); Asaphus (A. Homfrayi, &c.); Ogygia (O. Selwynii, &c); Trinucleus (T. Ramsayi, &c.); Calymene (C. parvifrons, &c.), and many others. Of Brachiopoda there are 7 genera and 18 species including three Lingulas, Lingulella Davisii and L. lepis, 7 species of Orthis, including O. calligramma and O. lenticutaris; 2 species of Obolella; 2 of Discina, and others Of Lamellibranchiata there are only 5 genera and 8 species known, including Modiolopsis trapeziœformis, Palæarca socialis, and P. amygdalus, Ctenodonta elongata, &c., and Redonia Anglica. Pteropoda of

[Llandeilo and Caradoc Beds, 71]

the genera Theca and Conularia are found, and 6 species of Bellerophon, and of Cephalopoda there are 5 species of Orthoceras. Of univalve shells we have only 3 species—Pleurotomaria Llanvernensis, Ophileta, and Raphistoma, and several other fossils needless to enumerate.

In all, 184 species are known at present in the Arenig beds, mostly characteristic of these strata, for only about 8 per cent. pass upward into this horizon from the Tremadoc beds, a proportion of which go down into the Lingula flags, and about 7 per cent. pass upward into the Llandeilo flags.

Though in Wales the base of the Arenig beds is clear, it seems as yet impossible to draw any definite physical boundary between the Arenig beds and the overlying Llandeilo slates, for there is nothing like unconformity, and no marked lithological difference in the passage from one to the other. We have already seen that there is a very limited passage of species from the Arenig slates into those of the so-called Llandeilo series.1

Just about this time an important episode took place in the history of the Llandeilo and Bala beds over large tracts of Wales and Cumberland, for a series of volcanic eruptions occurred on a great scale while the strata were being deposited (fig. 62, p. 322). To this subject I shall by-and-by return.

In North Wales the Llandeilo and Baja or Caradoc beds combined, attain a thickness of from 4,000 to 6,000 feet, consisting chiefly of slaty rocks, sometimes interstratified with grits and occasional bands of limestone, of which the Bala Limestone is the most conspicuous. The whole series ranges right round the mountains of

1 The Llandeilo flags of North Wales are very unlike those of Llandeilo, which are generally called upper Llandeilo beds.

[72 Llandeilo and Caradoc Beds.]

Cader Idris, Aran Mowddwy, Arenig, and the Moelwyns, resting on the lava beds and ashes, and overlaid on the east by Upper Silurian strata, fig. 57, p. 304. They also form, with igneous rocks, the larger part of the Berwyn mountains, and with the Arenig slates the whole of the ground between the Stiper Stones and the Upper Silurian rocks of Chirbury and Montgomery, fig. 13, p. 59. The typical Caradoc Sandstone, crossing the strike, ranges between Church Stretton and Caer Caradoc, from whence it stretches in a broad band northward towards the Wrekin, and southward to Corston. The greater part of South Wales is formed of slates and grits of Llandeilo and Caradoc age, lying west and north of the Upper Silurian and Old Red Sandstone strata, and the same formations, associated with volcanic rocks, rise like an island surrounded by Upper Silurian strata, in the country between Builth and Llandegley in Radnorshire.

In South Wales, where they were first described by Murchison, the Llandeilo beds consist of sandy calcareous flags, black slaty rocks, and beds of grit and sandstone. A few beds of limestone occur in them in Carmarthenshire, at Llandeilo, and in Pembrokeshire near Narberth; and the Bala limestone is found higher in the series in the Caradoc or Bala beds of Merionethshire. They are often highly fossiliferous. There is a much larger development of fossils in the Llandeilo flags than in the pre-existing Silurian strata. The Trilobites of the Llandeilo beds are mostly peculiar to it, and the genera Æglina, Barrandia, and Ogygia are very common, Ogygia Buchii being especially characteristic. Viewed as a whole, however, the Llandeilo beds, as already stated, pass insensibly into, and have many genera and species in common with the Caradoc
andeilo and Caradoc Fossils. 73]

FIG. 18.
llandeilo flag, Caradoc and Bala Fossils
Group of Llandeilo flag and Caradoc or Bala Fossils.

[74 Llandeilo and Caradoc Beds.]

Sandstone or Bala beds. 19 genera and 34 species of orals have been described in these lower Silurian strata, among which Heliolites and Petraia are perhaps the most common.

Fragments of Echinodermata are common, including Cystideans, common in the Bala Limestone, and one star-fish, Palœaster Caractaci. In all, more than 40 genera and 200 species of Trilobites have been described from the whole series of Lower Silurian British rocks, among the chief of which are species of Olenus, Agnostus, Ampyx, Lichas, Ogygia, Acidaspis, Asaphus, Harpes, Illœnus, Phacops, and Trinucleus (T. Caractaci). In the Caradoc beds alone, 23 genera and 111 species are known. Of bivalve shells there are 22 genera and 171 species of Brachiopoda, the most common of which belong to the genera Strophomena, Leptœna, Lingula, Orthis, and Rhynchonella.

Of the Lamellibranchiate molluscs there are 17 genera and 87 species known at present, prominent among which are Ctenodonta, Modiolopsis, Pterinœa, Palœarca, and Ambortychia. Of Pteropoda there are known 6 genera and 31 species, of which Theca is most abundant; 16 genera and 66 species of Gasteropoda, the most characteristic of which in point of numbers are Euomphalus (10), Murchisonia (15), Pleurotomaria, Cyclortema, and Holopœa. Of Nucleobranchiata, Bellerophon (14). Of the Cephalopoda there are 10 genera and 62 species—Cyrtoceras (5), Lituites (6), Orthoceras (42), Phragmoceras (1), and others. No fishes nor any other vertebrate animals have yet been found in the Lower Silurian rocks of Wales or elsewhere.

In Cumberland the Coniston Limestone is believed to be the equivalent of the Bala Limestone of North

[Igneous Rocks. 75]

Wales, and the assemblage of fossils in each is very nearly the same.

I have already mentioned the occurrence of an important episode characterised by volcanic eruptions, during the accumulation of the Lower Silurian strata in Wales. The proof of this is that in Carnarvonshire and Merionethshire extensive interstratified sheets of felspathic lavas and ashes are associated with the Silurian rocks on two horizons, the lower that of the Llandeilo beds, and the higher in the Caradoc series. I do not, however, wish to imply that between them there was a complete cessation of volcanic activity, but simply that in what is now the region of North Wales, there was for a time an interval of comparative repose.

If any one will examine the Geological Survey maps of North Wales, he will observe that opposite Barmouth, beginning with the hills on the south side of the estuary of the Mawddach, a great series of igneous rocks sweep round the country in a crescent form, including the mountains of Cader Idris, Aran Mowddwy, the Arenigs, and lastly the Moelwyns, the high southern escarpments of which overlook from the north the beautiful vale of Ffestiniog. These consist of felspathic lavas, and interstratified ashes or tufas, the whole being also associated with bands of Silurian slate, which are sometimes found to be fossiliferous, especially when bedding and cleavage coincide. Among these volcanic rocks, but especially in the Arenig, Tremadoc, and Lingula beds below them, there are numerous lines and bosses of greenstones (diorites, &c.), and also of more purely felspathic traps, which are not interbedded but distinctly intrusive. These I have elsewhere shown give evidence of the underground working of the

[76 Igneous Rocks.]

melted matter, the eruption of which to the surface through volcanic rents, produced the lavaflows and ashes already mentioned. The ashy beds are sometimes coarse and tufaceous, but were also often formed of fine volcanic dust, which being now consolidated into hard felspathic rocks, are at first sight somewhat difficult to distinguish from the associated lavas. Practice, however, renders it comparatively easy, and in distinguishing the difference, the observer is aided by the circumstance, that underneath each lava current the slates, once beds of mud, are apt to be baked and porcellanised at the point of junction with the originally hot lavas, which having in the meanwhile cooled, the slaty beds that rest on them are in that respect unaltered.

The second series of eruptions may be traced as follows. Near Bala, not far below the limestone, there are a few thin bands of volcanic ashes. These, as we go northward to the rivers Machno, Lledr, and Conwy, gradually thicken, and by-and-by get mingled in that slaty area with numerous thin and thick bands of felspathic lavas, the importance of which as large masses, culminates in Snowdon and the surrounding area, going northward by Glyder-fawr, Glyder-fach, Carnedd Dafydd, and Carnedd Llewelyn, and so on to Conway. South of Snowdon the same kinds of lavas and ashes are seen in force on the sides of Moel Hebog, and the great mass of Llwyd-mawr near Dolbenmaen.

Other large bosses of intrusive rocks, mostly felspathic, occur on Y-Foel-frâs, between Snowdon and Conway, another between Llanllyfni and Bethesda, a third near the eastern shore of Menai Straits, and many more including the beautiful mountains of Yr Eifl, or The Rivals, in the north horn of Cardigan Bay,

[Unconformity. 77]

known as the district of Lleyn. These, ere exposure by denudation, probably were the roots of the volcanos, or in other words the deep-seated centres from whence the explosive force of steam drove out the lavas and showers of ashes, which, during successive eruptions, with minor periods of repose, got interstratified with the mud and sand beds that were deposited in the sea of the Llandeilo and Bala or Caradoc period.

On a smaller scale similar volcanic rocks are interstratified with the Llandeilo and Bala beds of the Berwyn Hills, also of the Breidden Hills, and the hills west of the Longmynd and Stiper Stones towards Chirbury and Church Stoke, of the country between Builth and Llandegley in Radnorshire, and in North Pembrokeshire from the ground round St. David's, extending for many miles to the east, by Mathry, Fishguard, St. Dogmells and Mynydd Preselley.

The next question that occurs to me is, what was the nature of the physical geography of this area during the deposition of the Arenig slates, and also at a later epoch when the Llandeilo and Caradoc or Bala beds were being deposited.

With regard to the Arenig slates in Pembrokeshire and Merionethshire, I know of no signs of unconformity, that is to say, of a lapse of time unrepresented by the deposition of marine strata either in Pembrokeshire or in Merionethshire, unless there be some symptoms of it in the latter county. But when we go further north into Carnarvonshire, the case is different. There, at the widening of the Passes of Llanberis and Nant Ffrancon, the Lingula flags are not more than 2,000 feet thick, whereas further south, between Ffestiniog and Portmadoc, they are at least 4,000 feet in thickness. Furthermore, in those valleys in Caernarvonshire

[78 Physical]

there is, as yet, no certainty of the existence of the Tremadoc slates, and these ought to be found overlying the Lingula flags if the whole of the Silurian series were present. Still further, as we approach Caernarvon and Bangor, even the Lingula flags are absent, and the Arenig slates are found lying directly on the purple slates and conglomerates of the Cambrian series all the way from Bangor to Caernarvon. This clearly shows that the base of the Arenig slates has overlapped the whole of the Tremadoc slates and Lingula flags, in the area between the Ffestiniog and Portmadoc country and the neighbourhood of the Menai Straits, and an overlap so great means unconformity between the strata; or, in other words, in this area the strata of older date than the Arenig slates had been raised above the sea, and subjected to sub-aerial agencies of denudation, while the deposition of the Arenig slates was going on elsewhere. In this manner, therefore, it happens that the Arenig slates are now found resting directly on the Cambrian strata, without the intervention of the missing members of the series, viz., the Tremadoc slates and Lingula flags; and still further north, in Anglesea, these strata are also wanting.

The effect of this episode in the physical geography of the area seems to have been, that at this period a tract of land lay in the north-west of what is now Wales, and probably far beyond that district during the deposition of the Arenig strata on its borders, but what the features of that land were I cannot say, except that it may have extended to Ireland, where there is a similar unconformity, the Lingula flags and Tremadoc slates being also wanting in Wicklow. Probably the whole region was low and unimposing.

The next question that arises is, what was the

[Geography. 79]

nature of the physical geography during the time of the volcanic eruptions already mentioned? To me it seems to have been somewhat of this sort.

On the margin of the ancient land, or at some distance therefrom, volcanic eruptions took place in the sea-bottom somewhat of the nature of that which in 1831 took place in the Mediterranean between the islands of Pantellaria and the south-west coast of Sicily. This eruption was preceded by an earthquake on June 28, and on July 10 John Corrao from his ship saw a column of water 60 feet high and 800 yards in circumference spout into the air, succeeded by dense steam, which rose to a height of 1,800 feet. On the 18th the same mariner found an island twelve feet high, from the crater of which immense columns of steam and volcanic ashes were being ejected, 'the sea around being covered with floating cinders and dead fish.'1 The eruption continued into August, when, by the ejection of what is often called volcanic ashes, viz., pumice, scoriæ, and lapilli, on the 4th of that month the island was said to have been more than 200 feet in height and 3 miles in circumference. From that time it gradually decreased in size, owing to the action of the waves, and towards the close of the year the island had been destroyed and disappeared, leaving only a reef beneath the sea with a black rock in the centre, from 9 to 11 feet under water, and which probably marked the position of the funnel of the short-lived volcano. Before the eruption took place it so happened that Captain (afterwards Admiral) W. H. Smyth sounded on the spot in more than 100 fathoms, and this, added to 200 feet that the island rose above the sea, gives 800 feet as the height of the cone from the

1 Lyell's 'Principles of Geology,' vol. ii. p. 60, 12th editicn.

[8o Physical]

bottom of the sea to its summit. In a case such as this, it is easy to see that the ordinary marine sediments of the area would get intermingled with volcanic ashes, and possibly with submarine streams of lava.

Explosions of steam accompanied by floating cinders are mentioned by Darwin as occurring at intervals in the South Atlantic; and anyone who will tax his memory a little will recollect that a large proportion of the volcanoes of the world are islands, or in islands, in the Atlantic, the Indian Ocean, the Indian Archipelago, and the Pacific Ocean, south and north. It has been often remarked that almost all volcanoes are in the neighbourhood of the sea.

I think, then, at the time of the deposition of the Llandeilo and Bala beds of our area, our terrestrial scenery consisted of groups of volcanic islands scattered over the area of what is now North Wales and South Wales, and extending westward into the region of the Irish strata of the same age, and northward as far as the sea that then rolled where Cumberland now stands; for there also volcanic rocks occur in great force, all of the same general character as those found in Wales. There is however, this difference between the two areas, that, whereas in Wales ordinary sediments are plentifully interstratified with lavas and ashes, and sometimes even lithologically intermingled with volcanic ashes, in the Cumbrian area it is only for a few feet at the very base of the volcanic series that interstratifications take place, the whole of the rest of these Silurian volcanic rocks of Westmoreland and Cumberland being quite destitute of any intermixture of marine sediments. Exclusive of intrusive rocks, the whole consists of purely terrestrial lavas, volcanic conglomerates and ashes, the latter often well stratified, for where showers of ashes fall

[Geography. 81]

there layers of stratification will be formed, whether they fall in the sea or on land. It has been suggested by Mr. Ward that some of this fine volcanic dust fell into lakes that filled old craters or areas of subsidence during periods of partial repose, and this seems highly probable, for the finely divided matter is so beautifully stratified, that these beds were, and still are by some, mistaken for marine strata.

When we consider the vast amount of these products of ancient volcanoes, there can be no doubt that, rising from the sea, some of them must have rivalled Etna in height, and others of the great active volcanoes of the present day, and, as most volcanoes have a conical form, we can easily fancy the magnificent cones of those of Lower Silurian age. But that is all we know respecting them, and whether or not they were clothed, like Etna, with terrestrial vegetation, no man can tell. It is hard to believe that they were utterly barren, but as yet no trace of a flora has been found in Lower Silurian strata.

There is another point bearing on the physical geography of the time that has sometimes crossed my mind in connection with these island volcanoes, which is, that we may, with some show of probability, surmise, that then, as now, the prevalent winds of this region blew from the west and southwest, for the following reason. In Merionethshire and Caernarvonshire the various volcanic products gradually thin out and disappear to the west, between the ground south of the estuary of the Mawddach, and the neighbourhood of Tremadoc on the north. As we pass round the large crescent-shaped masses of lavas and ashes it becomes evident as a rule that the ashy series of beds show a tendency to thicken more and more in an easterly direction for a space, and finally to decrease in

[82 Physical Geography.]

thickness in the same direction, till, in the Bala country and further north, they are represented only by a few insignificant beds of ashy strata, a character of which the Bala limestone itself sometimes feebly partakes. The idea is, that the prevalent westerly winds had a tendency during eruptions to blow the volcanic dust and lapilli eastward, and that these materials fell thickest near the vents and at middle distances, and gradually decreased in quantity the further east they were carried.

To those unaccustomed to technical geological arguments a word of warning remains. Let no reader suppose that in Wales he will now find clear traces of these old volcanic cones and craters in their pristine form, such, for example, as the extinct craters of Auvergne and the Eifel. Semi-circular hollows surrounded by igneous rocks like those of Cwm Idwal and Cwm Llafar he will find plentiful enough, and these, in old guide-books and other popular literary productions, have sometimes been described as craters. So far from that being the case, such cirques or corries are ancient valleys of erosion, the rocks of which have been exposed to the weather perhaps ever since Upper Silurian times, and have been subsequently modified by glaciers, during the last Glacial Epoch, in days, comparatively speaking, not far removed from our own. The truth about these ancient volcanoes is, that long after they became extinct the whole Lower Silurian area was disturbed and thrown into anticlinal and synclinal curves, which suffered denudations before the beginning of the deposition of the Tipper Silurian rocks, and the positions in which the lavas and ashes now stand will approximately be best understood, if we suppose Etna by simlar disturbances to he half turned on its side,

[Lower Silurian Rocks, Scotland. 83]

and afterwards that the exposed portion should be irregularly cut away and destroyed by processes of long-continued waste and decay, partly sub-aerial and partly marine.

The remaining areas in Great Britain occupied by Lower Silurian rocks lie in Scotland. The southern district extends from St. Abbs Head on the east to Portpatrick on the west coast, forming the uplands of the Lammermuir, Moorfoot, and Carrick Hills, fig. 55, p. 287. They chiefly consist of thick banded strata of grits and slaty beds, much contorted, and in the western part of the area, where bosses of granite and other igneous rocks occur, they are often metamorphosed. The fossils which they contain prove them to belong to the Llandeilo, and Bala or Caradoc series.

In Wigtonshire great blocks of gneiss, granite, &c. are imbedded in the dark slaty strata near Corswall, and similar large blocks occur in Carrick in Ayrshire.  Where they came from I cannot say, for all the nearest granite bosses in Kirkcudbrightshire and Arran are of later date than the strata amid which these erratic blocks are found. I therefore incline to the opinion that they must have been derived from some Laurentian region, of which parts of the mainland and of the Outer Hebrides then formed portions, and when I first saw them I could, and still can conceive of no agent capable of transporting such large blocks, and dropping them into the graptolite-bearing mud, save that of icebergs. One of the blocks measured by me near Corswall, in 1865, is 9 feet in length, and they are of all sizes, from an inch or two up to several feet in diameter. Many persons have considered, and will still consider, this hypothesis of their origin to be overbold, but I do not shrink from repeating it, and I may mention that the same view

[84 Lower Silurian Rocks.]

with regard to these ancient boulder beds is held by Professor Geikie and Mr. James Geikie, who mapped the country.

The Lower Silurian rocks of the south, pass underneath the Old Red Sandstone and Carboniferous rocks of the midland parts of Scotland, and rise again on the north in the Grampian Mountains. A great fault,

1  I shall by-and-by have to notice the recurrence of glacial episodes at various epochs in geological history, a subject with which ever since 1855 I have had a good deal to do. (On Permian Breccias, &c. 'Journal of the Geol. Soc.' vol. ii. p. 185). It is difficult to make out the ground on which all the old, and many of the middle aged geologists, have cast aside the various evidences that have been adduced in support of the recurrence of glacial epochs or episodes, especially as I remember no argument that has been brought forward, excepting that in old times, the radiation of internal heat through the crust of a cooling globe, produced warm and uniform climates all over the surface, and that the further you go back in time the hotter they were. The Lias was accumulated in warm seas, and, if so those of the Carboniferous times were warmer, and those of Silurian times warmer still, and I have heard a distinguished geologist declare in a public lecture, that the tropical vegetation of the Coal-measures, was due to the heat that radiated outwards from the earth's crust, aided by that produced by the flaring volcanoes of that epoch! Undoubtedly there must have been a geologically prehistoric time, when internal heat may have acted on the surface, and perhaps the sun may have been hotter than now, and that also had its effect. I, however, can see no signs of these internal and external interferences since the times in which the authentic records of geological history have been preserved, and these extend backward earlier than the Lower Silurian epoch. I recollect the time when what passed for strong arguments were urged to prove that the former great extension of the Alpine glaciers advocated by Agassiz, and the existence of glaciers in the Highlands, Cumberland, and Wales, proved by him and Buckland, were mere myths. Now, however, there is such a persistent run upon the subject, that more memoirs have been, and still are being, written about it, than persaps on any other geological question. Coincident with this a beginning of the acceptance of the theory of the recurrence of glacial episodes, is slowly making its way both in England and the Continent.

[Scotland. 85]

proved by Professor Geikie, runs at the base of that so-called chain right across Scotland, from the neighbourhood of Stonehaven on the east coast, to Loch Lomond on the west. Its effect is to throw down the Old Red Sandstone on the south-east, partly against the Silurian rocks, and partly against volcanic tufas and other strata belonging to the Old Red Sandstone itself. From that region, nearly the whole of the Highlands, from the Grampians to the north coast of Scotland consists of Lower Silurian rocks often intensely contorted, and formed of quartz-rocks and flagstones, gneissic and micaceous schists, clay slate, and chlorite slate. Associated with these, there are certain limestones, sometimes crystalline, but where less altered, sometimes fossiliferous, fig. 55, p. 287. One of these, near the base of the Silurian series, runs in a long band from Loch Erriboll, on the north coast, southward to Loch Broom, where for a space of about fifteen miles it is lost, to reappear between the east side of Sleugach and Loch Carron. The same limestone is well seen in the Island of Lismore in Loch Linnhe, and here and there on the sides of Strathmore or the Great Glen (a line of fault), through which the Caledonian Canal was constructed. Elsewhere in the Highlands, further east, streaks of limestone occur. Immense masses of granite here and there rise in the midst of the strata, one of the smaller of which forms great part of Ben Nevis, the highest mountain in Britain, 4,406 feet in height, and another the splendid peaks of the Island of Arran. No interbedded igneous rocks have yet been found among the Silurian rocks of Scotland.

The strata of the Highlands, not of Lower Silurian age, are the Laurentian gneiss and Cambrian conglomerates and sandstones already mentioned, intersected by

[86 Lower Silurian Rocks, Scotland.]

so many noble Fjords between Sleat and Cape Wrath, while on the east there are large tracts of Old Red Sandstone, more or less extending from Thurso in Caithness to the Great Glen, Moray Firth, the river Spey, and yet further east. Fig. 55, p. 287.

In times within the memory of the writer, all these metamorphic rocks of the Highlands were classed in Wernerian style as Primitive strata, thrown down in hot seas before the creation of life in the world. The progress of research showed that gneiss and other rocks now called metamorphic, are of many geological ages; and the fortunate discovery of fossils in these strata, at Durness, by Mr. C. Peach, in 1854, showed them to be of Arenig ages a discovery the importance of which was at once seen by Sir Roderick Murchison, who by this means, revolutionised the geology of the greater part of the northern half of Scotland. Feeling anxious to have a second opinion respecting the justness of his new views, he asked me to accompany him on a long tour through the northern Highlands in 1859, when I mapped part of the country at Durness and Loch Eriboll, and the whole matter seemed to me so plain, that the wonder is, that any man with eyes ever dreamed of disputing it. In these days no one now thinks of denying the Lower Silurian age of the chief part of the gneissic rocks of Scotland, the features of which have been mapped by Professor Geikie, first in concert with Sir Roderick Murchison, and afterwards personally in more detail.

With regard to the physical geography of the time, little is certain but this, that almost the whole of the area now called Scotland was under the sea, during the time

1  See 'Geological Map of Scotland,' last edition, by Archibald Geikie, LL.D., F.R.S., 1876.

[Physical Geography. 87]

that these Lower Silurian strata were being deposited. The only sign of pre-existing land, is found on the west coast between Cape Wrath and Loch Torridon, where the Llandeilo beds lie alike unconformably on the Cambrian and Laurentian strata. This proves, that when the lowest Llandeilo beds began to be deposited, the underlying rocks formed the eastern margin of a territory, of which probably our Outer Hebrides was only a part, but how far it may have stretched westward it is impossible to say. However that may have been, it seems certain that long before the uppermost strata of the Lower Silurian rocks of Scotland were deposited, these fragments of an older land, which are still preserved on the west, had been long submerged and buried under the accumulating piles of the Silurian strata. That even then an extensive land lay not far off is certain, for the extent and great thickness of the Lower Silurian rocks affords a measure of the amount of waste of a pre-existing territory, the partial and gradual destruction of which, by all the agencies of denudation, provided mechanical sediments wherewith to form thousands of feet of Silurian strata of mud and sand, first consolidated, and long after metamorphosed into quartzite, gneiss, and mica-schist. This land may have occupied an area now covered by the Atlantic ocean.