CHAPTER LI

SUMMARY AND GENERAL DEDUCTIONS

The foregoing chapters comprise a connected narrative of the history of volcanic action in the area of the British Isles during the vast succession of ages from the early Archæan dawn down to the latest eruptions of Tertiary time. In this final chapter I propose to present a brief summary of the facts of largest import and widest interest which this protracted history has placed before us, together with a statement of deductions which may be drawn from them regarding the nature and progress of volcanism in the evolution of the globe.

1. Among the broad features which soonest arrest attention in such a survey is the geographical position of the theatre of this volcanic activity. In the distribution of volcanoes at the present time we are familiar with their tendency to range themselves along continental borders or in oceanic islands. The volcanic energy so conspicuous in the geological history of Britain has shown itself along the western or Atlantic margin of the European continent. When the eruptions have not been actually on the land itself, they have taken place within the shallow tracts near the land, where the lavas and tuffs have been interstratified with sediments derived from the adjacent coasts.

Moreover the volcanic rocks in Britain are ranged along the greatest length of the group of islands, in a general north and south line, from the south of Devonshire to the far Shetlands. It is on the western side of the country that they occur. East of a line drawn from Berwick by Leicester to Exeter, although the geological formations, ranging from the Carboniferous Limestone to the latest Pleistocene deposits, are there abundantly exposed to view, they include no contemporaneous volcanic rocks.

2. A second and still more remarkable feature in the geological history of Western Europe is the persistence of volcanic activity along the site of the British Isles. Evidence has been brought forward in these volumes that from the primeval time vaguely termed Archæan, onward to that of the older Tertiary clays and sands of the south-east of England--that is to say, through by far the largest part of geological history, as chronicled in the stratified crust of the globe--this long strip of territory continued to be intermittently a theatre of volcanic action. Every great division of Palæozoic time was marked by volcanic eruptions, sometimes over tracts hundreds of square miles in area and on a colossal scale. After a long period of quiescence during the Mesozoic ages, the renewed outbreak of volcanic energy in older Tertiary time, so marked over the western half of Europe, reached its maximum of development along the Atlantic border, from the north of England and Ireland through the chain of the Inner Hebrides to the Faroe Islands, Iceland and Greenland.

3. Not only has there been a remarkable persistence of volcanic activity over the comparatively limited area of the British Isles, viewed as a whole, but if we examine the different parts of this area we perceive that many of them, of relatively restricted extent, have been the sites of a recrudescence of volcanic action, again and again, through a vast succession of geological periods. While the whole region has been in different quarters and at different times affected, there have been districts where the volcanic fires have been rekindled after long intervals of quiescence, the new vents being opened among or near to the sites of earlier volcanoes. In the south-west of England, for example, the Middle Devonian tuffs and diabases were succeeded in the Carboniferous period by the eruptions of the Culm-measures, and in the very same tracts came last of all the lavas and tuffs of the Permian conglomerates. Still more astonishing is the record of volcanic energy in the south of Scotland, where, within a space of not many hundred square miles, there are the chronicles of the Arenig, Llandeilo and Bala eruptions of the Southern Uplands, the huge piles of lavas and tuffs of the Lower Old Red Sandstone, the long succession of the plateaux and then of the puys of the Carboniferous period, the groups of tuff-cones of the Permian period, and, lastly, the numerous dykes connected with the Tertiary volcanoes.

While some portions of the region have been specially liable to exhibitions of volcanic action, others have continuously escaped. Some of these "horsts," or stationary and unaffected blocks of country, have been surrounded by or have risen close to the borders of this volcanic district, yet have maintained their immunity through a long series of ages. Thus the Central Highlands of Scotland, though they were flanked on the south and south-west by the active volcanoes of the Old Red Sandstone, and again on the south by those of Carboniferous time, had no vents opened on their surface after the metamorphism of their schists. Still more striking perhaps is the immunity of the Southern Uplands. Though they were in large measure surrounded by the volcanoes of the Lower Old Red Sandstone, then by those of the Calciferous Sandstones and Carboniferous Limestone, and though they looked down on the Permian eruptions of Ayrshire and Nithsdale, which spread streams of lava and showers of ash along their flanks, these hills formed a solid block that seems to have resisted perforation by the volcanic funnels. Again, the tracts covered with Carboniferous Limestone in England and Ireland almost entirely escaped from invasion by volcanic eruptions.

We thus learn that even within comparatively restricted regions some portions of the terrestrial crust have been areas of weakness, liable to serve again and again as lines of escape for volcanic energy, while close to them other portions of greater solidity have been persistently left intact.

4. The sites of volcanic vents in all the geological systems wherein they occur in Britain have not usually been determined by any obvious structure in the rocks now visible. They comparatively seldom depend on ascertainable lines of fault, even when faults, probably already existent, occur in their near neighbourhood. This independence, to which, however, there are occasional marked exceptions, comes out more particularly in the coal-fields pierced by vents, for mining operations have there revealed the positions of many more faults than can be traced at the surface. If the sites of the vents have been fixed by dislocations or lines of weakness in the terrestrial crust, these must generally lie below the formations now visible at the surface.

There is one striking connection between the sites of the vents and ancient topographical features to which frequent reference has been made in the foregoing chapters. All through the long volcanic history, as far back as such features can be traced, we see that orifices of discharge for the erupted materials have been opened along low grounds and valleys rather than on ridges and hills. The great central hollow of the Scottish midlands was a depression even as long ago as the time of the Lower Old Red Sandstone, and though it has probably been several times since then filled up, and more or less completely effaced, its ancient features have been partially revealed by extensive denudation. This vast depression, 40 miles broad, between the Highland mountains on the one side and the Southern Uplands on the other, was the chief centre of volcanic activity in western Europe during the latter half of Palæozoic time. The vents of the Old Red Sandstone, Carboniferous and Permian series are scattered all over it, but few or none of them are to be found on the high grounds that bound it. Again, in Tertiary time, the great outpouring of lava took place in the hollow that lay between the ridge of the Outer Hebrides and the mainland of Scotland. This wide and long tract of low ground was buried under upwards of 3000 feet of lava and tuff, but these materials were erupted from fissures and vents within its own border and not from the mountains on either side.

But perhaps the most conspicuous example of any in which the vents keep to the valleys is that supplied by the Permian necks of Nithsdale and the neighbouring glens. These depressions are as old as Permian, and even as Carboniferous time, but they appear to be entirely hollows of erosion; at least they have yielded no evidence that their direction has been determined by lines of fault. The chain of vents can be followed from the lowlands of Ayrshire up to the base of the Southern Uplands, down the wide valley cut by the Nith in these hills and up some of the tributary valleys, and though the volcanoes continued for some time in vigorous eruption, not a trace of any contemporary vent has yet been met with on the surrounding hills.

While the position of volcanic vents in lines of valley may be generally due to guiding lines of fissure in the crust underneath, either within or below the rocks visible at the surface, there may sometimes be conditions in which other dominant causes come into play. The curious coincidence between variations in the upper limit of dykes and inequalities in the configuration of the overlying ground, suggest that where the subterranean magma has ascended to within a comparatively short distance from the surface, a difference of a few hundreds or thousands of feet in the depth of overlying rock, such as the difference of height between the bottom of a valley and the tops of the adjacent hills, may determine the path of escape for the magma through the least thickness of overarching roof.

5. Volcanic phenomena cannot be regarded as a mere isolated and incidental feature in the physics of the globe. During the short time within which man has been observing the operations of existing volcanoes, he has hardly yet had sufficient opportunity of watching how far they can be correlated with other terrestrial movements. Nor, when he endeavours to trace some such connection among the records of the geological past, has he yet collected materials enough to furnish a sufficiently broad and firm basis of comparison. One formidable obstacle is presented by the difficulty in determining chronological equivalents in separated groups of rock. Geologists have tried to discover whether the volcanoes of some particular period or region were in any way connected with such geological changes as extensive plication, dislocations of the crust, or elevation of mountain-chains. In regard to the volcanic history of Britain, various possible relations of this kind obviously suggest themselves. Thus the division of geological time comprised within the Lower Silurian period was undoubtedly an interval of considerable terrestrial disturbance in western Europe. The unconformabilities and overlaps in the series of formations belonging to that period, the frequent conglomerates, the great and often rapid changes in the thickness and lithological characters of the strata, all point to instability of land-surface and sea-floor. During these oscillations a prolonged and widespread series of volcanic eruptions took place. The volcanic manifestations began in Cambrian time and continued in intermittent activity till towards the close of the deposition of the Lower Silurian formations. It is certainly a significant fact that the Upper Silurian deposits, in their lithological characters, present a strong contrast to those that preceded them. They point, on the whole, to quiet sedimentation, during an interval of comparative calm in the terrestrial crust. With this evidence of tranquillity there is, over almost the whole of the British Isles, an entire absence of any trace of renewed volcanic activity. With the exception of the Dingle lavas and tuffs, in the extreme west of Ireland, not a single undoubted instance is yet known of an Upper Silurian volcano.

After the deposition of the Upper Silurian rocks an interval of great terrestrial disturbance ensued, and these rocks over a large part of Britain were intensely plicated and crushed. The movements, continued into the period of the Lower Old Red Sandstone, were, in their later stages, accompanied or, at least, followed by the vast outpourings of lava which now cover so much of the tracts of Old Red Sandstone in Scotland and Ireland.[438]

[Footnote 438: _Trans. Geol. Soc. Edin._ vol. ii. part iii. (1874).]

In proportion as the volcanic energy was vigorous, widespread and long-continued, we may expect it to have been connected with important terrestrial movements affecting extensive regions of the earth. The Tertiary volcanic history seems to afford a remarkable instance of this connection. A wide area of the European continent is dotted over with old centres of volcanic activity which were in eruption at successive epochs throughout the Tertiary period. Of all these centres the most important was that of the north-western basalt-plateaux, where floods of lava were discharged over many thousand square miles from Ireland to Greenland. The geological date of these outpourings probably coincides with the last great orographic movements that gave to the mountain-chains of Europe their latest elevation and dimensions.

But without entering into what must be for the present a field of speculation, we can be assured of one important fact in the connection of ancient volcanoes with movements of the terrestrial crust. A study of the records of volcanic action in Britain proves beyond dispute that the volcanoes of past time have been active on areas of the earth's surface that were sinking and not rising. We usually associate volcanic action with elevation rather than subsidence, and there are certainly abundant proofs of such elevation around active or recently extinct volcanoes. Many of the active vents of the present time, like Vesuvius and Etna, began with submarine eruptions and have been gradually upraised into land. It may be, however, that such uprise is merely a temporary incident, and that if we could survey the whole geological period of which human history chronicles so small a part, we might find that subsidence, and not upheaval, is ultimately the rule over volcanic areas.

Be this as it may, there can be no question that with the one solitary exception of the Tertiary volcanoes, which were terrestrial and not submarine, all the British vents were carried down and eventually buried under aqueous sediments. Even the Tertiary lava-fields have in many places sunk down below sea-level since their eruptions ceased.

That there are any Palæozoic volcanic rocks now visible at the surface is obviously due to subsequent movements not immediately connected with their original conditions of eruption, and to gigantic denudation. The amount of subsidence which followed on a volcanic episode was sometimes enormous, even within the same geological period, as one may see by observing the prodigious piles of sedimentary material heaped over the lavas and tuffs of Arenig time, or over those of the Lower Old Red Sandstone. I do not wish to maintain that the downward movement was necessarily a consequence of volcanic ejections, for we know that it took place over tracts remote from centres of eruption. But I have sometimes asked myself whether it was not possibly increased as a sequel to vigorous volcanic action; whether, for instance, the great depth of the Palæozoic sedimentary rocks in some regions, as compared with their feeble development in others, may not have been due to an acceleration of subsidence consequent upon volcanic action.

6. A review of the geological history of Britain cannot but impress the geologist with a conviction of the essential uniformity of volcanism in its manifestations since the early beginnings of geological time. The composition and structure of the materials erupted from the interior have remained with but little change. The manner in which these materials have been discharged has likewise persisted from the remotest periods. The three modern types of Vesuvian cones, puys and fissure-eruptions can be seen to have played their parts in the past as they do to-day.

Among the earliest igneous masses of which the relative geological date can be fixed are the dykes which form so striking a system among the Archæan rocks of the north-west, and show how far back the modern type of volcanic fissures and dykes can be traced. No relic, indeed, has survived of any lavas that may have flowed out from these ancient fissures, but so far as regards underground structure, the type is essentially the same as that of the Tertiary and modern Icelandic lava-fields.

The early Palæozoic volcanoes formed cones of lava and tuff comparable to those of such vents as Vesuvius and Etna. In the Lake District the pile of material ejected during Lower Silurian time was at least 8000 or 9000 feet thick. In the Old Red Sandstone basins of Central Scotland there were more than one mass of lavas and tuffs thicker than those of Vesuvius.

The puys of the later half of Palæozoic time closely resembled their Tertiary successors in Central France, the Eifel, and the Phlegræan Fields.

Nor, as regards extent and vigour, did the eruptions of the geological past differ in any important respect from those of the present time. There is assuredly no evidence that volcanic energy has gradually waned since the dawn of geological history. The latest eruptions of North-Western Europe, forming the Tertiary basalt-plateaux, far exceeded in area, and possibly also in bulk of material discharged, all the eruptions that had preceded them in the geological record.

7. Nevertheless, while the Tertiary eruptions showed no diminution of vigour, it is undoubtedly true that the volcanic energy has not manifested itself in a uniform way since the beginning of geological time. There have been periods of maximum activity followed by others of lessened force. Thus if we take a broad view of the general features of volcanic action during the Palæozoic ages in Britain, we see clear evidence of a gradual diminution in its vigour. The widespread outpourings of lava and tuff in the Silurian period in England, Wales, Scotland and Ireland were succeeded by the somewhat diminished, though still important, eruptions of the Lower Old Red Sandstone basins. The latter were followed by the still lessened outflows of the Carboniferous plateaux, which in turn were succeeded by the yet feebler and more localized eruptions of the Carboniferous puys, the whole prolonged volcanic succession ending in the small scattered vents of the Permian period. There were of course oscillations of relative energy during this history, some of the maxima and minima being of considerable moment. But though progress towards extinction was not regular and uniform, it was a dominant feature of the phenomena.

8. The Permian volcanoes were the last of the long Palæozoic series, and, so far as we yet know, the whole of the Mesozoic periods within the area of Britain were absolutely unbroken by a single volcanic eruption. The chronological value of this enormous interval of quiescence may, perhaps, never be ascertainable, but the interval must assuredly cover a large part of geological time. It was an era of geological calm, during which the Triassic, Jurassic and Cretaceous formations were slowly accumulated over the larger part of Europe. The stratigraphical quietude was not indeed unbroken. The widespread subsidence of the sea-bottom was interrupted here and there by important upheavals, and considerable geographical changes were in process of time accomplished. But, save in one or two widely separated areas of Europe, there were no active volcanoes over the whole continent.[439] Here again the scarcity or absence of intercalated volcanic rocks is in harmony with the general stratigraphy of the formations.

[Footnote 439: The Triassic eruptions of Predazzo and Monzoni were important, and traces of others are said to occur in the Cretaceous system in Portugal and Silesia.]

9. After the prodigious interval represented by the whole of the Mesozoic and the earlier part of the Tertiary formations, a time of disturbance arose once more, and the great basalt-floods of the north-west were poured forth. Evidence has been adduced in the foregoing chapters that this latest volcanic period was one of vast duration; that it was marked by long intervals of quiescence, and by repeated renewals of volcanic energy. Yet over the area of Britain the whole of its manifestations were probably comprised within the earlier (Oligocene and perhaps early Miocene) part of older Tertiary time. Since its eruptions ceased, another interval of profound quiescence has succeeded, which still continues. But this interval is almost certainly of less duration than that which elapsed between the Palæozoic and Tertiary outbursts. In other words, remote as the date of these Tertiary volcanoes appears to be from our own day, it comes much nearer to us than did the era of the last Permian eruptions to the earliest of the Tertiary series.

10. By the dissection which prolonged denudation has effected among the old volcanic centres of Britain, materials are supplied for studying the sequence of events from the beginning to the end of a volcanic period. These events have generally followed the same tolerably well-defined order.

In the case of fissure-eruptions, rents formed in the crust of the earth and communicating with the surface have allowed lava to rise and flow out above ground, either from the lips of the fissures or from vents opened along the lines of chasm. The thousands of parallel dykes in Britain remain as evidence of this mode of the ascent of the molten magma. Lines of large cones of the Vesuvian type may be presumed to have risen along guiding fissures in the terrestrial crust.

But it is evident from a study of the British examples that the existence of a fissure in the visible part of the crust is not always necessary for the production of a volcanic vent. In hundreds of instances, communication from the internal magma to the surface was effected by successive explosions, which finally blew out an orifice at the surface with no visible relation to any fissures or dykes. Of course, beneath the formations that now form the surface, and through which the necks rise, there may be lines of fault or weakness in older rocks which we cannot see. But, in what can be actually examined, vents have commonly been drilled through rocks independently of faults.

The discharge of explosive vapours was sometimes the first and only effort of volcanic energy. Generally, however, fragmentary volcanic materials were ejected, or, if the eruption was more vigorous, lava was poured out. In a vast number of cases, especially in the later ages of Palæozoic time, only ashes were projected, and cones of tuff were formed. In the earlier ages, on the other hand, there was a much larger proportion of lava expelled. Towards the close of a volcanic period, the vents were gradually choked up with the fragmentary materials that were ejected from and fell back into them. Occasionally, during the process of extinction, an explosion might still occur and clear the chimney, so as to allow of the uprise of a column of molten rock which solidified there; or the sides of the crater, as well as of the cavernous funnel underneath, fell in and filled up the passage. Heated vapours sometimes continued to ascend through the debris in the vent, and to produce on it a marked metamorphism.

There seems to have been commonly a contraction and subsidence of the materials in the vents, with a consequent dragging down or sagging of the rocks immediately outside, which are thus made to plunge steeply towards the necks.

When the vents were plugged up by the consolidation of fragmentary matter or the uprise of lava in them, the final efforts of the volcanoes led to the intrusion of sills and dykes, not only into the rocks beneath the volcanic sheets, but also, in many instances, into at least the older parts of the sheets themselves. These subterranean manifestations of volcanic action may be recognized in almost every district. They vary greatly in the degree to which they are developed. Sometimes, as in the Cader Idris, Arenig and Snowdon regions, they attain considerable importance, alike as regards the number and thickness of the sheets. In other cases, they are exhibited on so small a scale that they might be overlooked, as in the tract of Carboniferous puy-eruptions in the north of Ayrshire. But they are so generally present as to form a remarkably characteristic feature of the volcanic activity of each geological period from the earliest time to the latest. The basic sheets in the Dalradian series of Scotland display early and colossal examples. All through the successive eruptive periods of Palæozoic time, sills are found as accompaniments of superficial ejections.

The Tertiary basalt-plateaux supply numerous and gigantic examples of intruded sheets. Tertiary cones of Vesuvian type are not found in Britain, but where on the continent they have been sufficiently laid open by denudation, they present sometimes an astonishing series of sills. As a striking illustration of this structure reference may be made to the sheets of trachyte that have been injected between and have marmorized the Cretaceous strata on which Monte Venda stands, among the Euganean Hills.[440]

[Footnote 440: G. vom Rath, _Zeitsch. Deutsch. Geol. Gesellsch._, xvi. (1864), p. 461. E. Suess, _Sitzungsber. k. Akad. Wien._, lxxi. (1875), p. 7; _Antlitz der Erde_, vol. i. p. 193. E. Reyer, _Die Euganeen_, 1877. This volcano is further referred to, _postea_, p. 477.]

It is obvious that the time of intrusion of the sills cannot be precisely determined. They were not likely to be injected at an epoch when the volcanic magma could find ready egress to the surface. That they did not arise before such egress was obtained may be inferred from their petrographical characters, which are usually those of the later and not of the earlier outflows of the magma; and from the fact that they not only lie among the rocks below the volcanic series, but intersect the lower parts of that series, sometimes even the higher parts. We may therefore, with every probability, regard the sills as among the closing phases of a volcanic period.

As the lavas and tuffs of each volcanic period are intercalated among the successive geological formations, a definite beginning and end to the period are stratigraphically fixed. We see exactly where in the sedimentary series the first showers of ashes fell, and where the last mingled with the ordinary sand and mud of the sea-door. The same record shows that the volcanic accumulations were finally washed down, that they subsided with the rest of the ground around them, and that usually they were buried under overlying conformable sedimentary deposits. Thus cones of ashes and lava which may have been several thousand feet high completely disappeared.

10. A consideration of the distribution of the volcanic rocks in time shows not only how singularly uniform the course of volcanic activity has been, but that there is no evidence of the cessation of any of the broader petrographical types during geological history. Quite as much variety may be observed among the erupted materials of Tertiary time in Britain as among those of the early ages, when the earth was younger and its volcanic vigour might be supposed to have been greater and more varied than it is now. The table on the following page will make these features at once apparent. From this table it will be seen that while some of the acid rocks have not always been extruded, the basic masses have played their part in every volcanic period.

11. A study of the volcanic products of a long series of eruptions within the same geographical region may be expected to throw light on the changes that take place during the course of ages in the character of the internal molten magma. In a former chapter (vol. i. p. 27) reference was made to the subject of volcanic cycles and to the sequence, observed in various widely separated parts of the world, among the materials erupted from below. Allusion was likewise made in a later chapter (vol. i. p. 90) to the remarkable differences in texture and composition noticeable within some large bodies of eruptive material, and to the evidence which these differences furnish of a segregation or differentiation among the constituents of an eruptive mass after it has been injected into its position within the crust of the earth.

Table of the Periods of Volcanic Action in the British Isles and of the Chronological Distribution of the Volcanic Products.

Key to Columns ==================================== Gr = Granites, Granophyres, etc. Fe = Felsites, Rhyolites, etc. Da = Dacite, "Pitchstone" of Eigg. Tr = Trachytes. An = Andesites (Porphyrites). Ga = Gabbros. Do = Dolerites, Basalts (Diabases). Pi = Picrites and highly basic lavas. Tu = Tuffs, acid or basic.

+--------------------+-----+-----+-----+-----+-----+-----+-----+-----+-----+ | | Gr | Fe | Da | Tr | An | Ga | Do | Pi | Tu | +--------------------+-----+-----+-----+-----+-----+-----+-----+-----+-----+ | Older Tertiary | | | | | | | | | | | (Plateaux, dykes, | | | | | | | | | | | necks, bosses, | | | | | | | | | | | sills) | * | * | * | * | * | * | * | * | * | | | | | | | | | | | | | Mesozoic | | | | | | | | | | | No volcanic rocks.| | | | | | | | | | | | | | | | | | | | | | Permian | ··· | * | ··· | ··· | * | ··· | * | * | * | | | | | | | | | | | | | Carboniferous | ? | | | | | | | | | | Puy type | ··· | * | ··· | ··· | * | ··· | * | * | * | | Plateau type | ··· | * | ··· | * | * | ··· | * | * | * | | | | | | | | | | | | |{Devonian | ··· | ··· | ··· | ··· | ··· | ··· | * | ···| * | |{ | | | | | | | | | | |{Old Red Sandstone | | | | | | | | | | |{ Upper | ··· | ··· | ··· | ··· | ··· | ··· | * | ···| * | |{ Lower | * | * | ··· | * | * | ··· | * | ···| * | | | | | | | | | | | | | Silurian | | | | | | | | | | | Upper | ··· | * | ··· | ··· | ··· | ··· | ··· | ···| * | | Lower, Bala | * | * | ··· | * | * | * | * | * | * | | " Arenig | * | * | ··· | * | * | * | * | ···| * | | | | | | | | | | | | | Cambrian | ··· | * | ··· | ··· | * | ··· | * | ···| * | | | | | | | | | | | | | Uriconian | ··· | * | ··· | ··· | ··· | ··· | * | ···| * | | | | | | | | | | | | | Dalradian | ··· | ··· | ··· | ··· | ··· | ··· | * | ···| ? | | | | | | | | | | | | | Torridonian | | | | | | | | | | | | | | | | | | | | | | Lewisian | * | ··· | ··· | ··· | ··· | ··· | * | * | ··· | | | | | | | | | | | | +--------------------+-----+-----+-----+-----+-----+-----+-----+-----+-----+

From the history of volcanic action in the British Isles it is clear that differentiation is effected under three distinct conditions.

In the first place, a notable difference may be occasionally observed between two adjacent parts of the same mass of lava which has flowed out at the surface. Thus, in the Carboniferous picrite of Blackburn, there has been a separation of the heavy basic constituents, which have in great part settled down into the lower part of the sheet, while the lighter felspar has mainly come to the top. In this case the gradual transition from top to bottom suggests that the separation occurred after the lava had reached the surface and taken the form of a stream or sheet.

In the second place, segregation has taken place in the magma within the terrestrial crust after intrusion, for it is frequently observable in large bosses and sometimes in sills, the basic elements having tended to mass themselves towards the margins of the rock, leaving more acid material in the centre. The cases of Garabol Hill among the Dalradian schists of Scotland, of Carrock Fell among the Silurian strata of the Lake District, and of the Cramond picrite among the Carboniferous formations of Midlothian, with others that might be cited from various other regions and geological formations in Britain, prove to what a considerable extent a separation of ingredients may take place in a boss, and even sometimes in a comparatively thin sill before the molten mass consolidates.

In the third place, there is good evidence that already before the magma is either intruded or extruded, and while it still lies within the internal reservoir, it may not possess a general uniformity of composition, but may have become more or less heterogeneous. In regard to intrusive rocks, the extraordinarily banded gabbros of the Tertiary series of Skye obviously proceeded from a magma in which the molten material consisted in some parts mainly of felspar, and in others mainly of the ferro-magnesian minerals and iron-ores. Streams from these differently constituted parts of the magma were simultaneously or successively injected as sills into the older portions of the volcanic series, while, as the process of differentiation within the magma proceeded, still more felspathic liquid was left behind, to be thrust into cracks in the sills previously consolidated.

Moreover, the banded basalts of the Tertiary plateaux show that this heterogeneity was not confined to internal intrusions, but maintained its place even when the molten material was ejected to the surface. The differentiation indeed is not so striking there as among the sills of gabbro; but its presence, even in a less degree, proves that the separation of constituent minerals was not due to any general cooling of an erupted body of igneous rock, but was already developed in the reservoir from which the molten material was propelled to the surface.

Attention has been called to the remarkable similarity of structure between these banded intrusive rocks and some of the ancient gneisses. The resemblance is so close that we may with every probability infer that the gneisses acquired their characteristic banding as intrusive masses of igneous rocks, discharged from heterogeneous magmas, like that which supplied the gabbros of the Cuillin Hills. And as these gneisses belong to pre-Cambrian formations, we are thus led to the interesting result that the tendency to develop heterogeneity was already as characteristic of the magma-basins of the earliest geological time as it has been of those of later periods.

The evidence of differentiation presented by superficial lavas, and by intrusive sills and bosses, acquires great interest when considered in connection with the changes which are seen to have occurred in the character of the materials erupted during the course of a definite volcanic period. An attentive examination of the volcanic products of the various ages, so fully recorded in the geological structure of the British Isles, shows that a recognizable sequence in the nature of the materials erupted during a single volcanic period can be traced from the earliest to the latest times, and that, in spite of occasional departures, the normal order remains broadly uniform.

With the important exception of the Snowdonian region and possibly others, we find that the earlier eruptions of each period were generally most basic, and that the later intrusions were most acid. Thus the diabase-lavas and tuffs at the base of the Cambrian series of St. David's are pierced by quartz-porphyry veins. The andesites of the Lower Old Red Sandstone were succeeded by bosses, sills, and dykes of granite, felsite, and lamprophyre. The eruptions of the Carboniferous plateaux began with extremely basic lavas, and ended with trachytes, felsites, and quartz-porphyries. The basalts of the great lava-fields of the Tertiary period are pierced by masses of granophyre and even granite.

There has evidently been, on the whole, a progressive diminution in the quantity of bases and a corresponding increase in the proportion of acid in the lavas erupted during the lapse of one volcanic period. This sequence is so well marked and so common that it cannot be merely accidental. The acid and basic rocks, occurring as they do at each volcanic centre in the same relation to each other, are obviously parts of one connected series of eruptions. We seem to see in this sequence an indication of what was taking place within the subterranean magma. There was first an extensive separation of the more basic constituents, such as the ferro-magnesian minerals and ores, and the lavas which came off at that time were heavy and basic basalts, and even picrites. The removal of these elements left the magma more acid, and such rocks as andesites were poured out, until at last the deeper intrusive sills, dykes and bosses became thoroughly acid rocks, such as felsite, quartz-porphyry and granite, while if any superficial outflow took place it was such a rock as dacite.

In the case of the Tertiary volcanic series there is evidence that after the acid protrusions a final uprise of basic material occurred. No satisfactory proof of any similar return to basic eruptions has been detected among the Palæozoic formations. But it is possible that some of the basic sills and dykes, the precise age of which cannot be fixed, may really mark such a reversion, even in the earlier volcanic periods.

Some illustrative examples of volcanic cycles from other countries were cited in Chapter iii. To these I may add another instance which presents a close analogy to some of the phenomena characteristic of the British examples of Palæozoic as well as of Tertiary age. Monte Venda in the Euganean Hills, already alluded to (p. 474), may be cited as an interesting specimen of an older Tertiary volcano, which has been so dissected by denudation as to show not only the succession of its superficial discharges, but the position and order of its subterranean intrusions. The volcanic eruptions of this neighbourhood, judging from the area which they still cover and the height they reach, may have piled up a mountain rivalling or surpassing Etna in dimensions. In Monte Venda the lowest visible igneous rocks are sills of oligoclase-trachyte that have been thrust between and have highly altered Cretaceous (Tithonian) limestones. Other intrusive sheets of trachyte follow in the overlying Cretaceous strata (Neocomian and _Scaglia_). It is not until the older Tertiary formations are reached that undoubted tuffs and lavas occur, indicative of truly interstratified volcanic materials. These formations, consisting of nummulitic limestones and other strata together with fossiliferous tuffs, show that the volcano began as a submarine vent. It discharged dark basic dolerites and tuffs. The highest lava, however, crowning the summit of the mountain is a trachyte. There appears to have been a rapid decrease of the bases in the magma, for the later lavas were rhyolites, accompanied with rhyolitic tuffs of Oligocene age, and followed in the end by the black vitreous trachyte of Monte Sieva.

12. From the evidence detailed in these volumes, it appears that the sequence from basic to acid discharges was on the whole characteristic of each eruptive period. It is obvious, however, that as the protrusions of successive periods took place within the same limited geographical area, the internal magma during the interval between two such periods must in some way have been renewed as regards its constitution, for when, after long quiescence, eruptions began once more, basic lavas appeared first and were eventually followed by acid kinds. This cycle of transformation is admirably exhibited in Central Scotland, where the andesites of the Old Red Sandstone with their felsite sills are followed by the limburgites, picrites and other highly basic lavas at the bottom of the Carboniferous plateaux, succeeded in turn by the andesites, trachytes and acid sills of that series. When the puy eruptions ensued, the magma had once more become decidedly basic.

That the true explanation of these alterations is of a complex order may be inferred from the exceptions which occur to the general rule. I have alluded to the Snowdon region, where the acid rhyolites are followed by more basic andesites, and where the sills are also more basic than the superficial lavas. In the Arenig and Cader Idris country the sills are likewise more basic than the bedded lavas. Among the Carboniferous puys of the basin of the Firth of Forth, the sills are not sensibly more acid than many of the superficial basalts, and they even include such rocks as picrite. Possibly in this last-named region we see an arrested sequence, the volcanic protrusions having from some cause ceased before the general uprise of the more acid magma.

INDEX

Aa form of lava in the Sandwich Islands, ii. 187 Abereiddy Bay, i. 206 Abich, H., i. 32 Acid igneous rocks, silica percentage of, i. 14; devitrification of, 19; flow-structure of, 21; occur in thicker sheets than basic, 24; alternations of, with basic, 28, 61, 152, 157, 165, 207, 213, 233, 284, 318; ii. 236, 266, 278; metamorphic action of, i. 95, 96; connection with mountains, ii. 98; scenery of, 102. Acids, mineral, at volcanoes, i. 72 Acland, Mr. H. D., i. 133 Aegean Sea, volcanoes of, i. 1 Agglomerates, i. 31, 57, 58; in dykes, 70; Archæan, 120, 130, 135; Cambrian, 148, 149, 167; Silurian, 178, 180, 181, 184, 185, 194, 199, 206, 214, 237, 241, 244, 247, 253, 255; Old Red Sandstone, 279, 285, 289, 300, 313, 325, 338, 349, 352; Carboniferous, 381, 399, 402, 404, 427, 429, 439, 440; ii. 13, 24, 28, 29; Permian, 62, 64, 99; Tertiary, 194, 277, 278, 281, 289, 292, 293, 384, 400, 423 Allan, T., i. 363 Allotriomorphic minerals, i. 21 Allport, Mr., i. 95, 130, 131, 260, 451; ii. 11, 42, 102, 103, 104, 106, 370 Amber in Tertiary volcanic series, ii. 198 America, Western North, volcanic rocks of, i. 10, 100; ii, 267 Amygdales, origin of, i. 15; ii. 189, 221, 285, 290 Amygdaloidal structure, i. 15, 16, 17, 59, 274, 385; ii. 3, 31, 57, 129, 188 Analyses of Cambrian tuffs, i. 148, 149; of Cambrian diabases, 153; of Old Red Sandstone diabases, 274; of Old Red Sandstone andesites, 275; of Old Red Sandstone trachytes, 276; of Old Red Sandstone felsites, 278; of Carboniferous limburgite, 377; of Carboniferous basalts, 379; of Carboniferous trachytes, 380; of Carboniferous phonolite, 381; of Tertiary trachyte, ii. 139; of Tertiary dacite, 244 Anderson, Dr. Tempest, ii 261, 262, 263 Andesite, i. 24, 131, 136, 164, 165, 167, 178, 180, 184, 189, 190, 204, 212, 213, 214, 215, 229, 230, 245, 246, 247, 252, 274, 275, (analyses), 277, 292, 300, 306, 309, 315, 318, 325, 330, 333, 345, 377, 379, 386, 403, 421; ii. 45, 57, 96, 125, 137, 184, 236, 424 Anglesey, gneisses and schists of, i. 126; volcanic rocks of, 189, 219 Anhydrite deposits, ii. 54 Annandale, Permian volcanic rocks of, ii. 56, 58, 60, 61, 66 Antrim, Old Red Sandstone volcanic rocks of, i. 314; Tertiary volcanic rocks of, 47, 52; ii. 109, 110, 113, 139, 140, 199; basalts of, 192, 193, 199, 202, 206; clays and iron-ore of, 204; rhyolites of, 185, 364, 370, 371, 426, 445; deceptive agglomerate of, 188; rhyolitic conglomerate of, 195, 206; plateau of, 199; tuffs of, 202, 204; vents of, 271, 277; sills of, 298; central subsidence of basalt-plateau of, 448 Apatite, ii. 135 Apjohn, J., ii. 42 Applecross, volcanic vents in, ii. 292 Arans, the, i. 175, 176, 179, 184, 186, 207 Archæan period, i. 110, 111; volcanic rocks of, 120 Ardnamurchan, dykes and veins of, ii. 154, 320; basalt-plateau of, 208; vents of, 287; sills of, 318; gabbro of, 355 Arenig group, i. 175; lower limit of, 177, 185; top of, 178, 228, 246 ---- volcano of, i. 42, 175, 176, 179, 186, 207 ---- rocks in Scottish Highlands, i. 123, 126; in Merionethshire, 176, 179; of Shropshire, 189; of Ayrshire, 196; of Scottish Highlands, 201; of Anglesey, 221; of Lake district, 229; of Ireland, 239 Argyll, Duke of, ii. 113, 114, 198 Argyllshire, dykes of, ii. 127, 128, 138, 142, 146, 171, 172; vents of, 278 Arizona, explosion crater in, i. 58; laccolites in, 86 Arran, Old Red Sandstone volcanic rocks of, i. 298, 311; Carboniferous volcanic rocks of, 386, 392; possible Permian volcanic rocks of, ii. 58; granite of, i. 93; ii. 366, 367, 418; pitchstone of, i. 19; ii. 445; dykes of, 123, 139, 140, 142, 146, 154, 161 Arthur Seat, i. 364, 373, 378, 385, 386; ii. 67 "Arvonian," i. 145, 156 Asbestos in volcanic breccia, ii. 51 Ascension Island, cellular lava of, i. 15 Ashes, volcanic (_see_ Tuffs) Ashprington volcanic series, i. 262 Asphalt, ii. 79 Atherstone, i. 170 Augite, loose crystals of, in volcanic vents, i. 62, 178, 181; ii. 58, 79; lumps of, in volcanic vents, i. 352 Augite-aphanites, i. 178 Auvergne, old volcanoes of, i. 29, 32, 66, 70, 100; ii. 373 Aveline, Mr. W. T., i. 227, 230; ii. 32 Ayrshire, example of volcanic neck in, i. 56; Silurian volcanic rocks of, 192; Old Red Sandstone volcanic rocks of, 275, 282, 283, 285, 291, 331; Carboniferous volcanic plateau of, 102, 368, 388, 393, 398, 410; Carboniferous Puys of, 415, 416, 434, 440, 474; Permian volcanic rocks of, ii. 55, 58, 62 Azoic period, i. 109

Bäckström, Mr., ii. 266 Baily, W. H., i. 251, 252; ii. 198, 449 Bala group, i. 175, 190, 196, 201, 206, 207, 223, 242; limestone of, 47, 175, 229, 245, 251; volcanic rocks of, 186, 190, 207, 213, 221, 241, 248 Balbriggan, igneous rocks of, i. 244 Ballagan beds (Lower Carboniferous), i. 384, 387, 392, 393, 412, 447 Ballantrae, volcanic rocks at, i. 192, 199 Ballypallidy, tuffs and leaf-beds of, ii. 204, 429 Bamborough, Whin Sill at, ii. 2, 3, 5 Banding of igneous rocks, i. 84, 207; ii. 189, 294, 329, 354, 357, 476 ---- of gneiss, i. 116 Bangor group, i. 166 Banks, Sir Joseph, ii. 109 Barnavave, eruptive rocks of, ii. 421 Barrow, Mr. G., i. 201, 226, 272, 279, 380; ii. 147, 148 Basalt, columnar structure of, i. 24, 25; relation to gabbro, 78; altered by carbonaceous strata, 95; shells supposed to occur in, ii. 110; banded, 189; thickness of sheets of, 192; meaning of red layer between sheets of, 197, 203, 206, 254; metamorphism of, 272, 276, 337, 339, 340, 347, 355, 356, 357, 358, 362, 378, 383, 386, 397, 399, 400, 404, 413

---- pre-Cambrian, i. 119, 131; Silurian, 206, 207, 230, 245; Carboniferous, 378, 403, 407, 417; ii. 11, 45, 46; Permian, 57, 96; Tertiary, 125, 136, 183, 199, 208, 291

Basalt-conglomerate, ii. 195 Basic volcanic rocks, silica-percentage of, i. 14; devitrification of, 20; flow-structure of, 21; occur in thinner sheets than the acid, 24; metamorphic action of, 94; erupted at low levels, 98; scenery of, 102; converted into schists by deformation, 75, 114, 118, 119, 124, 129; alternation with acid, 28, 61, 131, 157, 165, 207, 213, 233, 284, 318;