CHAPTER XL.

ON THE ROCKS ATTRIBUTED TO VOLCANIC AGENCY TAKING

PLACE UNDER CIRCUMSTANCES DIFFERENT FROM THOSE BEFORE CONSIDERED.

Trap rocks-their general characters—their structure--prismatic-sphe

roidal — tabular-dykes.-Wernerian theory with regard to trap.-Arguments in favour of the aqueous origin of basalt-shown to be fallacious.—Differences between lavas and basalts explained.—Effects of heat modified by pressure.—Why submarine lavas cool slowly-causes that give rise to vitreous products—lamination of igneous products accounted for.—Process of devitrification accounted for by slow cooling-still slower cooling may produce basalt and other traps.—Why submarine lavas have cooled slowly.—Prismatic structure of trap accounted for, and greater frequency of dykes.— Trap rocks, at what periods formed.Three classes of volcanic products—their characters stated.—Distinctions

between plutonic and volcanic rocks. Having now concluded my intended sketch of the phænomena of existing volcanos, and attempted to explain the causes from which they originate, I shall consider in the next place, the influence which forces of the same description may have exerted in former times on the condition of our planet, the rocks that have been produced by their operations, or which have been altered in character and position by their agency.

It is this part of the inquiry, which connects the subject of volcanos with the other investigations of geology, and renders their study of interest, not merely to the chemist and the natural philosopher, but likewise to all who would attempt to explain the condition, past or present, of the globe we inhabit. I shall therefore proceed to notice the rocks, which, though differing in some respects from those produced by volcanos at the present day, appear, nevertheless, to be derived from the same cause acting under somewhat altered circumstances.

There are few parts of the world that do not offer examples of those Formations, which are comprehended by geologists under the name of Trap, including, as it does, in its most extensive signification, on the one hand basalts, greenstones, syenites, and wacke, and on the other porphyries, with base of felspar or claystone. To each of these general subdivisions are annexed sundry mechanical aggregates, in which pebbles or angular fragments of the rocks above-mentioned constitute the prevailing ingredients.

The above rocks have this peculiarity belonging to them, that they occur in connexion with all the formations which compose the crust of the globe, from the oldest to the most modern, resting on them in irregular tabular masses, occasionally alternating with, and still more commonly intersecting them at various angles.

When circumstanced in either of the two former ways with reference to the accompanying strata, they have been denominated beds, with what propriety will afterwards appear; when disposed in the latter way they are called trap-veins or whin-dykes ; whin being a provincial term, originally employed by the colliers in Northumberland to designate any hard stone, but now introduced into the general language of geology, for the purpose of indicating a rock, consisting of basalt, greenstone, or wacke, traversing the strata in the manner that has been represented.

I will consider, in the first place, the general structure of trap rocks, and afterwards those circumstances which may be regarded as peculiar, either to the one or other of the forms in which it is found.

Trap rocks, in some one of their different conditions, present examples probably of every kind of structure which has elsewhere been observed : examined on the small scale, we remark them amygdaloidal, porphyritic and granular; examined on the large, we find them in some instances slaty or fissile, as clinkstone; in others divided into thick tabular masses, as basalt and greenstone frequently are.

But that which peculiarly distinguishes rocks of the Trap family, is the tendency to split into prismatic, or, speaking more generally, polyhedral masses, which structure, though it exists likewise in granitic and a few other species of rocks, is nowhere so frequent, or so well displayed as in these. The columns vary in the number of their sides from three to six, seven, and even twelve; they are more generally straight, but not unfrequently curved ; in size they may be said to range from an inch to nine feet in breadth, and from a foot to 300 or more in height. They are sometimes continuous for a considerable space; but at other times are ob. liquely and irregularly divided by fissures or joints, the convex surface of one being inserted into a corresponding concavity of the next.

The columns are usually at right angles to the direction of the bed, but not always so; occasionally indeed they radiate from a central point, forming clusters without any determinate direction, and still more commonly they are placed so irregularly as to interfere one with the other. Sometimes in the same bed, one portion will be prismatic and the rest amorphous; whilst every intermediate condition, from that of jointed columns possessing an almost architectural regularity, to a total absence of all arrangement, will be perceived.

It has been usual to refer this kind of structure to the contraction which the mass underwent during its cooling down from a melted state ; and there is no doubt that a prismatic structure may arise from a cause of this kind, as we see exemplified in many modern lavas*, and in the shrinking of masses of clay, starch, &c. But there is one circumstance which seems to prove that the prismatic form of trap is owing to a different cause; namely, that in many cases the columns approximate so nearly, that not even the blade of a knife can be thrust in between them. Now in every instance in which the same kind of structure is produced by contraction, theory suggests, and experience confirms, the conclusion, that a certain interval would be left between the columnar masses so produced.

We must therefore look to some other cause for the columnar arrangement of trap, and probably the true solution will be suggested to us, after considering another kind of structure noticed as existing in these rocks, namely the spheroidal or globular. In this case the rock is either wholly or in part arranged in balls of various magnitudes. The globular form

* Contrast, for instance, the structure of the lava of Niedermennig mentioned in my fourth chapter with that of the columns of the Giant's Causeway.

a

a

is very conspicuous in the rock of the Shiant Islands; but according to Dr. Macculloch does not appear to be common. A tendency however to this structure is betrayed in most trap rocks by the manner in which they disintegrate, those even which are columnar exfoliating into spheroidal forms when exposed to the weather. Now it is evident that a series of globular concretions of trap, placed in close contact, whilst in a pasty condition, or in the state of transition from fusion to solidity, would be by mutual pressure converted into a succession of jointed columns, which, owing to slight differences in the compactness and consequent softness of the several parts of the mass, would rarely be exact in their sizes and in the number of their sides, but would exhibit all those variations which, in that respect, columnar basalt commonly displays*. Neither does it follow that they may not in some cases have shrunk, after the prismatic form has been communicated to them by mutual compression, since the latter force would begin to operate from the moment they ceased to be liquid; whereas the tendency to contract would continue up to the time at which the rock had sunk to the temperature of the bodies surrounding it.

I conceive therefore that the spheroidal structure will be found to be the one most prevalent in rocks of the trap family, and that the prismatic is in general only a consequence of it; the former indeed arising from that species of molecular attraction, which begins to display itself in all melted bodies from the moment they cease to be absolutely fluid up to the time at which they become completely solid. Hence the longer the interval between these two points, the more fully does this disposition operate, as has been shown by Mr. Gregory Wattt and others, who caused the particles of glass, and even of lava, to arrange themselves in spheroidal concretions, by allowing them, after being melted, to return to a state of solidity with sufficient slowness.

Having now considered the general structure of trap rocks,

This and the other facts which bear out my hypotheses are beautifully illustrated in the case of the Kase-Keller near Bertrich, described in my 79th page.

+ Phil. Trans. 1804.

I will next examine the peculiarities belonging to either of the two conditions in which they exist.

One of the most common forms in which the harder varieties of trap are found, is in large overlying masses, sometimes rising into high mountains, but more generally capping the summits of hills of comparatively low elevation. These latter sometimes would seem to indicate stratification; but the appearance they present is owing to their division into large flat tables, which again have a tendency to decompose, in an abrupt manner, at right angles to the seams of the stratification, thus exhibiting a series of mural precipices, ranging one above the other, from which the term trap, which in Swedish signifies a stair, has been applied to them.

In other cases they appear to alternate with the rocks of the country, but this appearance is most frequently, though not always, deceptive. Dr. Macculloch has shown that many veins of trap put on a form so far parallel to the stratification, as, when partially viewed, to possess the semblance of beds *. Their true nature may in these cases be determined by finding that the parallelism is not long maintained, but that any one such supposed stratum quits its place to intersect the adjoining and including stratified rock, or sends ramifications through the whole series.

In a few cases, where deep sections of cliffs afford opportunities for examination, it is found that irregular masses lie beneath the stratified rocks in some places, just as they surmount them in others; and that, from these also, veins proceed to the surface, or in other directions.

Without therefore denying that alternations of trap rocks with neptunian deposits may occur,-a consequence indeed which would necessarily ensue, if successive formations of the former rock had taken place at the bottom of water, whilst the latter was in the act of throwing down deposits of clay, limestone, or sand,-let us go on to consider the case of veins or dykes, to which class the great majority probably of stratiform masses of trap actually belong.

These dykes occur of all sizes, from a few inches to twenty or thirty yards in thickness. They extend in some cases many miles in length, as in the case of the great Cleaveland