While Spallanzani was engaged in investigating the nature of the action going on at Stromboli and other Italian volcanoes, his contemporary Dolomieu was laying the foundation of another important branch of vulcanology by studying the characters of the different materials of which volcanoes are built up. Since the publication of Dolomieu's admirable works on the rocks of the Lipari and Ponza Islands, science has advanced with prodigious strides. The chemist has taught us how to split up a rock into its constituent elements and to determine the proportions of these to one another with mathematical precision; the mineralogist has done much in the investigation of the characters and mode of origin of the crystalline minerals which occur in these rocks; and the microscopist has shown how the minute internal structure of these rocks may be made clearly manifest. We shall proceed to give a sketch of the present state of knowledge obtained by these different kinds of investigations, concerning the materials which are ejected from volcanic vents.

The most abundant of the substances which are ejected from volcanoes is steam or water-gas, which, as we have seen, issues in prodigious quantities during every eruption. But with the steam a great number of other volatile materials frequently make their appearance. The chief among these are the add gases known as hydrochloric acid, sulphurous acid, sulphuretted hydrogen, carbonic add, and boracic acid; and with these acid gases there issue hydrogen, nitrogen, ammonia, the volatile metals arsenic, antimony, and mercury, and some other substances. In considering the nature of the products which issue from volcanic fissures, it must be remembered that many substances which under ordinary circumstances do not exhibit marked volatility are nevertheless easily carried away in fine particles when a current of steam is passed over them. As we shall have to point out in the sequel, different volatile substances have a tendency to make their appearance at volcanic vents according as the intensity of the action going on within it varies.

The volatile substances issuing from volcanic fissures at high temperatures react upon one another, and many new compounds are thus formed. We have already seen how, by the action of sulphurous acid and sulphuretted hydrogen on each other, the sulphur so common in volcanic districts has been separated and deposited. The hydrochloric acid acts very energetically on the rocks around the vents, uniting with the iron in them to form the yellow ferric-chloride. The rocks all round a volcanic vent are not unfrequently found coated with this yellow substance, which is almost always mistaken by casual observers for sulphur. In many volcanoes the constant passage through the rocks of the various acid gases has caused nearly the whole of the iron, lime, and alkaline materials of the rocks to be converted into soluble compounds known as sulphates, chlorides, carbonates, and borates; and, on the removal of these by the rain, there remains a white, powdery substance, resembling chalk in outward appearance, but composed of almost pure silica. There are certain cases in which travellers have visited volcanic islands where chemical action of this kind has gone on to such an extent, that they have been led to describe the islands as composed entirely of chalk.

Some of the substances issuing from volcanic vents, such as hydrogen and sulphuretted hydrogen, are inflammable, and when they issue at a high temperature, these gases burst into flame the moment that they come into contact with the air. Hence, when volcanic fissures axe watched at night, faint lambent flames are frequently seen playing over them, and sometimes these flames are brilliantly coloured, through the presence of small quantities of certain metallic oxides. Such volcanic flames, however, are scarcely ever strongly luminous and, as we have already seen, the red, glowing light which is observed over volcanic mountains in eruption is due to quite another cause. The study by the aid of the spectroscope of the flames which issue from volcanic vents promises to throw much new light on the rarer materials ejected by volcanoes. Spectroscopic observations of this kind have already been commenced by Janssen, at Stromboli and Santorin.

Some of the volatile substances issuing from volcanic vents, are at once deposited when they come in contact with the cool atmosphere, others form new compounds with one another and the constituents of the atmosphere, while others again attack the materials of the surrounding rocks and form fresh chemical compounds with some of their ingredients. Thus, there are continually accumulating on the sides and lips of volcanic fissures deposits of sulphates, chlorides, and borates of the alkalies and alkaline earths, with sal-ammoniac, sulphur, and the oxides and sulphides of certain metals. The lips of the fissures from which steam and acid gases issue in volcanoes are constantly seen to be coated with yellow and reddish-brown incrustations, consisting of mixtures, in varying proportions, of these different materials, and these sometimes assume the form of stalactites and pendent masses.

Some of these products of volcanic action are of considerable commercial value. At Vulcano regular chemical works have been established in the crater of the volcano, by an enterprising Scotch firm, a great number of workmen being engaged in collecting the materials which are deposited around the fissures, and are renewed by the volcanic action almost as soon as they are removed. In fig. 6, I have given a sketch of this singular spot, taken from the high ground of the neighbouring Island of Lipari. From the village at the foot of the volcano, where the workmen live, a zig-zag road has been constructed leading up the side, and down into the crater of the volcano. On this road, workmen and mules, laden with the various volcanic materials, may be seen constantly passing up and down.

Vulcano appears to have been frequently in a state of violent eruption during the past 2,000 years—the last great outburst having taken place in 1786. In 1873 the activity in the crater of Vulcano suddenly became more pronounced in character, and the workmen hastened to escape from the dangerous spot, but, before they could do so, several of them were severely injured by the explosions. After this outburst, which did not prove to be of very violent character, the quantity of gases issuing from the fissures in the crater was for a time much greater than before, and the productiveness of these great natural chemical works was proportionately increased: but eventually the action died out almost entirely. The chief products of Vulcano which are of commercial value, are sal-ammoniac, sulphur, and boracic acid. At one time it was even contemplated that great leaden chambers should be erected over the principal fissures at the bottom of the crater of Vulcano, in which chambers the volatile materials might be condensed and collected. The change in the condition of the volcano has unfortunately prevented the carrying out of this bold project.

Besides the volatile substances which issue from volcanic vents, mingling with the atmosphere or condensing upon their sides, there are also many solid materials ejected, and these may accumulate around the orifices, till they build up mountains of vast dimensions, like Etna, Teneriffe, and Chimborazo. Some of these solid materials are evidently fragments of the rock-masses, through which the volcanic fissure has been rent; these fragments have been carried upwards by the force of the steam-blast and scattered over the sides of the volcano. But the principal portion of the solid materials ejected from volcanic orifices consists of matter which has been extruded from sources far beneath the surface, in a highly-heated and fluid or semi-fluid condition.

The fragments torn from the sides of volcanic fissures consist of the rocks through which the eruptive forces may happen to have opened their way; pieces of sandstone, limestone, slate, granite, &c., are thus frequently found in considerable numbers among materials which build up volcanic mountains. Thus, some of the volcanic cones in the Eifel are very largely made up of fragments of slate, which have been torn from the sides of the vents by the uprushing currents of steam. At Vesuvius masses of limestone are frequently ejected, and may be picked up all over the slopes of the mountains. These limestone-fragments frequently contain fossils, and Professor Guiscardi, of Naples, has been able to collect several hundred species of shells, transported thus by volcanic action from the rock-masses which form the foundation of the volcano of Vesuvius. The action of water at a high temperature, and under such enormous pressure as must exist beneath volcanic mountains, has often produced changes in the rocks of which fragments are ejected from volcanic vents. The so-called 'lava' ornaments, which are so extensively sold at Naples, are not made from the materials to which geologists apply that name, but from the fragments of altered limestone that have been torn from the rocks beneath the mountain, and scattered by the eruptive forces all over its sides. The chemical action of the superheated and highly-compressed steam on the rocks beneath volcanoes frequently results in the formation of beautifully crystallised minerals. Such crystallised minerals abound in the rock-fragments scattered over the sides of Vesuvius and other volcanoes, both active and extinct. They have been formed in the great chemical laboratories which exist beneath the volcano, and have been brought to the surface by the action of the steam-jets issuing from its fissures.

Of still greater interest are those materials which issue from volcanic orifices in an incandescent, and often in a molten, condition, and which are evidently derived from sources far below the earth's surface. It is to these materials that the name of 'lavas' is properly applied.

Lavas present a general resemblance to the slags and clinkers which are formed in our furnaces and brick-kilns, and consist, like them, of various stony substances which have been more or less perfectly fused. When we come to study the chemical composition and the microscopical structure of lavas, however, we shall find that there are many respects in which they differ entirely from these artificial products.

Let us first consider the facts which are taught us concerning the nature and origin of lavas, by a chemical analysis of them.

Of the sixty-five or seventy chemical elements, only a very small number occur at all commonly in lavas. Eight elements, indeed, make up the great mass of all lavas—these are oxygen, silicon, aluminium, magnesium, calcium, iron, sodium, and potassium. But even these eight elements are present in very unequal proportions. Oxygen makes up nearly one-half the weight of all lavas. Almost all the other elements found in lavas exist in combination with oxygen, so that lavas consist entirely of what chemists call 'oxides.' This is a most remarkable circumstance, which, as we shall presently see, is of great significance. The metalloid silicon makes up about one-fourth of the weight of most lavas, and the metal aluminium about one-tenth. The other five elements vary greatly in their relative proportions in different lavas.

In all lavas the substance which forms the greatest part of the mass is the compound of oxygen and silicon, known as silica or silicic acid. In its pure form, this substance is familiar to us as quartz, or rock-crystal and flint. Silica is present in all lavas in proportions which vary from one-half to four-fifths of the whole mass. Now, this substance, silica, has the property of forming more complex compounds by uniting with the other oxides present in lavas—namely, the oxides of aluminium, magnesium, calcium, iron, potassium, and sodium. Silica is called by chemists an acid, the other oxides in lavas are termed bases, and the compounds of silica with the bases are known as silicates. Hence we see that lavas are composed of a number of different silicates—the silicates of aluminium, magnesium, calcium, iron, potassium, and sodium.

The above statements will perhaps be made clearer by the accompanying table from which it will be seen that lavas are compounds in varying proportions of six kinds of salts—namely, the silicates of alumina, magnesia, lime, iron, potash, and soda.

Composition of Lavas.

Now, in some lavas the acid constituent, or silica, is present in much larger proportions than in others. Those lavas with a large proportion of silica are called 'acid lavas,' those with a lower percentage of silica, and therefore a higher proportion of the bases, are known as the 'basic lavas.' It is convenient to employ the term 'intermediate lavas' for those in which the proportion of silica is lower than in the acid lavas, and the proportion of the bases is lower than in the basic lavas.

The acid lavas contain from 66 to 80 per cent, of silica; they are poor in lime, magnesia, and oxide of iron, but rich in potash and soda. The basic lavas contain from 45 to 55 per cent, of silica; they are rich in magnesia, lime, and oxide of iron, but poor in soda and potash. In the intermediate lavas the proportion of silica varies from 55 to 66 per cent.

As the basic-lavas contain a larger proportion of oxide of iron and other heavy oxides than the acid-lavas, the former have usually a higher specific gravity than the latter; it is, indeed, possible in most cases to distinguish between these different varieties by simply weighing them in water and in air.

The basic lavas are usually of much darker colour than the add lavas—the terms acid lavas, intermediate lavas, and basic lavas correspond indeed pretty closely with the names trachytes, greystones and basalt, which were given to the varieties of lavas by the older writers on volcanoes, at a time when their chemical constitution had not been accurately studied. Fresh lavas of acid composition are usually nearly white in colour, intermediate lavas are of various tints of grey, and basic lavas nearly black. It must be remembered, however, that colour is one of the least persistent, and therefore one of the least valuable, characters by means of which rocks can be discriminated, and also that by exposure to the influence of the atmospheric moisture the iron present in all lavas is affected, and the lavas belonging to all classes, when weathered, assume reddish and reddish-brown tints.

Geologists have devised a great number of names for the various kinds of lava which have been found occurring round volcanic vents in different parts of the world, and the study of these varieties is full of interest. For our present purpose, however, it will be sufficient to state that they nearly all fall into five great groups, known as the Rhyolites, the Trachytes, the Andesites, the Phonolites, and the Basalts. The Rhyolites are acid lavas, the Basalts are basic lavas, and the Trachytes, Andesites, and Phonolites, different kinds of intermediate lavas, distinguished by the particular minerals which they contain.

Before we part from this subject of the classification of lavas according to their chemical composition, it will be well to point out that there exists a small group of lavas which stand quite by themselves, and cannot be referred to either of the classes we have indicated. They contain a smaller proportion of silica, and a much larger proportion of magnesia and oxide of iron than the other lavas, and may be made to constitute a small sub-group, to which we may apply the term of 'ultra-basic lavas.' Although much less widely distributed than the other varieties, they are, in some respects, as we shall presently have to point out, of far greater interest to the geologist than all the other kinds of lavas.

We will now proceed to consider the facts which are brought to light concerning the nature of lavas, when they are studied by the aid of the microscope. Although most lavas appear at first sight to be opaque substances, yet it is easy to prepare slices of them which are sufficiently thin to transmit light. In such thin transparent slices we are able to make out, by the aid of the microscope, certain very interesting details of structure, which afford new and important evidence bearing on the mode of origin of these rocks.

Host lavas are capable of being melted by the heat of our furnaces; but the different kinds of lava vary greatly in the degree of their fusibility. The basic lavas, or those with the smallest proportion of silica, are usually much more easily fusible than those which contain a high percentage of silica, the add lavas.

Now, it is a very noteworthy circumstance, that when a lava is artificially fused it assumes on cooling very different physical characters to those which were presented by the original rock.

If we examine the freshly-broken surface of a piece of lava, we shall, in most cases, find that it contains a great number of those regular-shaped bodies which we call crystals; in some cases these crystals are so small as to be scarcely visible to the naked eye, in others they may be an inch or more in length. Most lavas are thus seen to be largely made up of crystals of different minerals. The minerals which are usually contained in lavas are quartz, the various kinds of felspar, augite, hornblende, the different kinds of mica, olivine, and magnetite.

But when a piece of lava is melted in a furnace, all these crystalline minerals disappear, and the resulting product is the homogeneous substance which we call glass. If, as many suppose, lavas acquire the fluidity which they possess when issuing from volcanic vents as the result of simple fusion it is strange that artificially fused lavas do not agree more closely in character with the natural products.

A careful examination of different kinds of lavas, however, will show that they vary very greatly in character among themselves. Some lavas are as perfectly glassy in structure as those which have been artificially fused, while others contain great numbers of crystals, which may sometimes be of very large size.

If we prepare thin transparent slices of these different kinds of lavas, and examine them by the aid of the microscope, we shall find that lavas are made up of two kinds of materials, a base or groundmass of a glassy character, and distinct crystals of different minerals, which are irregularly distributed through this glassy base, like the raisins, currants, and pieces of candied peel in a cake. In some cases the glassy base makes up the whole mass of the rock; in others, smaller or larger numbers of crystals are seen to be scattered through a glassy base; while in others again the crystals are so numerous that the presence of an intervening glassy base or groundmass can only be detected by the aid of the microscope.

If thin slices of the glassy materials of lavas be examined with high magnifying powers, new and interesting facts are revealed. Through the midst of the clear glassy substance cloudy patches are seen to be diffused; and, if we examine them with a still higher power, these cloudy patches resolve themselves into innumerable particles, some transparent and others opaque, having very definite outlines. At the same time fresh cloudy patches are brought into view, which can only be resolved by yet higher powers of the microscope. In examining these natural glasses by the aid of the microscope, we are forcibly reminded of what occurs when the 'Milky Way' and some other parts of the heavens are studied with a telescope. As the power of the instrument is increased the nebulous patches are resolved into distinct stars, but fresh nebulous masses come into view, which are in turn resolved into stars, when higher powers of the instrument are employed.

In the Frontispiece, No. 1 illustrates the appearance presented by these volcanic glasses when examined with a high power of a microscope. Through a glassy base is seen a number of diffused nebulous patches, which are in places resolved into definite particles.

These minute particles of definite form, which the microscope has revealed in the midst of the glassy portions of lava, have received the name of microliths, or crystallites. The study of the characters and mode of arrangement of these microliths or crystallites has in recent years thrown much new light on the interesting problems presented by lavas.

In some glassy lavas the microliths or crystallites, instead of being indiscriminately diffused through the mass of the base or groundmass, are found to be collected together into groups of very definite form. In No. 2 of the Frontispiece we have a section of a glassy rock in which the crystallites have united together, so as to build up groups presenting the most striking resemblance to fronds of ferns. Around these groups spaces of dear glass have been left by the gathering up of the crystallites, which in other parts of the mass are seen to be equally diffused through it. In this formation of groups of microliths we cannot but recognise the action of those crystalline forces, which on frosty mornings cover our windows with a mimic vegetation composed of icy particles.

In other cases, again, the crystallites scattered through the glassy portions of lavas unite in radial groups about certain centres, and thus build up globular masses to which the name of 'sphærulites' has been given. No. 3 in the Frontispiece illustrates the formation of these sphærulites.

Now, a careful study of the microliths or crystallites has proved that they are the minute elements of which those wonderfully beautiful objects which we call crystals are built up. In some cases we can see that the crystallites are becoming united together in positions determined by mathematical laws, and the group is gradually assuming the outward form and internal structure of a crystal. In other cases crystals may be found which are undergoing a disintegrating action, and are then seen to be made up of minute elements similar to the crystallites or microliths of glassy rocks.

The conclusion is confirmed by the fact that if we take an artificially fused lava and allow it to cool slowly, it will be found that the glassy mass into which it has resolved itself contains numerous crystallites. If the cooling process be still further prolonged, these crystallites will be found to have united themselves into definite groups, and sometimes distinct crystals are formed in the mass; under these circumstances the rock frequently loses its glassy appearance and assumes a stony character.

In connection with this subject, it may be mentioned that some years ago a very ingenious invention was submitted to trial in the Works of the Messrs. Chance, of Birmingham. It had been suggested that if certain lavas of easy fusibility were melted and poured into moulds, we might thus obtain elaborately ornamented stone-work, composed of the hardest material, without the labour of the mason. The molten rock when quickly cooled was found to assume the form of a black glass, but when very slowly cooled passed into a stony material. Unfortunately, it was found that this material did not withstand the weather like ordinary building stones, and, in consequence, the manufacture had to be abandoned.

Now, the study of the products of volcanoes has led geologists to recognise the true relations between glassy and crystalline rocks.

In the amorphous mixture of various silicates which compose a glass, chemical affinity causes the separation of certain portions of definite composition, and these form the microliths or elements of which different crystalline minerals are built up. Under the influence of the crystalline forces, there is a great shaking or agitation in the mass, and the microliths of similar kind come together and become united, like the fragments in Ezekiel's valley of dry bones.

Although we cannot see this process taking place under our eyes, in a mass of lava, yet we may study specimens in which the action has been arrested in its different stages. In order to understand the development of an acorn into an oak-tree, it is not necessary to watch the whole series of changes in a particular case. A visit to an oak-thicket, in which illustrations of every stage of the transformation may be found, will afford us equally certain information on the subject.

In the same way by the examination of such a series of rock-sections as that represented in the Frontispiece, we may understand how, in the midst of a mass of mixed silicates constituting a natural glass, the separation of microliths takes place; these unite into groups which are the skeletons of crystals, and finally, by the filling up of the empty spaces in these skeletons, complete crystals are built up. The series of operations may, however, be interrupted at any stage, and this stage we may have the chance of studying.

We are able, as we shall show in a future chapter, to examine many rock-masses that have evidently formed the reservoirs from which volcanoes have been supplied, and others that fill up the ducts which constituted the means of communication between these subterranean reservoirs, and the surface of the earth. Now in these subterranean regions the lavas have been placed under conditions especially favourable for the action of the crystalline forces—they must have cooled with extreme slowness, and they must have been under an enormous pressure, produced in part by the weight of the superincumbent rocks, and in part by the expansive force of the imprisoned steam. We are not, therefore, surprised to find that in these subterranean regions, the lavas, while retaining the same chemical composition, have assumed a much more perfectly crystalline condition. In some cases, indeed, the whole rock has become a mass of crystals without any base or groundmass at all.

An examination of the Frontispiece will illustrate this perfect gradation from the glassy to the crystalline condition of lavas. No. 1 represents a glass through which microliths or crystallites of different dimensions and character are diffused. In Nos. 2 and 3, these crystallites have united to form regular groups. In No. 4, which may be taken as typical of the features presented by most lavas, we have a glassy groundmass containing microliths (a 'crypto-crystalline base'), through which distinct crystals are distributed. Nos. 5 and 6 illustrate the characters presented by lavas which have consolidated at considerable depths beneath the surface; in the former we have a mans of small crystals (a 'micro-crystalline base') with larger crystals scattered through it; while the latter is entirely made up of large crystals without any trace of a base or groundmass.

Now, as all lavas are found sometimes assuming the glassy condition at the surface, so when seen in the masses which have consolidated with extreme slowness, and under great pressure, in subterranean regions, the same materials are found in the condition of a rock which is built up entirely of crystals. Chemists have found that artificial mixtures of silicates in which soda and potash are present in considerable quantities, have a great tendency to assume the glassy condition on cooling from a state of fusion, and glass manufacturers are always careful to use considerable proportions of the alkalis as ingredients, in making glass. It is found, in like manner, that those lavas which contain the largest portion of the silicates of soda and potash (the 'acid lavas') most frequently assume the condition of a natural glass.

Geologists have given distinct names to the glassy and the perfectly crystalline conditions of the different kinds of lavas, the glassy varieties being found in masses which have cooled rapidly near the surface, and the crystalline varieties in masses which have cooled slowly at great depths. The names of these two conditions of the five great classes into which we have divided lavas are as follows:—

As vitreous rocks have little in their general appearance to distinguish them from one another, the glassy forms of the first four classes of lava have not hitherto received distinct names, but have been confounded together under the name of obsidian. If we determine the specific gravities of rocks having the same composition but different structures, we shall find that they become heavier in proportion as the crystalline structure is developed in them. Thus gabbro is heavier, but tachylyte is lighter than basalt, bulk for bulk, though all have the same chemical composition.

Nor are the crystals contained in lavas less worthy of careful study, by the aid of the microscope, than the more or less glassy groundmass in which they are embedded. Mr. Sorby has shown that the crystals found in lavas, exhibit many interesting points of difference from those which separate out in the midst of a mass of the same rock, when it has been artificially melted and slowly cooled. There are other facts which also point to the conclusion that, while the glassy groundmass of lavas may have been formed by cooling from a state of fusion, the larger and well-formed crystals in these lavas must have been formed under other and very different conditions.

The larger crystals in lavas exhibit evidence of having been slowly built up in the midst of a glassy mass, containing crystallites and small crystals. We can frequently detect evidence of the interruptions which have occurred in the growth of these crystals in the concentric zones of different colour or texture which they exhibit; and portions of the glassy base or groundmass are often found to have been caught up and enclosed in these crystals during their growth.

But when we find, as in the porphyritic pitchstones, a glassy base containing only minute crystallites, through which large and perfectly formed crystals are distributed, we can scarcely doubt that the minute crystallites and the larger crystals have separated from the base under very different conditions. This is indicated by the bet that we detect in these cases no connecting links between the embryo microliths and the perfect crystals; and a confirmation of the conclusion is seen in the circumstance that many of the crystals are found to have suffered injury as if from transport, their edges and angles being rounded and abraded, and portions being occasionally broken off from them.

Hence we are led to conclude that the larger crystals in lavas were probably separated from the amorphous mass in the subterranean reservoirs beneath the volcano, and were carried up to the surface in the midst of the liquefied glassy material which forms the groundmass of lavas. When we come to examine these crystals more closely, we find that certain very curious phenomena are exhibited by them which lend powerful support to this conclusion.

It is found convenient by geologists to designate those rocks which have consolidated in deep-seated portions of the earth's crust as Platonic Rocks, confining the name of Volcanic rocks to those consolidating At the surface; but Plutonic and Volcanic Rocks shade into one another by the most insensible gradations.

When the crystals embedded in granitic rocks, and in some lavas, are examined with the higher powers of the microscope, they are frequently seen to contain great numbers of excessively minute cavities. Each of these cavities resembles a small spirit-level, having a quantity of liquid and a bubble of gas within it. In fig. 7 we have given a series of drawings of these cavities in crystals as seen under a high power of the microscope. In No. 1 a group of such cavities is represented, one of which is full of liquid, while two others are quite empty; the remaining cavities all contain a liquid with a moving bubble of gas. In No. 2 two larger cavities are shown, containing a liquid and a bubble of gas; and it will be seen from these how varied in form these cavities sometimes are. In Nos. 3, 4 and 6 the liquid in the cavities contains, besides the bubbles, several, minute crystals; and in No. 6 we have a cavity containing two liquids and a bubble.

In the largest of such cavities the bubble is seen to change its place so as always to lie at the upper side of the cavity, when the position of the latter is altered, just as in a spirit-level. But in the smallest cavities the bubbles appear to be endowed with a power of spontaneous movement; like imprisoned creatures trying to escape, these bubbles are seen continually oscillating from side to side and from end to end of the cavities which enclose them. In fig. 8 a minute cavity containing a liquid and bubble is shown, the path pursued by the latter in its wonderful gyrations being indicated by the dark line. These cavities are exceedingly minute, and so numerous that in some crystals there must be millions of them present; indeed, in certain cases, as we increase the magnifying power of our microscopes, new and smaller cavities continually become visible. It has been estimated that in some instances the number of these minute liquid-cavities in the crystals of rocks amounts to from one thousand millions to ten thousand millions in a cubic inch of space.

What is the nature of the liquids which are thus imprisoned in these cavities contained in the crystals of lavas and granites? Careful experiments have given a conclusive answer to this question. In many cases the liquid is water, usually containing considerable quantities of saline matter dissolved in it. Sometimes the saline matters are present in such abundance that they cannot all pass into solution, but crystallise out, as in fig. 7—Nos. 3, 4, 5—where cubic crystals of the chlorides of sodium and potassium are seen floating in the liquid; in other cases the liquid is a hydrocarbon like the mineral oil which is present in great abundance in deep-seated rocks in many parts of the globe. But in some other cases the liquid contained in the cavities of crystals is found to be one which could scarcely be anticipated to occur under such circumstances—the gas known as carbonic add, which under extreme pressure can be reduced to a liquid condition. In cavities containing liquefied carbonic acid, if the rock be warmed up to 86° or 90° Fahrenheit the bubble suddenly vanishes, sometimes with an appearance like ebullition or boiling, as represented in fig. 9. Now the temperature which we have indicated is the 'critical point' of carbonic acid, and above that temperature it cannot exist in a liquid condition, however great may be the pressure to which it is subjected. The liquid has been converted into a gas which completely fills the cavity. The carbonic acid in the cavities of crystals has frequently been isolated and its nature placed beyond doubt by spectroscopic and ordinary chemical tests.

The presence of these liquids in the cavities of crystals clearly proves that the latter must have been formed under enormous pressure—a pressure sufficiently great to reduce, not only steam, but also volatile hydrocarbons and even gaseous carbonic acid, to the bulk of a liquid.

Such conditions of enormous pressure we may infer to exist in the deep-seated reservoirs beneath volcanoes, where, besides the weight of the superincumbent rock-masses, we have the compressing force of great quantities of elastic vapour held in confinement. The crystals of which granitic rocks are entirely built up exhibit clear evidence of having been all formed under these conditions of enormous pressure. The glassy base or groundmass of lavas, on the other hand, presents all the characters of materials that have cooled from a state of fusion. Most lavas consist in part of crystals, exhibiting fluid-cavities like those present in granite, and in part of a base, which has evidently been formed by the cooling of a fused mass. We are therefore justified in concluding that the crystals have been formed in subterranean recesses, and that the groundmass or base has consolidated at the surface. The bearing of these conclusions upon some of the great problems presented by volcanoes we shall have occasion to point out in the sequel.

One of the most interesting inquiries suggested by the study of the liquid-cavities in volcanic rocks is that of the cause of the apparently spontaneous movement of the bubbles which we have described as taking place in some of the smaller of them. The ingenious experiments of Mr. Noel Hartley have suggested to Professor Stokes an explanation which is probably the true one. It appears that these minute globes of vapour are in such a state of unstable equilibrium as to be affected by the smallest changes of temperature, and that the variations in the heat of the atmosphere, due to currents of air and the movement of warm or cold bodies through it, are sufficient to cause the oscillation of these sensitively poised bubbles.

The short account which we have been able to give in the foregoing pages of the researches that have been carried on concerning the nature of the materials ejected from volcanoes will serve to show that these investigations have already made known many facts of great interest, and that the farther pursuit of them is full of the highest promise. To the scientific worker no subject is too vast for his research, no object so minute as to be unworthy of his most patient study. In some of our future inquiries concerning the nature of volcanic action, we shall be led to an investigation of the phenomena displayed in the sun, moon, comets and other great bodies of the universe; but another road to truths of the same grandeur and importance is found, as we have seen, in an examination of the mode of development of crystallites, and a study of the materials contained in the microscopic cavities of the minutest crystals.