Art. LII.—The Physical and Chemical State and Probable Role of Water in Rock-magmas.
[Read before the Wellington Philosophical Society, 25th September, 1912.]
During recent years there has accumulated much evidence showing that an abundance of water is present in some or all rock-magmas, and may play a prominent part in their formation and subsequent consolidation. Thus a theory of so-called aqueo-igneous fusion has been evolved, more especially with respect to plutonic and hypabyssal rocks of acidic composition. Though there is a considerable amount of vagueness in the accessible literature dealing with the function of water in aqueo-igneous fusion, apparently many geologists and others distinctly hold the belief that the water present acts as a flux or solvent of silica, silicates, and other minerals not only when it is in the liquid state, but also when it is in the gaseous state. Moreover, some writers seem to assume that there is no practical limit imposed by temperature conditions on the solvent action of water. On the other hand, there are those who tacitly or explicitly make the assumption that the presence or absence of water and other “mineralizers” is a matter of secondary importance. For instance, Harker (6,* pp. 288 et seq.) certainly does not overemphasize the function of water as a solvent and mineralizing agent. Such a standpoint seems to be supported by the data obtained by Albert Brun, of Geneva, and published in his recent work entitled “Recherches sur l'Exhalaison Volcanique” (9). Brun maintains (pp. 254–55) that water is not present in lavas during extrusion, and, in fact, is wholly absent during ordinary volcanic activity. These conclusions are of doubtful validity, but if admitted would appear to lead to the corollary that water is not present in ordinary rock-magmas. It will, however, generally be granted that, notwithstanding Brun's new data, water is by far the most abundant of the volatile substances present in rock-magmas, and therefore that its role is relatively much more important than that of other “mineralizers,” such as fluorine, chlorne, &c. Hence in this discussion the latter are more or less neglected. They seem, indeed, to be of importance only in connection with the presence of certain minerals in veins genetically related to granitic intrusions.
Physical state of water.
If water is to be regarded as a solvent for silicates and other constituents of rock-magmas, the question of its physical state whilst so acting deserves consideration. The critical temperature of pure water being 365° C., and the critical pressure 200–5 atmospheres, some, if not most, writers assume that at higher temperatures, whatever the pressure, the water in a magma is necessarily in the gaseous state. As evidence of this the following quotation may be made: “He [Arrhenius] considers a magma as a complex solution containing various silicates, &c., and also gases, the latter, of course, including in the first place water, which above its critical temperature of about 365° must be ranked as a gaseous body” (6, p. 295). Again, with*
[Footnote] * This and other numbers enclosed in brackets refer to list of literature at end.
reference to so-called pneumatolytic action, Vogt quotes Arrhenius as saying, “The solution in aqueous gas now gradually cools, and one substance after another gradually separates from it. By reason of the great mobility of the solution, and its consequent strong capability of diffusion, the minerals (provided the cooling be not too rapid) are segregated in large crystals, such as characterize a so-called pegmatitic structure. Gradually also the constituents which longest retain a gaseous form—such as water and carbonic acid—escape” (1, p. 644).
If we consider the water in a magma to be chemically combined, no difficulty arises as to its physical state until the temperature is sufficiently high to destroy the combination. The more general view is to assume that the water is in solution. Again, there need be no discussion as to its physical state so long as the water or aqueous gas is truly in solution; but the temperature may become so great that steam separates from the magma in the form of bubbles. In this latter condition it can no longer be regarded as acting the part of a solvent for the containing magma. There is certainly need both for a more exact terminology in order to prevent confusion of thought, and for experimental work in order to furnish a better foundation for magmatic hypotheses.
The difficulties involved in the assumptions made by some writers are thus stated from a chemist's point of view by Dr. James Moir: “All sorts of geological authorities accept the belief that water can be made red-hot and yet preserve its solvent properties. Now, every chemist and physicist knows that above the critical point water can only exist as steam, entirely devoid of solvent powers except for other vapours; and this however high the pressure…. As rock-magmas are certainly not vapours, there is no possibility of anything except an uncombined emulsion of rock and steam” (7, p. 4). Dr. Moir relies upon the experimental work of Andreas Smits and J. P. Wuite (5, abstracts, ii, p. 985), who found that the solubility of sodium sulphate in water became zero at 365° C. This result, however, cannot reasonably be extended to silicate solutions without experimental proof. We know that the boiling-point of water containing dissolved solids is above 100° C. at ordinary atmospheric pressure, and also that water at temperatures above 200° C. exerts a strong dissolving power on silicates. In the light of these facts it appears probable that an aqueous solution of silica or a silicate will not necessarily cease to exist as such at 365° C. Moreover, there is some doubt as to whether the so-called critical temperatures of weter and other volatile bodies are really constants irrespective of pressures exceeding the critical pressures (2, pp. 460–61), and therefore it is to be hoped that experimental data showing the solubility of silicates in water at temperatures above 300° C. will soon be forthcoming.
The whole matter may be regarded from another point of view. The most enthusiastic supporter of aqueo-igneous fusion would not ask for the presence of more than 10 per cent, of water in any magma. Is there any real objection to regarding this amount of water as chemically combined with the silica and silicates of the rock-magma? The immense pressures prevailing in magmas may bring about a real chemical combination such as would be impossible at atmospheric pressure.
Hypothetical Classes of Magma.
A clear definition of the difference between water acting as a solvent and water chemicaly combined in magmas must be left to the physical chemist. In the following statements a return is made to the nomenclature
appropriate to the phenomena of solution. The well-known experiments of C. Barus show that at 185° C. and upwards water has a strong solvent action on soft glass. Lemberg shows that at 210° C. water slowly dissolves anhydrous powdered silicates (1, pp. 308, 643, 770). If the solubility of silicates does not materially diminish as 365° C. is approached, then, provided the pressure is adequate, a silicate solution will probably continue to exist as a solution at higher temperatures. We may assume, for example, a solution of silicic acid and hydrous silicates in an excess of water above that chemically combined, the whole at a temperature of 400° C., or even 500° C. Such a solution may correspond to some aqueo-igneous magmas. If the temperature of this hypothetical solution be raised it will reach such a point that the water not chemically combined becomes potentially gaseous—that is to say, it ceases to act as a solvent, and tends to separate itself from the magma, but is held (more or less) in solution by the prevailing pressure. Viscosity of the magma will also tend to prevent mechanical separation. There may now be an inclination for the water combined with silicates to break away from this union, but probably much or all as yet remains chemically combined, and the hydrous silicates still mutually dissolve one another, notwithstanding that the temperature is lower than that required for the ordinary fusion of anhydrous silicates. If the temperature still rises, presumably in the end the combinations with water are broken up. Since micas and amphiboles containing water form in both plutonic and volcanic rocks, and since analcite and, it is believed, calcite occur as primary minerals in various igneous rocks, there seems to be no difficulty in supposing that the ordinary magma contains dissolved or combined water at temperatures reaching or exceeding 1000° C. “Dissolved water” in such a magma is really dissolved steam, but the solution is a liquid, not “an uncombined emulsion of rock and steam.” At a temperature of, say, 1200° to 1500° C. chemical combination of water with silicates, as indicated above, may cease. Some steam will still be held in solution by pressure, but some probably separates in the form of gaseous bubbles. Magma at a high temperature (“superheated magma” of Daly) may be unable to dissolve more than a trace of steam, and, if so, bubbles of aqueous gas or steam will rise in the magma until either they reach a cooler portion where they can be redissolved, or are stopped by the solid rock that forms the upper boundary or roof of the liquid mass. Here the gaseous water and other volatile substances present may act on the roof rock, thus forming new magma. This magma will necessarily contain water.
From such considerations as those just stated it follows that there is probably a continuous passage from so-called aqueo-igneous magmas at temperatures not far above 365° C. to high-temperature magmas not requiring the presence of water as a flux. The writer, with considerable hesitation, divides magmas into three classes:—
Aqueo-igneous magmas, in which much water is present, some chemically combined, some in solution, and acting energetically as a solvent. Temperature range, say, 350° C. to 700° C.
Igneo - aqueous magmas, in which the water present is mainly chemically combined. Any water not chemically combined is a gas, held in solution by pressure, but probably not appreciably aiding in the fluxing of the silicates forming the main part of the magma. Temperature range, say, 700° C. to 1000° C., or more.
Fusion magmas, which maintain a liquid form without the assistance of water. Any water present is potentially gaseous, uncombined with silicates, and held in solution by pressure alone.
If in rock-magma a distinction between solution and chemical combination cannot be made then classes I and II must be merged. In any case, there cannot well be any hard-and-fast lines between, the three classes. Transition temperatures will depend to some extent upon conditions of pressure and of chemical composition.
Hypothetical Distribution of Magma in the Earth's Crust.
If the influence of pressure, which is practically constant for a given depth and that of chemical composition, which is of minor importance in the present discussion, be discarded, the presence or absence of liquid rock at a stated depth in the earth's crust depends within certain limits both upon the temperature and upon the presence or absence of water. Provided sufficient water (or other “mineralizer” or flux) is present, we may have aqueo-igneous magma at quite moderate depths. Below this comes igneo-aqueous magma, followed by fusion magma. Assuming that temperature continues to increase towards the centre of the earth, and that ordinary physical laws are not essentially modified by immense pressure, we may suppose, with Arrhenius, that the fusion magma ultimately becomes gaseous.* Into this zone, however, there is no present need to venture. With a patchy and limited distribution of water in the earth's crust we may have at relatively shallow depths disconnected reservoirs of aqueo-igneous or igneo-aqueous magma surrounded by solid rock, with fusion magma at some greater depth. Where water is absent we shall have fusion magma only, confined by a solid roof of considerable thickness. It is quite possible, and, indeed, probable, that the influence of increasing pressure is sufficient to prevent the formation of fusion magma, except in those parts of the earth's crust where the temperature gradient is somewhat above the average.
Hypothetical Formation of Magmas.
In order to narrow discussion, let us consider a segment of the earth's crust assumed originally to have a temperature gradient of 1° C. in 200 ft., and to contain appreciable amounts of water to a depth of ten or fifteen miles, but little or no water at greater depth. Such a segment will be solid to a depth of forty miles, below which may be ordinary fusion (universal) magma. If now, through earth-movements or other cause, the temperature of the segment rises, the fusion magma, owing to melting of the overlying rock, will extend upwards. Even though the arguments in favour of a solid earth be considered valid, yet it will doubtless be admitted that a sufficient rise of the temperature gradient will cause melting in places, and this is all that is here postulated. Sooner or later aqueo-igneous magma will begin to form at a depth of, say, fifteen miles, and will gradually work its way upward. Two cases, dependent on amount of heat supplied, are conceivable. In the less important of these, the temperature gradient failing to reach, say, 1° per 100 ft., the fusion magma does not eat its way upward into contact with the aqueo-igneous magma. The latter in this case will probably cease its upward movement six or seven miles below the surface. On the
[Footnote] * There are, of course, well-known astronomical reasons for believing that the earth as a whole is rigid, and in discussion upon this paper Mr. G. Hogben pointed out that the transmission of earthquake-waves through the earth seems to preclude the possibility of a liquid or gaseous interior, at least to a depth of eight hundred miles. Hence appears an evident weakness in the above assumptions.
other hand, if the temperature gradient becomes relatively steep the lower magma will ultimately junction with the upper. In view of the fact that the lower magma may reasonably be expected to contain at least a small amount of water and other mineralizers, either as gas or contained in an aqueo-igneous differentiate, such junction means transference of water and heat to the aqueo-igneous magma. Lateral extension of the fusion magma will enable it to maintain a supply of water in gaseous form to the aqueo-igneous magma. The latter, therefore, will work (or, in Daly's phrase, “stope”) its way upward at an accelerated pace. So the action will go on until either thermal equilibrium is established or some new condition connected with approach towards the earth's surface comes into play.
Most geologists will probably concede that in some such way as that indicated above solid rocks comparatively near the earth's surface may pass into the liquid condition. The prominent role here assigned to water will, however, be disputed, and doubtless strong arguments can be marshalled against statements resting on so hypothetical a basis.
Under the assumptions made in this paper a magma during formation will necessarily differentiate into two, or perhaps three, parts, distinguished by differences in water-content. The preference displayed by water for silica and alkaline silicates leads to the belief that the aqueo-igneous portion of the magma will be relatively light and acidic, whilst the fusion magma will be heavy, basic, and non-aqueous. The igneo-aqueous portion, intermediate in position, will probably be intermediate also in chemical composition, but may incline to acidity. Further differentiation, it is easy to imagine, will result through the cooling of portions of the magma to the point at which crystallization begins. The opening of a passage to the earth's surface, whereby part of the magma may be extruded, will also give rise to differentiation, which under some conditions may be of a varied nature. The absence of water probably limits differentiation very considerably; and if it be possible for a differentiated magma to rise as a whole to a temperature above that required for an ordinary fusion magma, then it may be assumed that convection currents will check differentiation, and tend to bring about an approximately uniform composition.
Formation of Granite.
An aqueo-igneous magma will in general have the composition of a granite or an acid diorite. Ultimate consolidation of a granitic magma is largely influenced not only by cooling, but also by conditions permitting the escape of water, such as obtain when a rising magma approaches the surface of the earth. The consolidation temperature of granite may be between 575° C. and 800° C.* (8, p. 342), but, according to some geologists of the French school, a lower temperature is more probable. It may be observed that under the assumptions made in this paper some granites probably represent more ancient re-fused granites, gneisses, and rocks of sedimentary origin, and thus their formation completes a cycle of change.*
[Footnote] * The quartz of granite shows by its etch figures that it was once β quartz, into which ordinary α quartz passes at 575° C. (552° C. according to Brun). The upper limit is 800° C., because at that temperature quartz inverts to tridymite. It may be suggested, however, that extreme pressure would modify these data.
Highly heated water escaping into fissures from a solidifying granite will naturally carry silica and silicates with it in solution. These will ultimately crystallize as pegmatite veins (6, pp. 294–96). The consolidation temperature may be below 365° C., and is with tolerable certainty below 575° C. (8, p. 342). No difficulty need be experienced in accounting for the various rare or peculiar minerals found in some pegmatite veins. These are generally due to the presence of volatile bodies rejected with water from the consolidating granite. Any special pneumatolytic hypothesis seems unnecessary. In New Zealand pegmatitic veins, as a whole, are remarkably poor in accessory minerals, tourmaline being the only one at all frequently observed. Hence arises one reason why the writer lays stress on water, and not on other volatile substances, as a “mineralizing agent.”
Syemtic rocks are quantitatively of no great importance, but present a great variety of types. Since they are usually found on the outskirts of granitic masses, it may be assumed that, as a rule, they represent differentiates from consolidating granite magmas. Their variety is perhaps due to varying conditions of temperature and of assimilation of solid rock, and more especially to variation in water and other mineralizers present. Some syenites may be derived from igneo-aqueous magmas.
Diorites, more especially the acid types, may in part be due to differentiation from a granitic magma. Some diorites may represent consolidated igneo-aqueous magmas, whilst others may be derived from fusion magmas.
Acid Hypabyssal Rocks.
The acid intrusives may be regarded as apophyses from granitic magmas. Water probably plays a prominent part in their formation, its presence reducing viscosity and preventing premature consolidation. Quartzporphyry is said to consolidate, like granite, at temperatures above 575° C. (8, p. 342).
Formation of Schists.
Where schistose rocks occur in highly folded mountain-chains, and, moreover, are associated with granitic masses, it is easy to account for their format on by invoking a theory of dynamo-thermo-metamorphism. Gently folded schists extending over a wide area, such as the quartz-mica-schists of Central Otago, present a more difficult problem. It has been suggested that the Otago schists are due to thermal metamorphism, induced by underlying granite. Such a theory implies the formation of igneous rock beneath a wide area. In Central Otago evidence in favour of such a view is afforded by the occurrence in the schists of quartz veins carrying the tungsten mineral scheelite.
It is commonly held that steam-pressure has much to do with volcanic eruptions (4, pp. 42, 45). Other imprisoned gases assist, and Harker suggests that gravitational pressure alone may cause fissure eruptions. Arrhenius supposes that sea-water may penetrate by means of fissures to a
magma near the earth's surface. The magma eagerly absorbs the water, and in consequence swells and becomes more fluid. The resulting pressure, which is analogous to osmotic pressure, forces the magma outwards, and thus volcanic eruption begins (4, p. 18).
According to the assumptions made in this paper, a fusion magma has little or no inclination to absorb water, and, if so, the hypothesis advocated by Arrhenius fails to hold good. Moreover, it is not easily understandable how large quantities of sea water could find their way into a magma against enormous opposing pressure.
An aqueo-igneous magma stoping its way upward during folding movements of the earth's crust may find a plane of weakness leading to the surface. In this case the probability is that water and other volatile substances will escape, whilst the magma, or those portions losing water, will consolidate.
It may happen that heat has accumulated in part of an originally aqueoigneous magma to such an extent that the temperature of this portion is raised well above 1000° C., and all or nearly all its water is free from combination. The opening of communication with the surface will relieve the pressure that keeps the aqueous and other gases confined, or holds them in solution. The result will be a series of violent explosions in the volcanic vent, analogous to geyser eruptions, which will scatter volcanic ash, pumice, &c., far and wide. When the gases have nearly exhausted them-selves there may follow effusions of acidic lava. The cooling of the magma as it rises in the volcanic conduit, together with the loss of water and the extrusion of lava, promotes part crystallization and differentiation. If to these changes we add the influence of magma drawn laterally to the volcanic vent or vents from regions outside the original disturbance, we shall have hypothetical data quite adequate to account for all the phenomena of vulcanism.
Volcanic action may originate through a fusion magma, nearly devoid of water, finding its way to the surface. Such an occurrence is rare, or perhaps impossible.
Hot Springs and Geysers.
Hot springs and geysers are usually attributed to expiring volcanic action. Since the early stages of vulcanism may afford escape for all or most of the water in a magma, a magmatic source for the heated waters of the final stage can be postulated only with difficulty. If, however, an aqueo-igneous or other magma containing water approach the surface it may give rise to boiling springs, the water of which (in part at least) will be of magmatic origin. Thus hot springs may indicate rise of magma, and precede as well as follow a period of igneous activity.
Summary and Conclusions.
The physical or chemico-physical state of water in rock-magmas is not a matter of indifference. By making assumptions that are consistent with known facts, but have not as yet been experimentally proved or disproved, the following conclusions may be reached:—
It is possible for an aqueo-igneous magma to form at temperatures not far above 365° C. In this magma there may be uncombined water acting as a solvent and physically in the liquid state. At somewhat higher temperatures uncombined water no longer acts as a solvent, and may be regarded as physically a gas, held in solution only by pressure. At still higher temperatures no combined water can exist in the magma and all
water present is physically a gas, either held in solution by pressure alone or actually segregated in the form of gas-bubbles.
The presence of water induces or aids the differentiation of a magma into two or more layers differing in chemical composition.
The presence or absence of water has much to do with the distribution and formation of magma in the upper layers of the earth's crust.
The escape of water, as well as cooling, is a determining cause in the consolidation of aqueo-igneous magmas.
Some plutonic rocks, more especially granites, may represent sedimentary rocks fused by aqueo-igneous action.
Differentiation in consolidating aqueo-igneous magma is partly explained.
A fairly efficient hypothesis of volcanic action and magmatic differentiation in connection therewith can be evolved.
Hot springs may precede as well as follow a period of volcanic activity. In the latter case their water is probably not magmatic.
By varying the permissible assumptions in connection with water in rock-magmas a great variety of hypothetical results may be obtained. The necessity for eliminating false assumptions by such experimental research as is possible therefore becomes obvious. Though experimental work in geo-physics is difficult and expensive, and in some respects can never be conclusive, much may be done in a properly equipped laboratory. Hence there is reason for hoping that in a few years' time our knowledge of the earth's interior may be greatly increased, and placed on a much more satisfactory basis.
1. Posepny, F.; Vogt, J. H. L.; Kemp, J. F.; and others: “Genesis of Ore-deposits.' Published by American Institute of Mining Engineers. Second edition, New York, 1902.
2. Preston, Thomas: “Theory of Heat.” Second edition, revised by J. Rogerson Cotter, 1904.
3. Chamberlin, T. C., and Salisbury, R. D.: “Geology.” Vol. 1, first edition, 1905.
4. Arrhenius, Svante: “Worlds in the Making.” Translated by H. Barus. 1908.
5. Journal of Chemical Society, December, 1909. Abstracts, ii, p. 985. Reference to Andreas Smits and J. P. Wuite in Proc. k. Akad. Wetensch. Amsterdam, xii, 1909, pp. 244–57.
6. Harker, Alfred: “The Natural History of Igneous Rocks.” 1909.
7. Moir, James: Inaugural Address in Journal of the Chemical, Metallurgical, and Mining Society of South Africa, vol. 11, No. 1, July, 1910.
8. Clarke, F. W.: “Data of Geo-chemistry.” Second edition. U.S.G.S. Bulletin No. 491, 1911.
9. Brun, Albert: “Recherches sur l'Exhalaison Volcanique.” Geneva, 1911. See also “The New Vulcanology,” by E. B. Bailey in Geol. Mag., dec. 5, vol. 8, 1911, pp. 268–73 and 311–16; and “Brun's New Data on Vulcanism,” by Alex. N. Winchell, in Economic Geology, vol. 7, No. 1, January, 1912, pp. 1–14.
10. Daly, Reginald A.: “The Nature of Volcanic Action.” Proc. Amer. Acad. Arts and Sci., vol. 47, No. 3, June, 1911, pp. 47–122. Summary in Journal of Geology, vol. 20, No. 5, July-Aug., 1912, pp. 471–75.