Hyperfusible, adj. (of a substance) capable of reducing the melting ranges in end-stage magmatic fluids.

The course of fractional crystallization of subalkaline magmas epitomized in the schematic reaction series is the net result of all the factors that contribute to phase equilibrium in the poly-component system concerned. Just what the components of the system may be in the strict sense of the phase rule we do not know. The ingredients are SiO2, S, CaO, F, Al2O3, H2O, MgO, Cl, Na2O, CO2, and a host of others. It has become the custom, merely because it is expedient, to express some of the ingredients as oxides, some as elements, and so on, but there is no fundamental basis for so doing. In the liquid magma they all occur combined in a great variety of compounds of which we know very little but which enter into the liquid upon an equal basis of mutual solution. We are not concerned particularly with the contribution each of these ingredients makes to the control of the general course of fractional crystallization as outlined. We know only that the net result of the interplay of all their contributions is as we observe it. Now in the list of ingredients there are mentioned certain substances such as F and CO2 which are gaseous at the ordinary temperature and pressure, and others such as H2O and S which are gaseous at only moderately elevated temperatures. These may be contrasted in this respect with such ingredients as CaO and Al2O3. In connection with a magma crystallizing under adequate pressure the contrast mentioned has no significance. Each ingredient, of whichever class, contributes its share to the general equilibrium and separates as a constituent of the solid phases formed. This is as true of, say, H2O and F, which taken alone are liquid or gaseous at the ordinary temperature, as it is of CaO and SiO2. Thus H2O of our list separates as a constituent of hornblende of the reaction series and both H2O and F as a constituent of mica, and although H2O or F as such would be fluid or gaseous at the existing temperature, the whole mass may become completely solid at a comparatively elevated temperature, let us say somewhat arbitrarily 500°. There need be no unsolidified residue. All of the magma ingredients enter into the solid mass on an equal basis. There is no reason for distinguishing two classes of ingredients.


Under other conditions where lower pressure prevails the sum of the partial pressures of all the ingredients of the magma may at some stage exceed the external pressure. A gas phase will then separate from the liquid and the gas phase will contain all of the ingredients of the magma, each in an amount proportional to its partial pressure. Certain ingredients such as H2O, F, and the like will contribute to the gas phase in large amount and certain others such as CaO will be practically absent. Those constituents that dominate in the gas phase are referred to ordinarily as the volatile components, though certain substances not usually to be regarded as volatile may enter the gas phase in notable amounts; as, for example, iron and titanium in virtue of their capacity to form volatile halogen compounds.

At the moment, however, it is desired to defer consideration of the formation of a gas phase and direct attention to crystallization in those deeper-seated bodies of magma that are the principal loci of magmatic differentiation. With this purpose in mind the term “volatile components” as applied to the substances mentioned may be laid aside and attention focussed upon another of their properties; viz., that of having low melting temperatures. Water is unquestionably the most abundant and most important of these constituents. It is liquid at ordinary temperature. Others solidify at even lower temperatures than does water. Now it has been pointed out that the scheme of fractional crystallization presented is the net result of the contributions of all the ingredients to the equilibria involved. The scheme does not need modification in order to include the effects of these low-melting ingredients. Their effects are already in it. The scheme as expressed does, however, emphasize the behavior of the most important products of crystallization of the magma; namely, the silicates. Thus it is made clear that the early subtraction of calcic plagioclase tends toward enrichment of the liquid in the low-melting alkalic feldspars. The behavior of the ingredients of outstandingly low melting temperatures (water, etc.) is not specifically discussed, though some aspects of it are implicit in the schematic outline; as, for example, in the formation of mica as a late member of the reaction series. What it is now desired to emphasize is that their behavior is analogous to that of other low-melting ingredients. If crystallization is only moderately fractional they may be used up by entering into the composition of crystals, either those currently separating or by reaction with earlier crystals, in which case all liquid disappears at a comparatively early stage. Or if crystallization is notably fractional, they may be continuously concentrated into the lower-melting residues, which eventually come to carry a rather high concentration of these ingredients. We may call them the hyperfusible components. Water is the principal of these, and the substances ordinarily called volatile components are important, but other substances to which the term volatile has no application belong here as well, such as molybdates, tungstates, and phosphates of the alkalis. Their behavior may profitably be regarded in the light of the reaction principle.

If we return to a consideration of the schematic picture of the reaction series (Fig. 1), we may say that, with the more notable degrees of fractionation, the later magmas will become of that nature which, in virtue of their silicate ingredients, is called granitic. Such granitic magmas will naturally tend to carry a concentration of water and other hyperfusibles which has reached a high value in the same manner as has that of any other low-melting ingredient; for example, alkaline feldspar. Further crystallization of silicates, if fractional, gives a liquid more and more enriched in water, etc., and develops magmas from which granitic pegmatites may be formed. It is important to note that there is no necessity that a pegmatitic liquor be left over from the consolidation of a granite; all of the pegmatite-forming material may enter into constituents of the granite, directly or reactionally, water in mica, chlorine in apatite, sulphur in a sulphide, and so on. Indeed, such using-up of the residual liquors of, say, the granite, may be their normal fate in any very deep-seated mass if they remain uniformly distributed in interstitial relation to the granitic minerals. Only when some portion of them is removed from this relation to form the filling of a fissure in the nearly consolidated mass may they be sufficiently free from the influence of earlier silicates to act independently and give clear evidence of their character in the nature of the materials deposited.

In a full differentiation sequence from basic types to acid types there will be a natural tendency to this greater concentration of hyperfusibles in the acid types. Nevertheless, each magma in the sequence, representing the ever-changing mother liquor, will have had its appropriate quota of hyperfusible constituents. Moreover, any magma of the sequence may crystallize in such a way that no great fractionation with respect to its silicate constituents occurs but significant fractionation in the way of concentrating hyperfusibles in the residual liquor does occur. It thus may come about that the residuum of a gabbroid magma, crystallizing under the appropriate conditions, may be dioritic as measured by the nature of its silicate ingredients yet may contain a fairly large concentration of hyperfusible constituents. The mutual reactions in which these residual liquors and the silicates participate will necessarily be different from those occurring at a granitic stage. For any approach to a full discussion of all the possible reaction effects at various stages our knowledge is totally inadequate. It may not be amiss, however, to make an attempt to discuss the effects obtained in a basic magma on the one hand and in an acid magma on the other.

Effects of High Concentration of Hyperfusibles in Gabbroid Magma

In the presence of a high concentration of hyperfusibles, principally water, the ordinary minerals of a gabbro, pyroxene and plagioclase become unstable; i.e., liquids of that nature are saturated with other phases and will transform pyroxene and plagioclase into these other phases to an extent limited by the amount of reacting materials available. The instability of the plagioclase is characterized by the fact that the one end member, albite, remains stable in contact with the liquid, whereas the other end member enters into new combinations. The result is albitization of some of the plagioclase which, theoretically, should be accompanied by at least a little albite newly deposited from the liquid. The calcic molecule of the plagioclase may in part enter, together with pyroxene material, into the constitution of the hydrous phase, amphibole. There will thus be a little new-formed amphibole accompanied by transformation of some pyroxene into amphibole (uralite). Under other conditions, the calcic molecule of the plagioclase forms the hydrous zoisite, or acquiring iron from pyroxene, the related epidote, and, instead of amphibole, chlorite may appear. The only feature of these reactions that has any suggestion of simplicity is the great stability of albite. The results of the action of these last residual liquors in basic magmas are, then, albitization, uralitization, saussuritization, chloritization, and epidotization, all occurring as late-magmatic processes. In cases of extreme fractionation in the way of enrichment of water in residual solutions zeolites may form, as well as minerals containing hyperfusibles other than water—datolite, thaumasite. Possibly some addition of water from extraneous sources is necessary for abundant formation of zeolites. It should be noted, too, that probably all of the above processes may be brought about as the result of access of extraneous solutions. Late-magmatic processes and secondary changes may thus be indistinguishable in some examples.

Under some conditions, perhaps involving addition of extraneous water, it may be possible to develop a liquid practically unfractionated with respect to its silicate ingredients but sufficiently enriched in water that all or nearly all the crystallization of the essentially gabbroid magma takes place at the stage where hornblende is the stable phase. The liquid crystallizes as hornblende almost exclusively. There is thus developed a hornblendite which has a sort of pegmatitic relation to gabbro. It is even urged by some writers that the amount of water present in some instances may be so great that nearly all the crystallization takes place at the time when albite and epidote are the stable expression of gabbroid composition and that a rock consisting of albite and epidote is formed (helsinkite) by direct crystallization from a magma. For hornblendite, and especially for helsinkite, it is probable that a hydrothermal-replacement origin is the true explanation, in most, if not all, cases.

Effects of High Concentration of Hyperfusible Constituents in Granitic Magma

When crystallization is characterized by high fractionation of the silicate ingredients the liquid acquires a relatively high content of hyperfusibles only when it has become much more siliceous and alkalic; let us say, granitic. The magma will become saturated with certain crystalline phases into whose composition the hyperfusibles enter. These will be hydrous phases, since water is the dominant hyperfusible. At the granitic stage the hydrous phase is mica. Other hyperfusibles besides water, notably fluorine, enter into its make-up and in its general composition it otherwise reflects the character of the liquid from which it forms. Thus in its potassic character it is strongly contrasted with the hydrous phase of earlier stages of the reaction series, hornblende. The formation of mica takes place by direct precipitation from the liquid and by reaction with phases crystallized earlier, which are thus partly transformed into mica.

Generally speaking it is in the granitic and related magmas that the greater concentrations of the hyperfusibles come into being. If all of the liquor remains uniformly distributed in the interstices of the mass from which it formed it may be completely consumed by reaction with the already crystallized ingredients of the mass.7 The principal product of the reaction will be mica but minute quantities of other minerals will necessarily form in order to account for other hyperfusibles. There may thus be a little calcite, a sulphide, and other minerals carrying hyperfusible constituents.

The late residual liquid of a granite, which we have pictured as capable of being used up by reaction with early-formed silicates, is made up principally of the low-melting silicates and associations of silicates, alkaline feldspar, quartz, muscovite, together with a considerable concentration of water and other hyperfusibles, and also such elements as are entrained in virtue of their property of forming low-melting compounds with the hyperfusibles or, indeed, with any constituent present, the alkalis, silica, etc. This liquid is of the nature that is ordinarily referred to as pegmatitic. In it there is a higher concentration, relative to that contained in normal magmas, of all the substances that enter into mineral deposits of magmatic origin except those that, in virtue of their entry into solid solution in early-formed minerals, may have been partly or largely removed. Among these latter may be mentioned chromium and less definitely nickel.

The crystallization of such highly aqueous magmatic residua has been considered by Morey and by Niggli. Morey decided that under deep-seated conditions the pressure would be adequate to prevent boiling and that there would be continuous passage, as crystallization proceeded, from magmatic melt to hydrothermal solution without the intervention of critical phenomena. He has also considered the behavior under less deep-seated conditions where boiling occurs, particularly with reference to the development of explosive phenomena in volcanism. Niggli has treated the subject especially in its bearing on the formation of mineral deposits. He likewise concluded that under deep-seated conditions the external pressure would be adequate to prevent boiling, but that, as further concentration of volatile constituents goes forward during crystallization, critical phenomena will be exhibited. This is, in its theoretical aspects at least, a very different matter from boiling, though whether any practical criteria can be adduced whereby its results might be distinguished from those incident to boiling is not so certain. The distinction lies in the fact that when critical phenomena occur only one phase is present other than the solid phases. This phase is developed from the liquid phase by a process involving no discontinuity in change of properties, though it is in many of its properties a gas-like phase. The familiar differences between gas and liquid no longer exist under critical conditions. In the case of boiling, two distinct phases are present in addition to the solid phases and they can be definitely referred to as liquid and gas. Niggli attaches great importance to the occurrence of critical phenomena and refers to that stage of mineral development from magmas as the pneumatolytic stage, thus giving to the term a rather specialized meaning which, if rigidly adhered to, would exclude from the category, pneumatolytic, the processes for which it was originally proposed. It would seem that, whereas the meaning of the term might reasonably be expanded so as to include the Niggli usage, it is hardly desirable to confine it to that significance.

Niggli also treats the course of crystallization under less deep-seated conditions where actual boiling occurs and gives some attention to the distillation process that results.

The investigators mentioned, and others, including the writer, were led to state conclusions regarding the phase equilibria in question at a time when there were few experimental data bearing very directly upon the question. It was necessary to use the results of phase equilibrium that offered no close approach to the natural system in composition or prevailing conditions. Recent experiments have done something to fill this gap and appear to indicate, in part, modification of the older conclusions, and in part, a choice among them.