It has been shown that the phenomena associated with the re action-relation in dry melts appear among the mineral phases of magmatic melts and can be interpreted in like manner. In the light of laboratory experiment available petrographical data per mit the conclusion that the pyroxenes, hornblendes and biotites form a discontinuous reaction series, pyroxene reacting with liquid to form hornblende, hornblende to form biotite. Each of these groups in itself as solid solutions is a continuous reaction series. It is in reactions of this kind that the ordered differentiation sequence gabbro-diorite-quartz diorite-biotitegranite, character istic of the calc-alkaline rocks now finds explanation ; the con tinual offset in composition of residual liquid in which water accumulates being accomplished by appropriate removal of crys tals, either by gravitative settling or squeezing out or filtering of interstitial liquid by earth movement during crystallization.
Volatile constituents are believed to play a prominent part in the evolution of rocks of the alkaline (Atlantic) suite. They are responsible for the desilication of the alkali aluminium poly silicates. Both carbon dioxide and water appear to be the princi pal agents in this process : Thus, reactions of the type— may represent equilibrium relations in a residual liquid rich in water (such as the granite stage of differentiation). If the pre cipitation of alkali-felspars, biotite and quartz is followed by their withdrawal, the concentration of nepheline molecules will lead finally to the separation of this constituent. Assemblages of nephe line syenites, urtites and ijolites may be differentiated in this way. The capacity of CO2 for desilicating polysilicates is attested by laboratory experiment ; and it is not improbable that carbon ates, by a process of absorption of limestone in a liquid already alkaline, have been effective in the generation of some alkaline rocks, especially those containing such minerals as calcite, can crinite, melanite and melilite.
More obscurity surrounds the origin of the potassic or leucitic rocks. Their spatial arrangement in relation to the Pacific and Atlantic suites, no less than their chemical characters reflects their intermediate position. They are characteristic of the region lying between the orogenic zones and the foreland or regions of sub sidence. The position of the potassic Roman province in rela tion to the Alpine orogenic zone on the one hand and the alkaline (Atlantic) basin of the west Mediterranean illustrates this rela tion very clearly. The formation of leucite by dissociation of orthoclase (incongruent melting) is an important reaction in the genesis of some potassic rocks. Normally in the presence of
abundant water under cover, the desilication of the polysilicate would appear to proceed a stage further in the formation of the orthosilicate which is precipitated in biotite—usually at the granitic stage—and the generation of leucite would in some cases appear to be connected with a loss of water from a biotite rich liquid (e.g., of the composition of minette) by its intrusion or extrusion near or at the surface. Leucite-bearing rocks are rare as plutonic rocks (the leucite being usually replaced by ortho clase and nepheline). The generation of the potassic series may indeed be dependent on a prior differentiation of a magma to a stage in which biotite is accumulated, as revealed, e.g., in mica rich lamprophyres.
Desilication of potash-felspar appears then as a characteristic reaction in rocks of the Mediterranean suite. Their relation to the Pacific suite is provided in the equilibria existing between potash f elspar, biotite, leucite and olivine ; to the Atlantic suite in the equilibria between alkali-felspars, felspathoids and quartz; in the potassic series desilication of the albite molecule playing a sub ordinate role.
The genetic relations between the three suites are amply attested and enforce the conclusion that no sharp line of division between them is to be recognized. Their source is a common stock magma. Nevertheless we appear to see in the relative distribu tion of the three series upon the face of the globe, the influence of tectonic processes as an external factor exerting a profound influence on the trend of differentiation.
Elementary books on petrology include—A. Harker, Petrology for Students, 6th ed. (1923) ; F. Rinne, Gesteinskunde, 9th ed. (1923) ; J. de Lapparent, Lecons de Petrographie (1923) ; L. V. Pirsson, Rocks and Minerals, 2nd ed. by A. Knopf (1926) ; G. W. Tyrrell, The Principles of Petrology (1926) ; F. H. Hatch, The Petrology of the Igneous Rocks, 8th ed. (1926) ; S. J. Shand, Eruptive Rocks (1927). More advanced works include:—A. Harker, The Natural History of Igneous Rocks (1909) ; J. P. Iddings, Igneous Rocks, 2 vols. (1909-13) and The Problem of Vulcanism (1914) ; H. Rosenbusch, Mikroskopische Physiographie, 4th eel. vol. 2 (1907), and Elemente der Gesteinslehre, 4th ed. by A. Osann (5923) ; F. v Wolff, Der Vulkanismus, 2 vols. (I1)14-22) ; R. A. Daly, Igneous Rocks and their Origin (1914) ; F. W. Clarke, The Data of Geochemistry, Bull. U.S.G. Survey, No. 770 (1924) ; P. Niggli, Gesteins und Mineral-provinzen (1923).