When we turn to the igneous rocks of older geological periods the same contrasted suites can be recognized. The midland valley of Scotland has been the theatre of igneous action both in Lower Devonian and Carboniferous times. The igneous assemblages of the two periods show, however, a striking contrast. The earlier lavas are distinctly of the Pacific type including andesites, basalts and rhyolites, while the Carboniferous extrusions and dike rocks include trachytes, teschenites, picrites and monchiquites—an Atlantic assemblage. Petrographical provinces are therefore clearly not permanent—a point to which we shall later return.
Each of the great suites—Pacific, Atlantic and Mediterranean is characterized by a set of rock types which reflects the chemical and mineralogical peculiarities. Arranged in tabular form the typical assemblages of the three suites are as follows :— The recognition of these three suites and their distinctive prod ucts constitutes without doubt the first step towards a natural or genetic classification of igneous rocks.
Study of the geological relations of igneous rocks clearly reveals that there is a general correspondence both in time and space between igneous action and movements of the earth's crust, whether these be simply vertical movements of elevation and de pression or tangential movements of folding and overthrusting which build up great mountain chains. As in crustal movements, so in igneous action, periods of activity have alternated with periods of quiescence. The movements of compression which led to the folded mountain chains fringing the Pacific were accompanied and followed by widespread vulcanicity and the movements of depres sion, which in Tertiary times led to the formation of the Great Rift Valley of East Africa, were attended and followed by a great suite of volcanic eruptions. The interdependence of igneous action and crust movement is even of a still more intimate kind, for not only the distribution of igneous rocks, but the distribution of different kinds of igneous rock is seen to stand in unmistakable relation to the great tectonic structures of the earth. The ex amples referred to above are illustrative of this conclusion. The
volcanoes of the Pacific cordillera erupted calcic andesites and basalts, and are associated with granodiorite and quartz-diorite intrusions ; the eruptions associated with the African Rift valley are in contrast, being of alkaline type and including soda-trachytes, phonolites and related alkaline lavas. Whether attention is con fined to Tertiary igneous rocks or extended to those of earlier date, the conclusion is enforced that rocks of the Pacific suite are closely associated with folded mountain chains and appear in regions subject to movements of compression and lateral thrust, and that the Atlantic or alkaline suites are associated with regions of subsidence, subject to tension and faulting. That a particular region at successive epochs may be the theatre of action of con trasted igneous suites plainly forbids the assumption that the dis tribution of rock types is due to initial heterogeneity of the crust. Rather it would appear that calcic and alkaline suites are ultimately derived from the same primitive magmas. The manner of their evolution is a problem of great complexity.
The mode of origin of the massive fringes of pegmatite mantling the "older granites" of the south eastern highlands of Scotland may perhaps provide the initial clue to the actual mechanism in volved. By Barrow these pegmatite fringes are conceived to represent the acid residual liquid squeezed out from partially crystallized granite magma under the influence of crustal stress. To an analogous process operating at deeper levels, and on a grander scale, Harker has made appeal to account for the separa tion in a horizontal sense of alkaline and calcic magmas from some common deep-seated stock.
Physical Chemistry of Igneous Rocks.—Natural rock mag mas are polycomponent systems with seldom fewer than six or seven oxides, and no adequate discussion of their physical chem istry is at present possible. It is the function of the petrologist to elucidate the nature and behaviour of these complex systems under different physical conditions. In the last 25 years much progress has been made on the experimental side and the founda tion for an extended exploration of the crystallization history of silicate magmas has now been laid. The main contributions have come from the Geophysical Laboratory at Washington. Investi gations initially concerned with the perfection of experimental methods of attack have been followed by a physicochemical study of simple oxides and their combinations. As yet these investiga tions have been almost wholly confined to dry systems but an attack on wet systems with volatile components is already in progress. Systems have now been investigated which approach in their complexity some of the simpler types of igneous rocks and the results obtained justify the conclusion that the fusion and solidification phenomena of igneous rocks are capable of systematic treatment. Starting with pure chemical substances the melting points, dissociation temperatures and inversion points of numerous rock-forming minerals have been determined, and the equilibrium phenomena of two and three component systems in volving the commoner oxides of igneous rocks have been quanti tatively elucidated. This study has embraced the six binary and four ternary systems of the oxides, CaO, MgO, and portions of the quaternary system itself have been explored. The investigation of mineral inversions has provided data throw ing light on the temperature of consolidation of igneous magma. Such of these inversions as are enantiotropic or reversible and take place without appreciable lag can be appropriately used as geological thermometers. The inversion a quartz 575(3 quartz is a striking example, as it is accompanied by a significant volume change whereby it is possible to recognize whether quartz at ordi nary temperatures has crystallized as a or 3 quartz. The criteria include complicated twinning and fracturing. It can thus be shown that the quartz of granites has consolidated as 13 quartz and therefore at a temperature above 575°. The quartz of many pegmatites and igneous quartz veins on the other hand has consoli dated as a quartz and therefore below 575°. (Note: some writers use a and j3 in the inverse senses.) The inversion pseudo-wollas tonite provides another case, from which it is concluded that wollastonite occurring in rocks has crystallized below 1,190°.