ART. XVI.-Relations of Subjacent Igneous Invasion to Regional Metamorphism; by JOSEPH BARRELL. (Continued from page 186.) PART III. INTERPRETATION OF DYNAMO-METAMORPHIC FEATURES IN THE ROOFS OF BATHOLITHS IN MOUNTAIN PROVINCES. PRELIMINARY STATEMENT. This article has reached the point where it appears that batholithic invasion is to be looked upon as one of two major factors in the control of the phenomena of dynamometamorphism, the other being crustal deformation. The following part will consequently be a study of significant features and their interpretation as phenomena of magmatic injection, chemical alteration, and lateral compression of batholithic roofs. FEATURES PRODUCED BY MOVEMENTS OF SOLUTIONS AND SELECTIVE CRYSTALLIZATION. Where blocks of biotite granite gneiss are inclosed in a coarse granite, it may frequently be observed that a concentration of biotite from the surrounding magma has taken place within and around such fragments. Pegmatitic seams in banded gneisses usually show no change at the margins, but may sometimes be seen to have a biotite lining which slightly permeates the walls and gives sharper definition to the seams. This tendency for biotite to separate by selective crystallization from highly aqueous solutions of magma is to be related with that normal sequence of crystallization in magmas whereby the black bisilicates, associated with smaller amounts of other minerals, crystallize out early, followed by the crystallization of plagioclases, then orthoclase, then orthoclase and quartz; each phase overlapping the adjacent ones. Now if the magma is rising through the foliated structure of a roof and especially if the amount of water is large, so that the crystals of any one generation are but a small part of the magma, the result is a vertical separation of the magma by fractional crystallization. The biotite will be strained out in one place and crystallize against biotite. Quartz and feldspar will rise higher and crystallize as pegmatite free from biotite. Where the magma rises through a foliated country rock, it should lead to a pronounced banding in composition, giving banded gneisses. It is in this way, by the movement of solutions along foliation planes, rather than by a mere recrystallization in place, that marked lamellar segregation of minerals in gneisses is most readily explained. Let a massive granite yield to lateral pressure by granulation or recrystallization: the minerals are broken down and reconstructed, but in this reconstruction laminæ of quartz and feldspar tend to alternate with those of mica. Mere molecular diffusion and crystallinic attraction would tend to build up individual minerals elongated in the plane at right angles to compression, but would tend to reconstruct the minerals about each dominating nucleus without extending one kind of mineral into continuous sheets. This kind of massive gneiss is fairly common in those types where granulation has exceeded recrystallization as the agent of deformation, in other words, where regional compression has been strong and crystallizing solutions scanty. Where, however, the solvents are more abundant, movement would take place readily along the foliation direction, very slowly across it, the biotite would be progressively concentrated into layers, the quartz and feldspar would tend to be carried farther along the planes and have the relations of intercrystallized lamina. If the solvents rise far and are concentrated into thicker sheets, it is a form of pegmatization carried on not as a result of a primary crystallization of magma, but as a result of rock-mashing in the presence of rising solutions. Pegmatization, as Van Hise has noted, may occur in association with formations which show no relationship to igneous intrusion. In granite gneisses which show a pegmatitic texture, as in the Hoadley Point, Connecticut, gneiss, the bands of quartz and porphyritic feldspar may be an inch across and separated by continuous sheets of biotite. In gneisses whose lamellar character is due to mashing and recrystallization, the emphasis is put here not merely upon the presence of crystallizers, but upon their passage. The crystallizers are to be interpreted as the rising emanations from deeper seated sources. DEVELOPMENT OF LIT-PAR-LIT STRUCTURE BY FORCE OF CRYSTAL LIZATION. Lit-par-lit injection is that form of magmatic intrusion which takes place where the magma has soaked into a highly foliated roof rock and has resulted in all grada tions of composition from unaltered country rock to pure igneous rock. The penetration of the magma has not been accomplished by the massive invasion of appreciable viscous fluids, but is more suggestive of capillary action and intercrystallization. The thinnest of parallel mica laminæ may be traced throughout their length without showing such crumpling as even the inertia of the least viscous fluid would have given them if the invading magma had all been fluid. From sheets of magma and of schist which are only of crystal thickness, all gradations in width may be traced into wide bands of country rock or equally wide dikes of pegmatite or granite. The phenomena are extensively displayed in Connecticut, and the writer is more particularly familiar with the large field of mixed rock known as the Waterbury gneiss, extending through western Connecticut from Torrington to Derby. Doctor G. O. Smith and the writer studied this area in 1906, and came to the conclusion that the best way for making field maps was, knowing the pure types of cover rock and granite, to estimate for each outcrop the ratio of the two. This would give data for deciding where, and on what basis, formation boundaries should be drawn. [Doctor Fenner25 has described the same features in an independent study of the highlands of New Jersey. He clearly shows that the hydrous magmatic emanations or differentiates may precede the magma lit-par-lit by penetrating small pores where their lower viscosity allows them much more rapid movement than the main magma. This penetration of solutions makes the rock more like magma in composition, as well as conducting magmatic heat in advance of the magmatic invasion, until finally, if the magma advances, it reaches a rock so modified that one would expect it to be readily assimilated.] The great bursting power of freezing water is well known, even when acting between surfaces such as joint planes, which permit free ingress and egress to the water. Becker and Day26 have discussed this power as exhibited by other crystals. The phenomena of feldspathization, as shown in lit-par-lit structure, suggest that for such mixed gneisses this factor should be elevated to a first place. A solution permeates and passes through a foliated rock. The temperature falls as the solutions flow outward. The 25 C. N. Fenner, Jour. Geology, 22, 594 and 694, 1914. G. F. Becker and A. L. Day, Jour. Geology, 24, 313-333, 1916. saturation point is passed and crystallization begins on particular foliation planes. Mica has mostly been left behind in the magma, so the growth is principally by the addition of quartz and feldspar. The continuation of growth requires the pressing apart of the walls of each lamina by the force of the growing crystals. Unevenness of crystal growth keeps the whole in a sufficiently porous condition to permit the passage of more fluid. The expansion on many planes requires some mass movement of the rock, resorption elsewhere, condensation, mashing, or crumpling. A slight accession of heat and gases or pressure in one locality will turn crystallization into solution and resorption, permitting more readily the expansion of other and adjacent parts by intercrystallization. Rock flowage under such conditions must be relatively easy, since the whole is within the temperature range of crystallization and slight changes of equilibrium will turn the balance from one side of the equation to the other. The chief loss of energy is that carried off by the solutions whose continuous passage is needed to maintain the critical conditions. The different minerals are doubtless quite differently susceptible to stress-differences. Although under normal conditions of great pressure but no stress-differences, the micas crystallize before quartz and feldspar, if stress-differences become great, then recrystallization of mica appears to proceed more readily than that of feldspar and quartz. The pressing apart of the magma walls may thus result in a schistosity imposed while the intercrystallization is still going forward. The process of infiltration proceeds until any degree of granitization has taken place. The process is not marked by the development of vein structure; rather each minute seam is like a pegmatite or granite in texture. The crystallization is, therefore, not so much against the walls as in between the walls. As long as the solutions are passing, they are impelled by the greater pressures of the magma below, and the excess pressure over that of the surrounding country rock at any place is expressed by the hydrostatic head needed for overcoming the frictional resistances to flow. This hydrostatic head is an essential factor, for it means that the fluid can support the pressure of the growing crystals and these do not need to resist porosity by mutual pressure at their points of contact. A pegmatite vein may therefore hold a porous texture and yet be solid during the time of growth, in much the same way as a porous sandstone may be strong and yet serve as a passage for circulating waters. The difference, however, is that the hydrostatic head of the meteoric waters is less than that of the sandstone, with the result that the recrystallization of the sandstone eliminates the porosity and develops a quartzite. In the pegmatite, on the contrary, the porosity would be maintained as long as the country rock could yield before an excess of hydrostatic head from the magmatic emanations. The last stages in the development of the crystallization would be a final elimination of the porosity. It would seem that this is a rather fundamental theoretic principle whose existence it is necessary to postulate in order to obtain a workable understanding of the phenomena of regional granitization as shown by lit-par-lit structure. DEVELOPMENT OF BANDED ORTHOGNEISSES AS A RESULT OF SUCCESSIVE INJECTIONS. The foliation of the granite gneisses was once looked upon as an evidence of sedimentary origin. With the recognition of their igneous nature and the evidences of later mashing, a contrary interpretation very generally came in, regarding the foliation as wholly due to rock flowage in the solid state. Observations began to multiply, however, which showed that a very appreciable portion of gneisses owed their parallel structure to flowage while in the fluid or partly fluid state-the protoclastic structure of primary gneisses. This is contrasted with what Becke called the crystalloblastic structure that developed by rock flowage—in the true compression gneisses. Adams especially has shown the importance of deformation during crystallization for the Laurentian gneisses, and Leith in his Structural Geology calls attention to the dominance of protoclastic gneisses. It is desired here to emphasize a somewhat different phase of the subject which may be called intermittent injection and crystallization in batholithic roofs. It seems to be associated especially with regions under differential crustal compression, and gives banded orthogneisses.27 It may grade from the previously described 27 [The nomenclature of the gneisses is in danger of confusion. It is well, following Van Hise, to make the term gneiss one of purely structural significance. Rosenbusch includes in the main division gneisses of igneous |