In contrast to sedimentary and igneous rocks, which form under near-ambient conditions and high-T, respectively, in metamorphic rocks we have in addition to the increasing temperature from slate to gneiss also an increasing pressure which has to be subdivided into two components (1) hydrostatic pressure or confining pressure and (2) directed or oriented pressure (=shear stress). Changes in pressure are especially decisive if volatile components such as water, carbon dioxide or halogens are involved in the process. The layered structure and separation into melanosome (biotite, chlorite..) and leucosome (feldspar, quartz) may have different origins. It runs the gamut from a relic sedimentary banding reflecting the original structure of the sedimentary parent material , e.g., in metagraywackes, through metamorphic differentiation of gneissic metamorphites into felsic mobilization known from migmatitic rocks of injection gneiss or arterites at the brink of granitization. The distortion of the layered structure by idioblastic growth of minerals depends on the stress executed on the metamorphic rocks and the position of the minerals in the crystalloblastic series based on lattice parameters inherent to the mineral with the two end members (high-low) rutile, magnetite on one end and quartz and feldspar on the opposite end of the series (see also stress minerals and anti-stress minerals).

It is an approximation of mine mainly based on my observations and experience gathere as a field geologist and petrographer. On a micro- or nano-scale there are certainly still some further interesting answers to be expected.

H.G.Dill

felzic and mafic minerals

Gneiss with dark-and-light bands is similar to zebra dolomite which is a combination of dark-and-light bands with the knowledge of the difference in environmental conditions of their formation.

But in your question is a major question whether dark band is predominant or light band? To put it simply, I go back to the example of zebra dolomite. What is the philosophy of the zebra striped which is named after the zebra dolomite?

In the book of "How the Zebra got it Stripes and Other Darwinian just so Stories", Leo Grasset describes the stripes well, and I think it can help to find The answer from a deeper perspective.

Sumit, this is a good question. I can see easy answers for some metasediments that show sedimentary lamination that might transform into metamorphic layering. A schist that has been deformed twice may develop crenulations that become gneissic layering. But exactly how does a homogeneous granite become a granite gneiss with its distinctive light and dark layers?

The alternation of clays and sandstones during metamorphism formed quartz-biotite gneiss, where quartz-light and biotite is dark

I agree that it can either be differential diffusion of Fe, Al, and Mg to form mafic minerals versus diffusion of Na, K, Ca, Cl, and F to form felsic minerals or primary layering of former sediments exhibiting graded bedding of sedimentary particle in sands, silts, and clay in lamellae containing different amounts of Fe, Al., and Mg in mafic minerals versus particles containing greater amounts of Na, K, Ca, and Si in felsic minerals.

If heterogeneities existed in the protolith, tectonic transposition can be a major factor. e.g.

Myers, J., 1978. Formation of banded gneisses by deformation of igneous rocks. Precambrian Res. 6, 43–64. doi:10.1016/0301-9268(78)90054-2.

And illustration in

Ramsay, J.G., Graham, R.H., 1970. Strain variation in shear belt. Can. J. Earth Sci. 7, 786–813. doi:10.1139/e70-078.

Normally, the presence of light and dark band of gneiss is due to the seggregation of felsic and mafic mineralization.

The alternating darker and lighter bands in Gneiss are because of mafic and felsic minerals therein.

It happens that the heat had affected both minerals light and dark as we knew that before as it was a granite with light coloured minerals such as feldspars, quartz and micas. Moreover the existence of dark minerals biotite and amphibole as weak had experienced metamorphism whereby foliation created these bands. Gneisses are subjected to dynamic metamorphism that allows them to be well layered by squeezing out the fluid in them.

because it is geochemical differentiation

I wonder if density is a factor in this differentiation. For example, biotite is light (in weight) and quartz and feldspars are heavier. But if you had biotite and amphibole together in the dark bands, then density is not an important factor.

When geochemical differentiation occurs, minerals are formed where the concentration of chemical elements has stopped.

Sand shale siltstone volcanic ash, marine pelagic material. I did a comprehensive layer by layer examination of gneiss in Colorado I found a place where I could walk the metamorphic grade and degree of gneissification back to very fine grained “protolith” within only a few thousand feet. It was part of my thesis in metamorphic stratigraphy of part of the affront Range of Colorado. What I discovered was that the dark layers always had hornblende and no biotite and the light layers had biotite and no hornblend. There was actually a metamorphosed massive sulfide deposit and a clay foot wall that had become sillimanitic. The bands are relicts of ancient bedding...at least that is how it was interpreted in the banded metamorphic units I mapped. I am sure there are geochemical and tectonic schlieren but go with the bedding concept even in granulite fancies. I enclose a fully transformed clay bed from a geothermal system. It is residual mesothermal leached clay metamorphosed to anthophyllite and dickite. The bedding is a geochemical relict...just as if it were sapropelic soil. The material was an andesite blocky mud from under a volcanic lake probably. When I saw the sillimanitic material in the front range I did not recognize it’s being something similar but at amphibolite grade.

Steve Johnson. Your explanation may well be relevant for in-situ high grade metamorphism of sedimentary rocks, but how does a gneissic layering develop in a rock that was originally an intrusive granite? For example a granite gneiss that has no hornblende, but shows quartz-feldspar-rich laminae vs biotite-rich laminae.

To comment on Brian P. J. Stevens 's observation it would appear that quartz-feldspar bands have similar density versus biotite which is much lighter and "floated" on top of the other layer.

I think these papers could help you;

They're available on the ResearchGate Server for download.

Jorge, you are getting closer to the real story.

Sumit Bhowmick Here, Quartz is typically abundantly found in gneiss. The bands that form on gneiss rock are due to the various rocks that are a part of its make-up. The rock is further characterized by its alternating light and dark bands of minerals. It forms from volcanic rock, shale, or granitic.