Synopsis: Spotted eye-like structures, made up of densely packed biotite ‘pupils’ mantled by quartz and feld-spar, constitute about 10% by volume of a high-grade quartzofeldspathic gneiss located on the road to Purnamoota Homestead about 10 km north of Broken Hill. These spotted structures have an uneven distribution within rocks exposed at this locality. In some parts of the outcrop they tend to be closely spaced (fig. 1), whereas elsewhere they are less abundant and the spotted gneiss merges into an essentially homogeneous gneiss composed of quartz (34%. by volume), plagioclase (24%), Kfeldspar (17%), biotite (20%), and garnet, muscovite, and opaques (5%). This homogeneous gneiss is believed to have developed into a ‘matrix’ gneiss phase holding the ‘eyes’ which have a felsic mantle to mafic clot ratio of 4:1.

The whole eye-like structure appears to be a distorted prolate spheroid with its axis set close to the foliation plane of the gneiss. The internal biotite clot forms another prolate spheroid with its axis inclined at approximately 70° to that of the enveloping quartz-feldspar mantle. Although most of the eyes are isolated within the matrix, some are linked to hold three or four biotite aggregates (fig. S1) and rarely a vein-like patch which contains some ten or more partly linked spots of biotite may be found. In addition to the biotite, xenoblasts of almandine occur within some of the eye-like structures.

Whole-rock chemical analyses (including both major oxides and trace elements) show that the homogeneous gneiss and the spotted gneiss are very similar. The most notable feature of the geochemistry of the spotted rocks is the close chemical correspondence between the matrix and the whole eyes. Moreover, the compositions of corresponding minerals in these phases are very similar. For example, plagioclase is consistently about An_35–40, the biotites are all very iron-rich (Fe/(Fe + Mg) ≈ 0.87) with compositions near siderophyllite, and K-feldspar has the composition Or₉₄Ab₆.

It has been suggested that the biotite spots formed by alteration of garnet probably during a retrograde metamorphic event. However, studies of the microstructure of the biotite and garnet in the spotted gneiss show that biotite of the spots does not replace garnet. Furthermore, the chemical similarity between the spots and the matrix is best explained by an isochemical rearrangement of components of the matrix phase to form the eyes. Thus, a more likely origin is best related to the displacement of matter along chemical potential gradients possibly induced by deformation in the rock system—a process described by the term metamorphic differentiation.

Veins in the Purnamoota Road gneiss are of two main contrasting types—regularly-disposed veinlets which are composed almost entirely of K-feldspar and quartz, and irregularly shaped discontinuous trondhjemitoid variants rich in biotite spots and carrying rare garnet porphyroblasts. The field relationship between these two main types of veins is difficult to discern. In some parts of the out-crop the veins occur adjacent to one another and locally appear to merge. Veins with modal (and chemical) compositions intermediate between the spotted and the K-feldspar-rich veins have also been recorded from the outcrop.

Despite a wide whole-rock chemical variation, the minerals in all the veins are similar chemically and compare closely with corresponding minerals in the host gneiss. One exception may be the plagioclase in veins rich in K-feldspar where normative calculations indicate a composition near An₁₄.

Interpretation of the trace element data counts against the veins having formed by partial melting and the variable K-feldspar contents place both main vein types well away from ternary minima in the system Q-Ab-Or. Calculations based on the compositions of co-existing biotite and garnet in the host quartzofeldspathic gneiss and in associated prograde pelitic schists indicate an equilibrium temperature of 650° ± 50 °C at pressures between 3 and 4 kilobars. This fact, together with trace element data and the calcic nature of the plagioclase in the homogeneous gneiss and the matrix phase, places some doubt on the suggestion that partial melting has played a signficant role in the genesis of the veins.

Although mechanisms of vein formation such as igneous injection or metasomatism are considered as other alternatives for vein formation, we suggest that metamorphic segregation (as proposed for the origin of the spots in the gneiss) will also account for the development of the veins.