EARLY ARCHEAN (GREATER-THAN-3.2 GA) FE-OXIDE RICH, HYDROTHERMAL DISCHARGE VENTS IN THE BARBERTON GREENSTONE-BELT, SOUTH-AFRICA

Ironstone pods (Fe2O3 = 72.5-97.2 wt%), interpreted to have formed by early Archean, sea-floor-related hydrothermal activity, are described from the ca. 3.5-3.1 Ga Barberton greenstone belt, South Africa. Most of the pods are elliptical in shape and have their longest dimensions subparallel to the local stratigraphy. They rest on silicified ultrama fic rocks and are overlain by ferruginous shales, silicified sandstone s, other coarse clastics, and barite horizons. The ironstone pods grad e along strike into laminated iron-oxide facies banded iron-formation that is inferred to represent periodic discharge of iron-oxide- and si lica-rich flocculates from the pods. The massive texture of the ironst one pods and their lack of internal sedimentary features suggest that they formed directly on the sea floor; relict hydrothermal discharge c himney structures have been recognized. Honeycomb-like cavities and po ssible fluid-flow channel textures attest to primary porosity. The iro nstone pods are dominated by massive, locally coarse-grained, specular hematite and goethite, with lesser amounts of quartz and an X-ray amo rphous Fe-Al-bearing silicate. The pods are distinct from typical mass ive sulfide gossans in total trace-element concentrations, including p recious and base metals, indicating a different origin. In contrast, t he ironstone pods have similarities in major oxide and trace-metal con centrations to iron-oxide deposits presently forming, in part, from lo w-temperature hydrothermal fluids on the sea floor (for example, Red S eamount iron-oxide deposit). Fluid-inclusion studies on quartz show do minant type I primary fluid inclusions with salinities of 4.7 to 15.8 wt% NaCl equiv. and homogenization temperatures (T(h)) of approximatel y 90 to 150-degrees-C; no evidence for boiling is seen. Type I fluid i nclusions are most likely dominated by NaCl-CaCl2-H2O, although their measured eutectic temperatures are consistent with the presence of FeC l2; these inclusions represent the end-member hydrothermal fluid. More saline (24-29.6 wt% CaCl2 equiv.), lower temperature (T(h) = 33-109-d egrees-C), type II and IIa fluid inclusions represent pulses of a hydr othermal fluid of NaCl-CaCl2-FeCl2 (or MgCl2)-H2O, or possibly FeCl3-b earing composition, mixing with ambient sea water in the case of type II inclusions. The various fluid types are interpreted to be indicativ e of intermediate compositions of an evaporitic brine. The lowest sali nity measurement for a type I inclusion of 3.1 wt% NaCl equiv. provide s a constraint on the salinity of unenriched (by evaporation) sea wate r. The T(h) data give a minimum trapping pressure of approximately 7 b ars, which equates to a minimum sea-water depth of approximately 60 m. The lack of precise pressure estimates has not enabled calculation of maximum water depths above the pods. Quartz deltaO-18 analyses averag e 17.3 +/- 0.6 parts per thousand (1sigma; n = 5). Calculated deltaO-1 8(H2O) for two samples are -1.41 parts per thousand and -1.2 parts per thousand, respectively. These estimates are within error of the postu lated Barberton early Archean sea water value of approximately 0 parts per thousand. Quartz/hematite-goethite mineral pairs yield temperatur es inconsistent with fluid-inclusion T(h) data, suggesting isotopic di sequilibrium.