Barton Peninsula is an ice-free area located in the southwest corner of King George Island (South Shetland Islands, Antarctica). Following the Last Glacial Maximum, several geomorphological features developed in newly exposed ice-free terrain and their distribution provide insights about past environmental evolution of the area. Three moraine systems are indicative of three main glacial phases within the long-term glacial retreat, which also favoured the development of numerous lakes. Five of these lakes were cored to understand in greater detail the pattern of deglaciation through the study of lacustrine records. Radiocarbon dates from basal lacustrine sediments enabled the reconstruction of the chronology of Holocene glacial retreat. Tephra layers present in lake sediments provided additional independent age constraints on environmental changes based on geochemical and geochronological correlation with Deception Island-derived tephra. Shrinking of the Collins Glacier exposed the southern coastal fringe of Barton Peninsula at 8 cal ky BP. After a period of relative stability during the mid-Holocene, the ice cap started retreating northwards after 3.7 cal ky BP, confining some glaciers within valleys as shown by moraine systems. Lake sediments confirm a period of relative glacial stability during the last 2.4 cal ky BP.

A fossil hydrothermal system on Barton Peninsula, King George Island, Antarctica, formed a series of lead-zinc- and pyrite + native sulphur-bearing epithermal quartz ± calcite veins, filling fault-related fractures in hydrothermally altered volcanic rocks of Eocene age. The lead-zinc veins occur within argillic hydrothermal alteration zones, whereas the pyrite + native sulphur veins are found within advanced argillic alteration zones. Fluid inclusion data indicate that the vein formation occurred at temperatures between about 125° and 370°C (sphalerite deposition formed at 123–211°C) from fluids with salinities of 0.5–4.6 wt.% eq. NaCl. Equilibrium thermodynamic interpretation of mineral assemblages indicates that the deposition of native sulphur in the upper and central portions of the hydrothermal system was a result of the mixing of condensates of ascending magmatic gases and meteoric water giving rise to fluids which had lower pH (<3.5) and higher fugacities of oxygen and sulphur than the lead-zinc-depositing fluids at depth. The δ34S values of sulphide minerals from the lead-zinc veins (δ34S = −4.6 to 0.7‰) are much higher than the values of pyrite and native sulphur from the pyrite + native sulphur veins (δ34S = −12.9 to −20.1‰). This indicates that the fluids depositing native sulphur had higher sulphate/H2S ratios under higher fo2 conditions. Sulphur isotope compositions indicate an igneous source of sulphur with a δ34SΣS value near 0‰, probably the Noel Hill Granodiorite. Measured and calculated δ18O and δD values of the epithermal fluids (δ18Owater = −6.0 to 2.7‰, δDwater = −87 to −75‰) indicate that local meteoric water played an important role for formation of lead-zinc and native sulphur-bearing quartz veins.