Microthermometry, laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS), and Raman spectroscopy have been used to determine the temperature, apparent salinity, and composition of individual fluid inclusions in adularia-sericite Au-Ag epithermal veins from the Karangahake, Martha, Favona, and Waitekauri deposits, southern Hauraki goldfield, New Zealand. Quartz veins contain colloform to crustiform bands that alternate with coarse-grained quartz and amethyst. The ore mineralization occurs only in colloform to crustiform bands.

Analyses of individual fluid inclusions by LA-ICP-MS identify Na as the most abundant cation, together with variable concentrations of K, Ca, Rb, Sr, Sb, and As. Rare inclusions have detectable Li, Al, and Ba concentrations, although recorded Al concentrations with values up to 231 ppm in Al-free quartz may reflect an accidentally captured mineral phase rather than fluid itself. The Na content ranges from ~260 to 10,200 ppm for inclusions in quartz and ~9,700 to 13,700 ppm for inclusions in amethyst. Antimony is the second most commonly detected element in both quartz- and amethyst-hosted inclusions; this element is also detected in the host mineral. Concentrations of Sb and As range from 0.3 to 988 ppm and from 3.33 to 418 ppm, respectively, and are most commonly detected in inclusions from the Karangahake and Martha deposits. The poor correlation between the Na content with either Sb or As suggests that Sb and As were transported as neutral hydroxyl complexes of Sb(OH)3 and As(OH)3. Both Au and Ag occur at concentrations that are less than their respective detection limits (ppm).

Geochemical modeling of the microthermometric and LA-ICP-MS data obtained from individual fluid inclusions suggests that fluids responsible for the quartz deposition were neutral to alkaline and that adiabatic boiling is the most effective mechanism for both gold and silica precipitation. The presence of single-phase vapor-only fluid inclusions in some mineralized samples indicates that local flashing may have contributed to deposition of Au and Ag.

Assuming adiabatic boiling under hydrostatic pressure, samples from the Karangahake deposit (Maria vein) were deposited from low-salinity fluids (<3.9 wt % NaCl equiv) at temperatures between 225° and 262°C and at depths of 270 to 575 m below the former water table. The average deep reservoir fluid temperature estimated from the Na/K geothermometer is 287°C, and the steam loss during boiling ranges between 8 and 17%.

Fluid inclusions in quartz from the Martha deposit trapped dilute fluids with salinity less than 1.7 wt % NaCl equiv. The coexisting liquid-rich (homogenization temperature, Th = 189°–225°C) and vapor-rich inclusions (Th = 205°–243°C) suggest formation at depths of 200 to 400 m below the water table. According to the Na/K geothermometer, the deep reservoir fluid temperature was near 295°C, and the steam loss during boiling ranged between 15 and 23%. Pseudosecondary inclusions in amethyst display salinity around 4.0 wt % NaCl equiv and homogenization temperatures between 218° and 241°C. Secondary inclusions are slightly more dilute (3.2–4.2 wt % NaCl equiv), with homogenization temperatures between 213° and 242°C.

Fluid inclusions in quartz from the Waitekauri deposit homogenize from 210° to 265°C and contain less than 1.2 wt % NaCl equiv. A thin quartz vein that occurs between the Jubilee and Scotia deposits contains coexisting liquid- and vapor-rich inclusions; their homogenization temperatures indicate a formation depth of 300 m below the former water table. The calculated deep reservoir fluid temperature is around 283°C and the steam loss is estimated to be between 13 and 18%.

LA-ICP-MS analyses show that in some cases different fluid inclusion assemblages (FIAs) within a single sample trapped fluids with variable chemistries. These differences likely reflect modification of a single parent fluid through mineral dissolution and precipitation, water/rock interactions, boiling and vapor loss, conductive cooling, and mixing.