The Witwatersrand 'basin' is the largest known gold province in the wo
rld. The gold deposits have been worked for moren than 100 years but t
here is still controversy about the ore forming process. Detailed petr
ographic studies often reveal that the gold is late in the paragenetic
sequence, which has led many researchers to propose a hydrothermal or
igin for the gold. However, observations, such as the occurrence of ro
unded, disc-like gold particles next to irregularly shaped or idiomorp
hic secondary gold particles in the same sample, suggest an initial de
trital gold source within the Witwatersrand strata. Mineral chemical a
nd isotopic data, together with SEM cathodoluminescence imaging and fl
uid inclusion studies, provide evidence for small-scale variations in
the fluid chemistry - a requirement for the short-range mobilization o
f the gold. The existing data and observations on the Witwatersrand ro
cks support a model of hydrothermally altered, metamorphosed placer de
posits, with at least two subsequent gold mobilization events: hydroth
ermal infiltration in early Transvaal time (2.6-2.5 Ga) and during the
2.020 Ga Vredefort impact event.