ORE-DEPOSITS ASSOCIATED WITH MAFIC MAGMAS IN THE KAAPVAAL CRATON

Mafic and ultramafic magmatism played an important role in the 3.5 Ga long history of the Kaapvaal craton. The oldest (3.5 Ga) greenstone be lts contain mafic and ultramafic volcanics that erupted in an oceanic environment, probably in oceanic plateaus. Then followed a series of c ontinental flood basalts, from the similar to 3.4 Ga old Commondale an d Nondweni sequences, to the 180 Ma Karoo basalts. The history was dom inated, however, by the emplacement, 2.1 Ga ago, of the Bushveld compl ex, an enormous layered ultramaficmafic-felsic intrusion. Three types of ore deposits might be found in such a sequence: Ni-Cu-Fe sulfides i n komatiites of the greenstone belts; ''Noril'sk-type'' Ni-Cu-PGE depo sits in the Karoo and other flood basalts; and deposits of Cr, platinu m-group elements (PGE) and V in the Bushveld and other layered intrusi ons. Only the latter are present. It is tempting to attribute the abse nce of komatiite-hosted deposits to the specific character of the ultr amafic rocks in Kaapvaal greenstone belts, which are older that the 2. 7 Ga komatiites that host deposits in Australia, Canada and Zimbabwe, and are of the less-common ''Al-depleted'' type. However, a review of mantle melting processes found no obvious connection between the chara cter of the mantle melts and their capacity to form ore deposits. The lack of this type of deposit may be due to differences in the volcanic environment, or it may be fortuitous (the Barberton and other belts a re small and could fit into deposit-free parts of the much larger Aust ralian or Canadian belts). Still more puzzling is the absence of Noril 'sk-type deposits. The Karoo and older flood basalt sequences appear t o contain all the important elements of the volcanic sequences that ho st the Siberian deposits. It is now recognised that these deposits for med through the segregation of sulfide from magma flowing rapidly thro ugh conduits en route from deeper magma chambers to the surface. An ex ploration approach aimed at understanding the fluid dynamics of such s ystems seems warranted. Although the Bushveld intrusion has been studi ed for decades and its deposits are taken as type examples of magmatic mineralisation, the origin of its PGE deposits remains unclear. Opini on is divided on the relative importance of sulfide segregation from m agma filling a large chamber at the time of emplacement, and the scava nging of PGE from fluids circulating through cumulates at a late magma tic stage. Answers to these questions may come from studies designed t o gain a better under-standing of the mechanisms through which the mag ma chamber filled and solidified.