Portal de Periódicos da CAPES

Upper mantle composition inferred from laboratory experiments and observation of volcanic products

1970; Elsevier BV; Volume: 3; Linguagem: Inglês

10.1016/0031-9201(70)90061-0

1872-7395

M. J. O’Hara,

earthquake and tectonic studies

The significance to geophysicists and geochemists of recent developments in petrological thought on magma origins and partial melting of the upper mantle are reviewed and new arguments favouring a particular upper mantle composition are presented. A garnet-lherzolite-in-kimberlite model of upper mantle composition is satisfactory. Erupted lavas are not primary but have left behind substantial volumes of igneous eclogite accumulates in the upper mantle. Geochemical calculations which ignore the effect of this fractionation on the reaction relationships encountered in partial melting yield invalid conclusions. The depth of magma generation in the upper mantle may be much greater than 100 km. The present state of hypotheses of andesite generation suggest that the availability of water in the upper mantle may have controlled the early development of the continents but too many assumptions are involved to justify any firm inferences on this basis about differences in mantle composition under continents and oceans.