Petrologic discoveries made over the last ten years have shown that in
regions of continent-continent collision, rocks of the continental cr
ust can be subducted to much greater depths than previously considered
reasonable. Documentation of such extreme subduction has come from di
scovery of diamond, coesite and other high pressure phases; in many ca
ses these phases, metastable at the Earth's surface, are only preserve
d as inclusions in strong, refractory minerals that either have a broa
d pressure range of stability or exhibit very sluggish kinetics of rea
ction to low pressure forms. The Alpe Arami peridotite of the Swiss Al
ps displays extensive exsolution of FeTiO3 rods in the oldest generati
on of olivine. The shape, orientation and abundance of the titanate ro
ds provide strong indication that the phase originally exsolved was th
e orthorhombic perovskite phase stable only at pressures greater than
10 GPa (300 km depth) at mantle temperatures. We show here that the di
slocation substructure of the oldest generation of olivine is younger
than the titanate rods and similar to that observed in peridotites the
world over and in experiments; the slip systems represented are incap
able of producing the unique and unexplained lattice preferred orienta
tion (LPO) displayed by this generation of olivine. We also have condu
cted preliminary experiments to investigate the maximum solubility of
FeTiO3 in olivine. Our results suggest that the solubility of TiO2 imp
lied by the abundance of titanate precipitates may be impossible under
any conditions of olivine stability. On the other hand, the measured
solubility in wadsleyite (beta-olivine) under the conditions of our ex
periments is comparable to that inferred for Alpe Arami olivine. This
latter observation combined with the determination that the titanate r
ods and LPO of this generation of olivine are the oldest features yet
identified in these rocks, leads us to speculate that this massif has
been brought to the Earth's surface from within the mantle transition
zone, at depths of 410-660 km. The only mechanism by which we can envi
sion this to have been accomplished is for the Lepontine gneisses that
now surround the massif to have been subducted to great depth followi
ng collision of Africa and Europe, and to have picked up the peridotit
e on their way back to the surface by buoyant upwelling.