The internal behaviour of the Earth is controlled by convection hence knowledge of mantle viscosity is critical to our understanding. The viscosity of the mantle on large scales has been measured from post-glacial rebound and geoid inversions. These have shown that there is a significant contrast in viscosity between the upper and lower mantles which has been invoked to explain the varied behaviour of subducting slabs at the 660 km discontinuity. The slabs show a range of behaviour from ponding just above the 660 km discontinuity to penetrating the discontinuity and thickening as they enter the lower mantle. It is apparent therefore that the size and impermeability of the 660 km viscosity contrast affects the style of mantle convection and cooling rates but the resolution of the geophysical observations is not sufficient for a complete understanding of the fate of material around the 660 km discontinuity. Here we propose to measure the rheology of the minerals and assemblages present in the transition-zone and lower-mantle and so provide the constraints needed for a fuller understanding of the viscosity contrast between the upper and lower mantles.