Precision patterning

inclined with respect to the shear plane, the observations could possibly be matched. Roughly horizontal shearing of perovskite associated with a deforming Tonga slab may therefore explain the S-wave observations of Wookey et al. as well as the apparent absence of large splitting from the lower mantle in SKS and ScS phases that travel almost vertically beneath the Tonga subduction zone. The work of Wookey et al. and McNamara et al. suggests that strong anisotropy may be a common feature of regions where subducting slabs enter the lower mantle. Accurately modelling the full waveform interaction of seismic phases with anisotropic structure, rather than just their arrival times, is a formidable task. But applying such approaches to waves that sample the mid-mantle could help to explain the dynamics of mantle convection. ■ Karen M. Fischer is in the Department of Geological Sciences, Brown University, Providence, Rhode Island 02912, USA. e-mail: [email protected]