Pure metals at a low temperature have an exceedingly anisotropic conductivity tensor in high magnetic fields, and one which is commonly strongly dependent on the orientation of the field relative to the crystal axes. A polycrystalline sample is therefore a mixture of very different species, and a general account of what is involved in calculating the bulk properties of polycrystalline copper serves to illustrate the characteristic features that recur in studies of magnetoresistance. It has probably been taken for granted too readily in the past that carefully prepared single‐crystal samples are homogeneous, and such puzzling observations as the material defects than from any intrinsic property. In such a case as this, defect size and free path may be comparable, generating a peculiarly difficult set of problems. The aim of the talk will be to set the scene for more specialized contributions rather than go into the details of any one aspect.