Much of the current interest in continental drift arose from the discovery of large divergences between the polar wandering curves determined from the palaeomagnetic directions in different continents. The geological lines of evidence for continental drift, each indecisive yet as a whole striking, were thus supported in a remarkable way by quantitative evidence from an entirely separate branch of geophysics. The palaeomagnetic evidence from Europe and North America is of particular interest as almost all geological periods have been sampled in these continents back to 1200 My. A systematic westward displacement, of the order of 20 degrees to 30 degrees, of the polar wandering path from American rocks from that determined from European rocks has been demonstrated. That the mean geomagnetic field is axially symmetrical seems a secure deduction from the mechanics of the mantle-core system. The possibility that the discrepancy could be due to the field having had an axial but non-dipole character during these periods can be dismissed on palaeoclimatic grounds. The palaeomagnetic data can therefore be used to reconstruct the northern hemisphere in pre-Triassic times: Europe and North America were closer together and in low latitudes. Comparison between the Precambrian of the western United States and the Keweenawan system appears to provide evidence of displacements within North America before the world-wide orogenic epoch of 1000 My ago.