We describe the discovery that the natural remanent magnetisation (NRM) of certain rock formations in Britain that are Eocene or older have directions that differed significantly from the Earth’s present field and from one another. In 1954 the first author, a third year research student in the Department or Geodesy and Geophysics (DG&G) at Cambridge University, observed that the poles corresponding to these old geomagnetic field directions fell consecutively on a path beginning in the Proterozoic in Arizona, swooped across the Pacific Ocean to the coast of eastern Asia and from there northwards to the present north geographic pole by the middle Tertiary. This was the first path of apparent polar wander (APW) based on the NRM of rocks. This path ought to have been reproducible in all continents had they not moved. Thus was born the first successful physical test of Wegener’s Theory of continental drift. The work had its origins several years earlier with the construction in the later 1940s of a very sensitive magnetometer by P. M. S. Blackett at Jodrell Bank, a field station of the University of Manchester. In the summer and autumn of 1951, the second author, a recent geology graduate from Cambridge University, showed that Blackett’s instrument, which had been made for an entirely different purpose, was well suited to measuring the NRM of rocks (palaeomagnetism). In the following years Blackett-type magnetometers became the means by which the British path of APW was observed. Although Creer’s path was based only on nine poles we show that subsequent work fully justifies its use as a starting point for the successful global test of Wegener’s theory that was carried out in the following years. Since then, palaeomagnetic poles and the APW paths derived from them have become the basis for constructing the general frame of reference for palaeogeographic maps, for mapping the past distribution of continents, oceans and orogenic belts, the location of ancient climatic zones and many other applications.