As early as the seventeenth century, scientists realised that a pendulum swings more slowly at the equator than it would at the North Pole. Newton predicted that gravity increased with latitude, and that the Earth could not be perfectly spherical. Although various experiments were undertaken to determine the exact degree of this ellipticity, none proved successful until physicist Edward Sabine (1788-1883) embarked on a series of expeditions across the world. Based on pendulum measurements from a wide range of latitudes, from Jamaica to Spitsbergen, his results were very different to mathematical predictions, and far more accurate; Charles Babbage would even complain that they were too good to be true. In this account, which first appeared in 1825, Sabine explains his methodology and presents his findings. His book opens a fascinating window into nineteenth-century geodesy for students in the history of science.