1. Geological observation provides abundant evidence of fractures in the upper crust, the relative displacements of the sides of a fracture being anything from a few centimetres to thousands of metres. It is generally believed that earthquakes are produced in these displacements, but little attention seems to have been given to the quantitative relations involved. The geologist sees the record of the total displacement, but has no direct information about how long it took to occur, or whether it took place all at once or by several stages. Except in cases where an earthquake has been associated with fracture of the outer surface, and we can say how much displacement occurred up to the time of inspection, he has no means of knowing whether it took seconds, hours, years, or geological periods to form. Even in these rather rare cases the displacement is not more than a few metres, and extrapolation to the larger known faults is impossible. Seismology on the other hand is well adapted to the study of sudden shifts, but the instruments do not record large movements spread over a long time; thus the geological and seismological data are largely complementary. It is known also that a large earthquake is often followed by a swarm of small ones, called aftershocks, apparently from the same focus and continuing for, possibly, several months; there may be a thousand in a single series. The geological displacement observed long afterwards will include those due to all the aftershocks, supposing, as is generally done, that they all are due to shifts on the same fault.