Abstract
A previous formula of the author's has been adapted to give the volume V of the region in which the material is near breaking-point, just prior to an extreme earthquake, in terms of the released energy and breaking-strength near the focus. In the light of recent earthquake energy-magnitude data of Gutenberg, Richter, and Benioff, it is inferred that V is at least of the order of the volume of a sphere of radius 25 km. Comparison with other estimates of the size of strained regions in earthquakes suggests that V may possibly reach the order of the volume of a sphere of radius 50 km. It is suggested that a factor in the occurrence of extreme earthquakes is the existence, in the zone of strain, of strength rather greater than in the majority of major earthquakes of magnitude one unit or more less than the extreme value.
