Propagation of glacier surges has been discussed in terms of stresses acting in the three major zones of the surge. The steeply sloping front of a surge appears sufficient to explain the thickening of a glacier and the rise in velocity of ice motion which takes place across this zone in terms of accepted stress–strain rate data for ice. Explanation of the high velocities which occur in the next zone in spite of little change in the available shear stress is more difficult, but the experimental results of Barnes and Tabor on ice close to the melting point appear to offer an explanation of the unusually high rates of flow. In the tensile zone, where velocities slow down, the net lowering of the glacier surface after the surge has passed is explained in terms of the depth of crevassing and easier flow of ice at melting point when under tensile and shear stresses.