Slow crack growth in glass, and the corresponding mechanical fatigue phenomenon, in the presence of water or water vapour are explained in terms of a stress corrosion reaction between water and glass accelerated by a tensile stress. The specific reaction between glass and water, however, is not clear. From the analogy between the structural changes induced by thermal and mechanical treatments and the corresponding property changes, the glass dissolution and hydrolysis reactions of silica glass are considered unlikely to be accelerated by tensile stress. On the other hand, water diffusion into a silica glass, or hydration, is accelerated by tensile stress. The relationship between the water diffusion rate and the slow crack growth rate is discussed.
