Abstract Corrosion or erosion of the surfaces of metals under water is in many cases found to be actively promoted by their motion through the water, or by the motion of the water over them. In such cases the relative parts played in the action by corrosion and erosion respectively are not readily distinguished. Even where corrosion products are found the action may still be primarily one of erosion, the latter merely creating the conditions favourable to corrosion. It may, for example, have the effect of removing a surface layer or protective scale. On the other hand, where corrosion products are not found the process is frequently claimed to be a corrosive one, the only function allowed to erosion being that of removing the products of corrosion as fast as they are formed. No doubt in some instances this latter point of view is justifiable, but in many cases the reason for attributing the action to corrosion rather than to erosion is merely that the mechanical nature of the cause is not sufficiently apparent, through a failure to recognise the remarkable ability of water and other nearly incompressible fluids, under certain circumstances, to produce momentary pressures far in excess of the yield-point strength of ordinary materials. The object of this note is to show that in many instances a satisfactory mechanical explanation of such action can be advanced by associating it with the phenomenon known as water-hammer, and, in particular, to show that the same principles afford a satisfactory explanation of the impact pressures produced by drops of water impinging on metallic surfaces.