In the event of there being only one “boil,” but that of important extent, perhaps the best way to proceed is to place clay upon the surface for a few feet around it, the weight of which should not be less than the normal pressure of the earth removed by excavation, _vide_ Chapter V.: and by the insertion of a cylinder, with a properly designed bottom, to prevent the issue of soil to a level some feet below the lowest intended excavation upon the site, or to the depth at which the greatest flow of water is obtained; care being taken that no run of sand or soil is allowed and that the water is not charged with earth, either by a thick gravel layer until the sand ceases to pass through it, or by other means, remembering that the motive power of running sand is that of the pressure of water. The flow should be gently led away, if necessary, to a sump, and discharged by pumping, thus draining the works. The sump should always be lined, not only to prevent any flow upon or into the surface causing saturation, but so as to control and regulate the water some feet below the lowest level of the foundations, the great point being to draw downwards the water from the earth and then to carry it away. An iron or wooden cylinder sump is to be preferred to one constructed of sheet piles or timber-lined, as the soil may be so loose that any shaking or vibration should not be allowed, and a close joint is not easy to attain in timber pile work. It is well to deposit some clay puddle for a little distance round the sump after sinking it in order to weight the soil and steady the cylinder. When a discharging outlet has been made it should be maintained in perfect condition, or a slip or subsidence may occur from leakage or diversion of the flow. It may also happen that no discharge is necessary, for if the “boil” is only, as it were, of local extent and caused by a weak vein in otherwise stable soil, the water may rise to the head level and there remain; the equilibrium being restored which had been destroyed by reason of the upper strata of an impervious character being excavated. The removal of the normal pressure upon the loose soil to some extent, liberates the water, and the head ceasing to be balanced by it, the earth cannot do otherwise than allow water to upraise it. A calculation of the weight of the earth removed may give an indication within which will be the probable head level of the supply, and usually it is much below a head of water that would equal the normal pressure of the soil. Supposing the weight of the earth to be twice that of water, when 10 feet of soil is excavated before a “boil” began, its removal would approximately be equivalent to a 20 feet head of water.