Near the entrances to tunnels slips appear to be most frequent in cuttings in chalk. This would seem to lead to the belief that in places where it is known the chalk soil is likely to be troublesome from land-water and springs, it would be advantageous to prolong tunnels beyond the economic depth of a cutting, and even to continue them until such a depth as 40 feet is reached, to so arrange the gradients that they drain the interior, and to provide a complete system of pipes and drains, even if side galleries have to be driven to tap the water, before it reaches the lining, to relieve the sides, crown, and invert of a tunnel so that no water can pour down the roof or walls unless under control. In a tunnel so situated and liable to water-pressure, the thrust of the soil will be very variable, and cannot be foreseen. At one place during construction, the walls and lining may be finished without movement of the earth, at another, the pressure maybe great and act unequally, either upon the side walls or the arch. As a rule, at the entrances to tunnels the pressure is greater upon the arch than the sides, for then the whole of the wedge-shaped mass within the boundary of the angle of repose of the soil is disturbed, and its cohesion impaired or destroyed, and therefore it presses upon the arch, this pressure tending to counteract the lateral pressure; but as the depth increases, the load from this wedged-shaped mass becomes less upon the arch, although generally greater than any lateral pressure, because the earth above is not disturbed or impaired, consequent upon the depth of the hill being greater and the cohesion and side-pressure of the earth tending to support it, but the lateral pressure is increased because the normal pressure of the soil due to the depth is augmented. It is this disturbance of such friable soil as chalk at tunnel entrances, causing the particles to be loose and separated, and in a state especially disposed to percolation of water, that probably causes the earth to be in a condition favourable to slips, and for them particularly to occur at or about the entrances to tunnels. When, therefore, the depth of open cutting at the entrances is reduced, any slip cannot be of the same magnitude as it would be if it happened at a greater depth. The circular or one closely approaching it would appear to be the best form for the lining, where variable or great pressure, vertical or lateral, is to be expected; for the pressure in a tunnel will always be unequal, and the surface of the earth must be supported. In some experiments to join substances by pressure it was found that though great pressure forms chalk into hard blocks, the particles are not firmly united, and that they separate along the surfaces of contact of the original particles and not through them; these tests tend to show that masses of chalk are usually in a state not indisposed to separation. The same result occurred in similarly testing pulverised sandstone.