Before determining upon the necessary precautions that must be taken, the deposits should be thoroughly examined and tests made of the amount of soluble and insoluble material in a certain mass, as although the earth may have the appearance of being mud, it may contain a preponderating proportion of insoluble particles of sand, and its character may hardly be that of mud or silt. When a marsh rests upon clay it usually requires only reasonable care, but if upon mud or loose sand it may always be in a more or less unstable condition. The chief means of consolidation in the construction of embankments of soft soil upon soft ground may be stated to be as follows:— 1. Draining, which must be deep, and should be effected before the earth embankment is deposited. 2. Weighting the ground outside the slopes with firm material so as to consolidate the soil by intermixture with it. 3. The formation of terraces of earth, and by allowing the ground to subside until movement is arrested, but there may be a porous seam of harder soil, such as gravel, under the soft ground, then care must be taken not to dam back the flow of the underground waters by the subsidence of the added earth or the whole embankment may slide, become softened and finally be swept away; therefore, through drainage must be maintained under an embankment. 4. Making the load as light as possible consistent with due consolidation. 5. Constructing an embankment in equal and regular layers of ashes or other dry, light, stable soil, and by drying the earth to be deposited by separation, wind, sun, or by burning. To prevent slips in such situations, draining, although causing subsidence, should be executed as long as possible before the embankment is tipped, is more effectual and certain in its action than any system of lessening the load or of counterweighting the soil, which may thereby temporarily be brought to rest but afterwards give way; and any damming back of water may cause additional sinking, and change the condition of the earth from a state of dampness to that of saturation. It may be easy to calculate the weight per square foot that will have to be sustained by the ground from an embankment and the heaviest train, but to make the necessary allowance for the effects of vibration is not so readily computed. The ground can be weighted to ascertain the limit of its sustaining power. The test should extend for a period of several weeks to be thoroughly reliable, which time can seldom be afforded upon works; therefore a considerable excess of sustaining power over the load should be allowed.