_Mud_ is always objectionable. It frequently contains organic slime and organisms which encourage the putrefaction of hides put in it to wash or soften. It also almost invariably contains iron as one of its constituents, and hence stains leather, and gives bad coloured liquors. It is not easily got rid of by filtration, as large filter-beds are expensive and difficult to keep in order, and much space is required to clear water by subsidence. Some filter easily cleaned offers the best chance of success. The Pulsometer Company supply such a filter, consisting of sponge tightly packed below a perforated piston. To cleanse the filter a stream of water is passed the reverse way, and the piston raised, and worked up and down, either by hand or power, so as to loosen and knead the sponge. The Atkins "water scrubber," in which sand may be used as a filtering medium, seems also well adapted for the purpose. If lime be precipitated by Clark's, or other process, it usually carries down the mud with it. Rain water and the water of streams in mountain districts of hard igneous rock are generally nearly free from mineral constituents. This is the case with the Glasgow water from Loch Katrine, and the Thirlmere water which is to supply Manchester. Such water, if cold enough, and free from mud and organic impurity, is the best for almost every purpose in tanning. Most river water, however, and all spring water, is contaminated with mineral matter which it has dissolved out of the soil and rocks through which it has flowed. The principal of these mineral constituents are lime and magnesia. These occur both as sulphates and chlorides, and as hydric carbonates, or "bicarbonates." The sulphates and chlorides constitute "permanent" hardness, while that due to bicarbonates is called "temporary," from the fact that on boiling, half the carbonic acid is driven off, and the lime or magnesia is deposited as an insoluble neutral carbonate, thus softening the water. Any water which can be softened in this way by boiling may also be softened by the addition of a suitable quantity of lime, thus:-- Calcic hydric carbonate. Lime. Chalk. Water. (CO_{3})_{2}CaH_{2} + Ca.(OH)_{2} = 2CaCO_{3} + 2OH_{2} This is Clark's process, and the chalk may either be separated by subsidence, which quickly takes place, or by a special filter (Porter-Clark). Thus the Bristol water, which from determinations by Mr. W. N. Evans, contains considerable temporary hardness and but little of permanent, may be almost completely softened by Clark's method. (For method of determining hardness and quantity of lime required, see p. 97).