This thin mixture, however, if left quiet, very quickly resolves itself into a sediment of white fibres with clear water above them. If now the superambient water can be drawn off, and the sediment pressed flat, paper is the result. But it is not so easily done as said. In order to catch the sediment in the most satisfactory way, the milky fluid containing the rag fibres is allowed to flow in a thin stream over a long, shallow trough, which is kept moving onwards and is also so arranged that it has a sideways tremble, backwards and forwards as well. The effect of this is that when the further end of the trough is reached, on its floor there is a thin continuous film of slightly matted fibres, the water from which has flowed away along the sides of the trough. Now another device comes into play; a thin light roller of wire presses lightly on the wet film, and by this pressure the little fibres are pressed upon each other so that they mat, interlace, and cohere together. The paper in this state is of course very delicate, but by reason of a quick drying and carrying off on light rollers it soon acquires the strength necessary to enable it to hold together until it is quite dry. Then it goes through several other stages, the most important of which is pressing. The various processes can obviously be modified easily enough so as to make a thin paper or a thick one. Paper made in this way with a vibratory trough is called “machine made,” and by reason of its fibres laying more or less in a uniform direction the resulting paper is more easily torn in one direction than in the other. If a circular piece be cut out of such pulp and laid on water it will tend to fold up two of its opposite sides. But ancient paper was made in a trough held by hands and given a lateral movement, then pressed and dried in some simple way. By such a procedure the fibres are thoroughly mixed, and do not lie in one direction more than in another, so that if a circular bit of such paper be laid in water it will turn upwards at the edge evenly all round, and look like a little saucer. I have mentioned a light wire roller which presses the wet film more or less into shape just before it leaves the long trough to be dried. As long ago as the thirteenth century in Europe the fact that devices could be impressed upon the undried film by thinning it and making it more transparent where touched appears to have been known, and from about that time onwards “watermarks” have fortunately been applied in the same manner, namely, so as to come in the middle of the first leaf of the pair forming a folio. It is also fortunate that the “chains” or wires forming the rollers have always been laid in the same way; it is certainly the obvious one, but obvious ways are not always adopted. The result is that by observing the direction of the strong chain marks and the fine “laid” marks between them much information concerning the folding of the original sheets can be obtained.