When the blood is withdrawn from the body, it sets, coagulates, or clots, becoming converted from a fluid into a jelly. This process occasionally occurs in disease, whilst the blood is still contained within the vessels. It takes place in different animals with various degrees of rapidity; thus in the blood of birds less than a minute suffices to complete the change, while a quarter of an hour or twenty minutes may elapse before the blood of the horse becomes a solid mass. Violent muscular efforts made by the animal before the blood is drawn, or the rapid cooling of the blood, effected by surrounding the vessel into which it is drawn with ice, retard coagulation almost indefinitely. If the process of coagulation be carefully watched, it will be seen that on account of the corpuscles being heavier than the plasma, they sink through that fluid, their descent being aided by their disposition to adhere together by their broad surfaces, forming rouleaux (see fig. 181), and so presenting their edges to the fluid. The white corpuscles, though heavier than the plasma, are lighter than the red, and hence form a thin layer on the surface of the latter, the whole being surmounted by a moderately thick layer of plasma. Whether this separation of blood into layers have taken place or no, the whole mass becomes first viscous and then solid, the difference resembling that seen in the white of egg before and after boiling, or in gelatine before and after setting. The consistence of the clot is about equal to that of red-currant jelly. After the lapse of a few minutes a further change may be seen. The clot contracts, and minute drops of a clear fluid begin to exude from the surface. This is the separation of the serum from the clot; and, as the contraction continues for many hours, the clot is ultimately covered and surrounded by a layer of clear fluid of considerable depth. When the separation of the freshly-drawn blood into three layers has taken place, the contraction of the upper layer of plasma, being unhindered by the presence of corpuscles, causes the upper surface to be much depressed in the centre, and its colour being yellowish, such clot is spoken of as being " buffed and cupped ". It is, of course, not observed when clotting has taken place too quickly for the corpuscles to sink through the plasma. Coagulation of the blood is believed to result from the breaking up into two parts of a proteid substance named fibrinogen, naturally existing in solution in the plasma. One of these parts is a globulin which, under the influence of a ferment existing in the white corpuscles, remains in solution; the other is fibrin, which immediately solidifies, forming a delicate net-work in the meshes of which the corpuscles are entangled, and which, subsequently contracting, squeezes out the serum. The calcium salts also play an important part in the process.