The reduction in the volume of commercial articles made of clay and placed in kilns varies greatly. With bricks, terra-cotta and pottery it must not, usually, exceed 40 per cent. or the warping and cracking which occur will be so great as to make the articles useless. The fineness of the particles exercises an important influence on the kiln shrinkage of a clay, and the latter is frequently reduced in commercial clayworking by adding burned clay ground to a coarse powder to the plastic clay before it is used. Sand is sometimes added for the same purpose, though its more frequent use is to reduce the shrinkage in drying. Quartz and other forms of free silica expand on heating, so that clays containing them in large quantities shrink very slightly or may even expand. As clays shrink equally in all directions it is usual to state the contraction in linear instead of volume form. Thus instead of stating that a certain clay when moulded into bricks, dried and burned, shrinks 18 per cent. by volume, it is customary to state that it shrinks 3/4 in. per (linear) foot. For many purposes, it is sufficient to regard the linear shrinkage as one-third the volume-shrinkage, but this is not strictly accurate. The _fusibility_ of clays is a characteristic which has been very imperfectly studied. Most clayworkers and investigators employ the term 'fusibility' in a special sense which is apt to be misleading. Owing to the extremely high temperatures to which refractory clays can be heated without even losing their shape, it is almost impossible to fuse them completely. In addition to this, clays are not perfectly homogeneous materials and some of their constituents melt at lower temperatures than others. For this reason a clay may show signs of fusion at 1100° C., but it may be heated for some hours at 1800° C. and yet not be completely melted! Consequently no single 'fusing point' can be stated. In practice, a suggestion made many years ago by Seger (7) is used; the clay to be tested is made into a small tetrahedron (fig. 6), heated slowly until it bends over and the point of the test-piece is almost on a level with the base. The temperature at which this occurs is termed the 'fusing point' though it really only indicates the heat-treatment which is sufficient to soften the material sufficiently to cause it to bend in the manner described. In spite of the apparent crudeness of the test this 'softening point' appears to be fairly constant for most refractory clays.