Crude Brandywine Summit clay will act exactly as national clay does in regard to crazing, and English china clay No. 7 will act exactly as the Key's clay does in regard to shivering. The Brandywine Summit clay is a coarse, open clay, that does not burn at all hard at potter's biscuit heat; in fact, can almost be cut with a knife, and a piece from one-fourth to one-half inch thick can easily be broken with the hands. English china No. 7 is a short, very fine grained clay, that burns soft. Near Millville, N. J., is found a clay that is very peculiar. This clay becomes semi-vitreous at the heat mentioned, and is an extremely heavy, close, hard and strong appearing clay, when burned, yet it shivers. I could cite a number of instances of crazing in hard burning clays and crazing in soft burning clays, of shivering in hard burning clays and shivering in soft burning clays, crazing and shivering in both plastic and non-plastic clays, in coarse grained and fine grained clays.

We see, then, that both crazing and shivering are independent of plasticity, grain and hardness to which clay burns; in fact, they are both independent of physical characteristics, not only of the original physical characteristics, but even of those characteristics left after burning, except this peculiar weakness or strength, which is not apparent to any tests but those of the glaze itself.

Can we find a point in common between the crazing clays? and can we find an opposite point in common between the shivering clays? If we can, I think it will be perfectly safe to assume we have found what leads to crazing and what leads to shivering. This common point will be found in the ratio of silica to alumina. All the crazing clays will be found to be high in alumina and low in silica, and all the shivering clays high in silica and low in alumina, and there is a comparatively small margin of safety. I do not mean to claim that there is a sharp line, which, if overstepped in either direction, will inevitably lead to either crazing or shivering. The amount of flux in the clay and the heat at which the clay is burned have some effect upon this point. For high heat clays, with small percentage of flux, a ratio of three and one-half of silica to one of alumina, or 70 per cent silica, 20 per cent alumina, is about the proper proportions, and a clay running as high as 74 per cent silica is almost certain to shiver, and one running as high as 25 per cent alumina is almost certain to craze.

Finely ground flint added to a crazing clay in sufficient quantity will, I think, invariably produce a shivering mixture. A crazing clay and shivering clay can be mixed together in such proportions as to either craze, shiver, or stand perfectly. In adding ground flint we produce often a mixture that is weak to ordinary tests, while a strong burning silicious clay added does not produce such a mixture. Therefore, I have always found it better to work with mixtures of silicious and aluminous clays rather than with an aluminous clay and silica.

So far the law of expansion and contraction has not entered into my discussion on the second method of avoiding crazing in porous bodies, but we must not lose sight of the law. If my experiments made with lime bodies are to be relied upon, lime bodies will be porous bodies at low heats, and yet will possess a comparatively low melting point. Such bodies will, then, possess two good qualities: From their porousness and weakness they will not have the strength or capacity for producing much strain upon the glaze, and from their low melting point will not have a very great tendency to produce strain. Slight capacity and slight tendency together ought to make a non-crazing body. This is confirmed by practice. Clays containing much lime are less liable to craze than clays that do not contain lime. The lime must be in a finely divided state and uniformly distributed, or it will lead to other difficulties as obnoxious as crazing. The direct addition of lime in the form of well-ground paris white will serve the purpose, but would be too expensive except where the cost of body is a slight percentage of total cost or where profits are great.

A short description of my methods in handling a new clay will probably do more to illustrate the application of the principles I have endeavored to set forth than anything else I might write. After having determined the adaptability of the clay to the purpose in hand by examining and testing its plasticity, shrinkage, liability to check in drying, texture, and, if a material requiring repressing, the action of the clay in the repress should be thoroughly looked into - after, as I have said, testing all these points, the next point I ascertain is the heat at which the clay burns. I probably would put the clay in a regular kiln, burning to whatever heat I happened to be using, and from its appearance determine whether the clay was apt to be a porous clay or to be one easily vitrified. I would then make some special tests on this point. If the clay should be one that is easily vitrified I would treat it as a vitrified body under the rules I have laid down for such bodies; that is, use as high a heat glaze as I could and burn until the glaze was thoroughly fused and body thoroughly vitrified. Ninety-nine times out of the hundred this would lead to perfect goods so far as crazing is concerned. If the body should be a porous body, I would glaze it with a good glaze that works well at the heat I was using, and, after getting a number of trial pieces from kiln, would subject them to severe tests of alternate heating and cooling. If crazing resulted, I should add silica sand, or fine silica, or, preferably, a silicious clay that burns hard. I should try several different mixtures, probably adding sufficient silica or silicious clay in my extreme mixture to produce shivering. I should then test these trial pieces as before. Wo will say I made mixtures containing 4 per cent, 8 per cent, 12 per cent, 16 per cent, 20 per cent, and I found that the 4 per cent mixture crazed and the 20 per cent mixture shivered, the 8 per cent., 12 per cent and 16 per cent mixtures standing well, I should use the middle, or 12 per cent, mixture; but, were I adding silica, it is more likely that both the 4 per cent and 8 per cent mixtures would craze and the 16 per cent and 20 per cent mixtures shiver, leaving me only the 12 per cent mixture good. I should then repeat my trials with a 9 per cent, 10 per cent, 11 per cent, 12 per cent, 13 per cent, 14 per cent and 15 per cent mixture, and from the result of these tests determine where my safest mixture lay.