So it may be that the natural glazes derive their strength and elasticity from the combination of iron and alumina.

Thorough annealing of the glaze by slow cooling of kiln certainly must produce a tougher and more elastic glaze than where kiln is cooled rapidly. Some glazes require rapid cooling from the time the kiln is finished until the glaze has hardened or set, but I know of no reason to prevent slow cooling from that time until the kiln is emptied. Taking glazed ware from kiln while hot is certainly a very prolific source of crazing.

Some of the coloring chemicals seem to weaken or lessen the elasticity of a glaze in a very marked degree. Among them I will particularly mention copper and uranium. Others seem to strengthen the glaze, as cobalt and chromium. Iron I have previously mentioned as a possible strengthening element in certain combinations. A thick coating of glaze is less elastic than a thin coating. Sometimes we see a glaze crazed where the glaze is thick, and not where it is thin. This is a sign of close, strong body under the glaze. The body possesses sufficient strength to break glaze, and does so in the thick spots, but in the thin spots the elasticity of glaze keeps it from rupture. We sometimes see just the opposite of this; that is, crazing takes place in the thin spots, but not in the thick. This is almost a certain sign that the body under glaze is silicious and unable to break a thick coating of glaze, but perfectly able to break a thin coating; it is also a pretty sure sign that the shivering point is close at hand. I regret that I am unable to give more explicit directions for producing a tough, elastic glaze, but I have already explained the difficulties attending experiments and investigations from which sure deductions regarding elasticity can be made. The work of a glazer is not of such a character as to lead to accurate investigations in this direction. My conclusions are that much more can be done to stop crazing through the body than through the glaze, for all my experience has been that if on a given body a certain good glaze crazes badly, when burned at a certain heat, it is extremely difficult to make a glaze that will not craze on that body, if burned at the same heat, and, if one does succeed, the same course of procedure is not likely to succeed on a different body. Yet we have the natural slip glazes as a proof that combinations are possible that will be successful on the majority of bodies. All that I am able to say, in summing up, is to get plenty of alumina, say 10 to 12 per cent, in the glaze, either by using felspar or clay; also use sufficient paris white or gypsum or plaster of paris to be equivalent in calcium to from 4 to 8 per cent of lime. Make the silica— that is, total silica, derived from the clay, the felspar and the ground silica that may be added - about 40 per cent in low heat glazes, up to as high as 73 per cent in extreme high heat glazes. Combine these three elements with the proper fluxes to effect thorough fusion at the heat used and the result will be a good glaze. If it crazes on the body used, it is easier and more certain to alter the body than to try to make a more elastic glaze.

We now come to the second method for the prevention of crazing in porous bodies, viz.: by making the body so weak that it cannot produce sufficient strain upon the glaze to break it. I have used the words weak and strong throughout this article because I knew of no others to substitute for them, and yet I am afraid they will be misleading without explanation. By a weak body is not meant an open, coarse body. Fine, dense and strong looking bodies often possess this quality, in the sense in which I use it, in a very marked degree, while many open, coarse bodies are very strong. This element of weakness is impossible to determine by appearance, but is soon determined by the action of a glaze upon a body possessing it, or its opposite is determined by the body's action upon a glaze. I do not even believe that what I would call a weak body under the action of a glaze would be weak if tested in a testing machine; judging from appearance, hardness and ring, in many cases, it would not.

In order that my readers and self may have the same understanding of these two words, weak and strong, as used in this article, I will cite a few experiments and their results as illustrations.

If we take national china clay and make it into small tile, one-fourth inch thick, subject them to a potter's ordinary biscuit heat, and glaze them on one side with an ordinary lead glaze, they will nearly all come from the gloss kiln crazed. If we take a clay known in the trade as Key's wad clay, and treat it in exactly the same way, the tile will all come from the kiln in hundreds of pieces. I would call the national clay a strong clay and the Key's clay a weak clay, because the national clay burst or ruptured the glaze and the Key's clay was broken by the same glaze. If we repeat the experiment, simply varying it by glazing the tile all over, the national clay tile will still craze, but the Key's clay tile, if they have no sharp edges, may come from the kiln in perfect condition, and will probably remain so. In this latter case, the glaze being on both sides of the tile, the forces exerted balance each other, and the body is under a compressive strain, instead of a bending or twisting strain. If the tile have sharp edges we will find those edges in the Key's clay tile nipped off, as though it had been done with a pair of pincers. Both of these clays are very plastic, tough clays, the Key's clay being the tougher of the two. They are both fine grained clays and they both burn very hard at the heat mentioned. I think that Key's clay would be pronounced the stronger by almost anybody, if tested in the usual manner - by ring, hardness, etc.

The only difference in appearance between the two sets of tile, with rounded edges and glazed all over, will be in color and in the fact that one crazes and the other does not. This is not chance, but is due to some inherent difference in the clays. I have already shown the difference between the two when glazed on one side only, and, from that difference, claim the Key's clay as being the weaker clay, and I claim that it is this same peculiar weak element that makes the Key's clay a non-crazing clay when glazed all over. The clay has not the particular kind of strength necessary to break its coating.