Were I a thorough analytical chemist, I should analyze the clay and then mix from this analysis. I have never yet been fooled when I had a complete analysis to work from. Were my original clay a shivering clay I should follow the same plan, simply using a strong burning aluminous clay to mix with it instead of a silicious clay.

Some may wonder whether, in trying to avoid crazing by the method here outlined, there is not danger of producing a body so weak as to be useless as a building material. With proper care I think not. If silica, finely ground or in the form of fine sand, is used, there is great danger of this result, but if a hard burning silicious clay is used there is not Occasionally a clay is met with like the Brandywine Summit clay, previously mentioned, that is aluminous, and yet so refractory as not to burn reasonably hard at a practicable heat. With such a clay one would think that the mixture of a shivering clay would result in a very poor building material, but, if the proper clay for admixture is selected, this is not so. For instance, the clay from Millville, N. J., mixed with Brandywine Summit clay, makes a strong, dense brick, that neither shivers nor crazes. The two clays are as near opposites as we often see. The Brandywine Summit clay contains about 25 per cent alumina, 65 per cent silica, 7 per cent water and only 3 per cent of iron and fluxes. The Millville clay contains alumina, 11 per cent.; silica, 74 per cent; water, 2 per cent, and 13 per cent of iron and fluxes. If we mix these two clays, three parts of the first to two parts of the second, we obtain a mix that will contain 19 2-5 per cent alumina, 68 3-5 per cent silica, 5 per cent water, 7 per cent of iron and fluxes. If 20 per cent of fine ground flint should be added to the Brandywine Summit clay it would stop the crazing, but it would make a brick too soft for use. The chemical composition of such a mixture would be: Alumina, 20.83; silica, 70.83; water, 5.83; iron and fluxes, 2.5. It is rather rare to find a single clay which contains all the properties desired by the glazer, and most of those who undertake the glazing business will have to make up their minds to mix clays, and, in many cases, to transport a certain percentage of their mixtures some distance. I have forgotten to call attention to the peeling in slip and glaze work. The more silicious the body the more difficult to obtain perfect adhesion of slip, so that special attention will have to be paid to this point. Perfect adhesion can be obtained, but not with the ease with which it can be done upon an aluminous clay.

All that I have said is from a single fire standpoint, except in one illustration, where I mentioned a previous biscuit burn. This was done merely to prevent any one from thinking the shivered or broken samples might have broken during the burning instead of afterwards. In the single fire process, body and glaze get exactly the same fire, and when one is under-fired the other is underfired, and when one gets a hard fire, so it is with the other.

The double fire process complicates the crazing problem to a very great degree. The same laws apply, but their application is much less apparent. There is a double set of variations in the burning, producing an endless variety of combinations, of which I might mention the four principal combinations: First, hard burned biscuit or hard burned body, with hard burned glaze; second, soft burned body, with soft burned glaze; third, hard burned body, with soft burned glaze; fourth, soft burned body, with hard burned glaze. The biscuit, or first burning, is at a much higher heat than the gloss, or second burning, except in hard porcelain, which I do not include as a double fire process, for, while it goes through two burnings, the gloss, or final burning, is so much above the biscuit that it may be considered the only real burning the goods receive, and they are really in the same condition as single fired goods. In the regular double fire process, then, we have these four principal conditions, with a fifth which might be mentioned, viz.: medium biscuit and gloss. With a body Just about at the extreme limit for shivering, that is, quite silicious, and at the same time containing considerable felspar and 4 or 5 per cent of lime, I have seen shivered goods, crazed goods and perfect goods, all produced from the same kiln. The heat used in biscuit was such as to vitrify the body in the extremely hot places and the body was quite soft in the extremely easy places; in fact, I have seen plates come from the kiln vitrified on one side and soft on the other, and it was not from bad firing, either, but from the extreme delicacy of the body due to the lime combination. The same kilns and same burner produced enormous quantities of what would be called uniformly burned goods when a less delicate body was used. In this lime body the high heat biscuit was almost certain to craze, whether burned hard or easy in gloss. The medium biscuit would stand if well fired in gloss, and would craze if it got an easy gloss fire. The easy biscuit was indifferent under an easy gloss fire, sometimes standing, sometimes crazing, and sometimes even shivering. The easy biscuit, with hard gloss fire, was sure to shiver. A whole article on the subject of crazing might be written and the facts drawn from this one experience. I cite this extreme case, and claim that in the double fire process this same condition of affairs is always apt to exist to a certain extent. I believe that the tendency to such a condition can best be reduced to a minimum by using a body containing little flux, either spar or lime, and burning the biscuit at an extremely high heat Such a body will harden slowly under the firing and will feel changes in temperature slightly, consequently the variations in the heat of biscuit burning will produce very slight variations in the condition of body. This body should be coated with a glaze that stands considerable excess of heat without blistering, and then the heat in gloss should be carried to such a point as to insure thorough fusion of glaze, even in the easiest part of kiln. Once having obtained such a combination that will not craze, I think the manufacturer could feel relieved of the fear that crazing might suddenly develop. I have carefully examined English goods made by the double fire process that have an excellent reputation for not crazing, and I find invariably that they are a soft but tough body, coated with a completely fused glaze. Such a body could only be made, with safety, upon the lines laid down above.

Analyses of five of the best makes of English enameled brick bodies show an average composition as follows: Alumina, 21 per cent; silica, 74 per cent; iron and fluxes, 5 per cent This would correspond to an analysis in the unburned clay about as follows: Alumina, 19.62 per cent; silica, 69.16 per cent; water, 6.54 per cent; iron and fluxes, 4.67 per cent The extreme variation in the five analyses was not over 2 per cent in silica and not over 1 per cent in alumina. Three of the samples were from England proper, one from Scotland and one from Wales. This shows either that these makers understand the law of chemical composition, or that practical experience has made them unknowingly observe the law.

To-day chemical science does not play a very important part in the business of the clayworker, because the chemical expert and practical clayworker have not been combined in one person, but the day is coming when they will be combined, and data will be obtained, based upon chemistry, that will enable the chemist-clayworker to predict to a nicety from chemical analysis and an examination of the unburned clay the action of such clay under the various heats and conditions to which it may be exposed. When that day comes our business will be a science rather than a matter of guesswork.

Henry R. Griffen.