Investigations to determine the causes of a type of crazing which develops on certain glazed ceramic wares, some time after firing, revealed the fact that in many cases such crazing is caused by an expansion of the body. This is probably due to the combination of water with the body, while the glaze remains practically constant in size. A body may be tested for this type of crazing by placing the glazed specimens in an autoclave and subjecting them to a steam pressure of 150 to 175 pounds per square inch for one hour. Bodies showing the least tendency to develop this type of crazing are those which have a low porosity, are not very soluble in sulphuric acid, and have a low ignition loss above 110°C.

The effects of the composition of some ceramic bodies on their moisture expansions were investigated. With the exception of mixtures containing blast-furnace slag or magnesite, most bodies with high absorptions developed considerable expansion. Bodies with low absorptions composed of pure clay and feldspar developed practically no moisture expansion while those composed of clay, feldspar, and more than 20% silica developed considerable moisture expansion. By adding magnesite, the resistance of a ceramic body to moisture expansion was improved. Additions of whiting, iron oxide, and titanium oxide were not as beneficial as magnesite. It was found that crazing caused by moisture expansion of bodies can usually be prevented by proper manufacturing control.

Introduction.One of the sources of loss in the manufacture of clay products, especially of the larger kind, chemical stoneware, glass pots, etc., is due to the lack of strength of the clays in the dry state and at the burning temperatures. Suficient strength to resist the shocks and strains of handling in the green state is necessary. A t certain periods in the burning the ware loses strength and is liable to warp, crack, or break, causing large kiln losses. The natural remedy would be to improve the strength of the clay body by artificial means or to introduce stronger clays.The effect of the addition of organic compounds tc clays has been studied by Acheson,2 Minton3, Seger and Cran~er,~ and K~e r n e r ,~ and it has been shown that the strength of the clays can be increased by the addition of gallo-tannic acid, catechu, straw emulsion, dextrine, or starch.The effect of the addition of alkalies and acids on the volume shrinkages of clays in the plastic state has been reported by Bleininger and Fulton,6 and Kerr and Fulton,' but our information as to the effect of these electrolytes on the strength of clays is meagre. Bleiningers has reported that cast bodies containing alkalies may be stronger than bodies not containing alkalies. The writer9 has noted that the rate of disintegration in water is much slower for the deflocculated than for the flocculated clays.

The characteristics of glazes are too frequently regarded as the only important factor governing the resistance of glazed ware to thermal shock crazing. I n this investigation it was found that thermal shock crazing was caused, in some cases, by craze-like cracks which developed in the bodies. The importance of selecting suitable bodies, as a means of controlling crazing, became evident a t once when it was found that bodies with a high resistance to thermal shock were not so liable t o craze when glazed and subjected to sudden temperature changes as those with a low resistance to thermal shock. I n general, glazed ware on which the glazes are under fairly high compression are more resistant to thermal shock crazing than those on which the glaze is under either low compression or tension. Glazed ware which is susceptible to thermal shock crazing may be resistant to moisture crazing and vice versa.

Several new methods of studying the surface reactions of slags, glass, and Portland cement clinker with refractories were compared. The penetration of the slag into refractories and the slump of slag spheres disclosed considerable information regarding surface reactions which other tests failed to show. For the more refractory slags, such as Portland cement clinker, acid open-hearth, copper reverberatory, and hasic openhearth slags, these reactions were governed largely by the relative acidity or basicity of the slags and the refractories, Le., basic slags showed the greatest reaction with acid refractories and vice versa. For the more fusible materials, such as glass, lead blastfurnace slag, and coal ash, the relative acidity or basicity of the slag and refractory is less important. With these slags, the low porosity of the refractory, which prevents the penetration of the very fluid slags, probably is more important.