Journal of The American Ceramic Society-Petersen, Jones, and Allen relaxation develops through the viscous flow of the low-temperature component. However, in the beginning of such a heat-treatment, the strains increase temporarily because of the comparatively rapid shrinkage of the undercooled low-temperature component as it approaches equilibrium a t the treating temperature. This effect merely delays the detection of the relaxation which ultimately allows the high temperature component and the whole mass of the glass to expand accordingly. This expansion together with the shrinkage of the low-temperature component appreciably increases the intermolecular space. These effects together with the changes which heat-treatment causes in the expansivity are responsible for the extraordinary density variations that are peculiar to some Pyrex-brand glasses when they are heat-treated.Previously, an empirical equation was developed to express the density-time relation as the density of ordinary glasses is altered by annealing treatments a t constant temperatures. Because of the rather unusual expansion effects observed in two-component glasses, this equation is obviously inadequate in the case of these glasses. However, it is shown that the equation has a limited applicability even under such circumstances.The equation was designed to take into account the change in viscosity as a glass advances toward equilibrium while at a constant annealing temperature. For this reason, presumably, it generally yields results that are much more satisfactory than those obtained when the logarithmic equation is applied to the same experimental data. The first equation has been found to be superior not only when applied to density data, as in this investigation, but also when applied to the changes that occur in the refractivity when a glass is annealed a t a constant temperature. Moreover, when both equations are modified so that they apply to the relaxation of strains in an annealing glass, the same superiority for the first equation has been found. As the logarithmic equation is derived from the assumption that the viscosity of a glass remains constant as long as the temperature is constant, experience in applying these equations t o various data suggests that the change in viscosity as glass advances toward equilibrium at any annealing temperature is an important factor in determining the rate of that advance and also in determining the rate a t which strains relax. ABSTRACT This method shows great promise in the study of different types of fractures of porcelain enameled specimens. An analysis of the stress conditions in the enamel-iron-enamel system before, during, and after thermal shock is also included.A process is described which locates microscopic cracks in an enamel layer.

A study was made of the spectrophotometric characteristics of a clear and an opaque titaniabearing enamel. Three and six per cent additions of green, blue, brown, and red oxides were made to the base enamels. After firing the enameled samples over a range of time and temperature, spectrcphotometric and Xray data were obtained. The results indicate that the lack of stability of the opaque colored enamels is due to the change from a blue-white to a cream-white color on increased firing treatment. The addition of a cream component tends to shift the hue of the fired enamel toward the red end of the visible spectrum. The green and the blue colors were much less stable than the red. X-ray data show that the total anatase present decreases, which is evidence that a cream-white color developed. X-ray data did not indicate any change in crystal structure of the anatase, rutile, or color oxide. Titania present in the glass of the clear enamel did not cause color instability.

Four antimony cover-coat enamels, four fluoride cover-coat enamels, gnd four groundcoat enamels were formulated to give varying coefficients of expansion. The calculated values were 250, 275, 300, and 325 X 10-l. The observed c d c i t n t 8 of expansion are included in the paper. The enamels were applied to standard 2-quart mdding pans and were tested for thermal shock and impact resistance using the standard methods of the Enameled Utensil Manufacturers' Council.The results show that thermal-shock resistance is controlled chieflp by the coefficient of expansion of the ground-coat ahdey&coat enamels. The best results were obtained by using a cove-coat enamel with a low coefficient of expansion and a grotmdhoat with a high coefficient of expansion. No conclusive evidence was found to indicate any effect of the coefficient of expansion of the ground-and cover-coat enamels on impqt resistance.The solubility resistance and reflectance curves for the enamels are also incladed.

Iiclative t o the control which must be exercised in conducting these tests, it is sufficient to say that this effect depends largely on the degree of accuracy required. Many factors in slip preparation as well as test methods were found to influence the results obtained, but sufficicnt standardization may readily be effected without undue effort or complication to permit results of acceptable accuracy for all practical purposes. For the sake of reliability, duplicate test runs are recommended when there is ally doubt as to the interpretation of a result or where close comparisons are required. In a majority of instances, however, a single test for any one property should suffice.Because ball-clay and kaolin slips contain divergent amounts of water and because different weights of slip were used in conducting the casting tests by the Biichner funnel method, no direct comparisons could be made of the casting time and water retention between a ball clay and a kaolin. Where such comparisons are desired, it may be found expedient to select such quantities of slip that a standard amount of clay is used in the funnel in contrast with the foregoing procedure in which both specific gravity and quantity of slip differ according to the type of clay used. The water content of slip could also be standardized, in which case an amount of water ( I 50'%) as previously used in the ball-clay slips would be satisfactory. Errors, however, E m meled Coo king 1 rte n s ils 11.7 resulting from the settling of aiiy particularly thin slil's thus prepared should be avoided.All tests were conducted on clay slips containiug a substantial proportion of nonplastic constituent. This procedure is recommended not only because of the reduction in testing time involved but because of difficulties inherent in the handling of 100% clay slips and the forming of satisfactory specimens. Whereas 50 clay-50 flint mixtures were adopted as the standard for a majority of the tests herein reported and would be satisfactory for general clay testing, i t is recommended for plant control that a nonplastic constituent of the amount and character employed in a production body of interest be used in combination with'the clay or clays under consideration; for example, feldspar or other nonplastic would partially or totally replace the flint according to the body composition itself. Test results simulative of production expectancy mould thus be obtained.It might be advisable in routine testing to increase the amount of nonplastic beyond the 50-50 ratio and to use a larger Biichner funnel because a reduction of testing time would probably result without too great a sacrifice of reliability. This procedure is recommended only for standardized routine testing purposes.ABSTRACT The various methods for determining the degree of resistance of porcelain enameled cookingware to acid solubility are discrussed. A method that overcomes many of the difficulties encountered in such a test and offers greater accuracy and reproducibility is described in detail.