A detailed study of the system PbO‐Fe2O3, largely by X‐ray diffraction analysis, has been made. The results indicate the existence of three intermediate phases: (1) a phase (beta) extending from PbO·5Fe2O3 to PbO·6Fe2O3, (2) a phase (gamma) extending from PbO ‐ 2Fe2O3 to PbO ‐21/2Fe2O3, and (3) a phase (delta) 2PbO·– Fe2O3. Structures are proposed for the beta and gamma phases that explain their solubility limits, magnetic properties, and very great similarity in structure and in lattice dimensions.

The fracture strength of fused silica rods w a s determined i n vacuum a n d in the saturated vapors of water, ethyl alcohol, n-butyl alcohol, n-propyl alcohol, acetone, ethyl acetate, undried benzene with "less than O . O Z~o water," and dried benzene. The relation between the lowering of the fracture strength of the f u s e d silica caused b y the vapor and the corresponding decrease in surface f r e e energy of quartz w a s found to b e consistent with the Griffith theory of brittle fracture.

No abstract

of that company entered into many constructive discussions and supplied the production data. E. F. Schweitzer made some of the measurements and calculations. The authors are greatly indebted' to E. J. Serfass for carrying out the gas microanalysis.

In a recent paper, Kolthoff (18) made a critical survey of the literature regarding the solubilities of metallic sulfides, and pointed out the surprising disparity to be found in the various values reported. He showed that the determinations for each sulfide fall more or less distinctly into two groups, in one of which, represented principally by the results of Weigel (33) and of Biltz (3), the solubilities are of the order of 10-6 moles per liter, while in the other, represented principally by the results of Bruner and Zawadzki (6), they are very much less. Kolthoff showed that the values in the first group cannot possibly be correct, and suggested that they might really represent the solubilities of oxidation products present at the surfaces of sulfides. This idea is supported by the well-known fact that oxidation products are usually present at the surfaces of sulfide minerals (27), and by the data in table 1, in which the solubilities of some possible oxidation products of a few of the sulfides (4, 7, 8, 19, 28) are compared with the solubilities of the corresponding sulfides as reported by Weigel.Kolthoff gave an improved list of solubilities and solubility products, but made no attempt to correct for activities, which, in many cases, cause appreciable changes in the values. Since accurate solubility and free energy data of the more, important sulfides are required for research work in pyrometallurgy and in flotation, it wras considered advisable to make a careful study of the results of previous investigators, and to recalculate their results with the aid of recent activity data.1In order to determine ion-activity coefficients, it has been assumed that the activity coefficients of potassium and chloride ions are equal in potassium chloride solutions, and that the activity coefficient of any ion depends only on the ionic strength (20). ZINC SULFIDE Maier (21), using entropy data and the value 43,000 calories (2) for the heat of formation, obtained AF2°s --41,600 for zinc sulfide. From this, 1 Unless otherwise specified, free energies, activity data, electrode potentials, and ionization constants are from Lewis and Randall (20).