LONDON Institute of Metals (Annual General Meeting), March 7. G. A. HANKXNTS and C. W. ALDOUS: Minimum dimensions of test samples for Brinell and diamond pyramid hardness tests. The metals investigated include copper, brass, aluminium and steel. A width of test-specimen of 4 £ times the diameter of the impression is satisfactory for accurate Brinell tests. For Brinell tests, the limiting value of the ratio of thickness of test sample to depth of impression for accurate results appears to be a characteristic of the test material; a value of the ratio of 6 is required for mild steel, about 15 for copper and more than 20 for spring steej. For diamond pyramid hardness tests a limiting value of the ratio of test-sample thickness to impression diagonal of 1J gives results which are practically independent of test-sample thickness except with soft copper and soft brass. I. G. SLATER: Note on the influence of gases in an 8 per cent copper-aluminium alloy on normal and inverse segregation. In a sand-cast ingot, 3 in. in diameter by 3 in., segregation is inverse with very gassy melts but normal with degassed melts. GILBERT RIGG: The diffusion of zinc and iron at temperatures below the melting point of zinc. When clean rolled zinc sheet is heated in close contact with clean iron, diffusion commences at below 300 ° C. and is fairly rapid at above 380 ° C.; it proceeds by the formation of cones of diffusion products, which spread out from isolated points where the contact between the metals is most perfect, and gradually penetrate into the zinc and across its surface. Two well-defined layers of diffusion products are formed, a thin layer of constant thickness (about 0-08 mm.) containing about 17 per cent iron being next to the iron, and a thicker layer containing 0-11 per cent iron outside this. On continued heating, the thin layer moves towards the zinc, being continuously converted into the zinc-rich layer; this would seem to indicate that the principal diffusion constituent is the iron. H. G. GOUGH, H. L. Cox and D. G. SOPWITH: A study of the influence of the intercrystalline boundary on fatigue characteristics. With the object of studying the process of fatigue in relation to crystalline boundaries, tests under alternating torsional stresses have been made on three specimens of aluminium each consisting of two crystals. The distribution of slip bands showed that the effect of the boundaries on the distribution of stress was extremely slight, each crystal of each specimen behaving as if it alone composed the whole specimen. It appears that the presence of intercrystalline boundaries may considerably strengthen the constituent crystals against fatigue; but that the effect of the boundaries on the distribution or even on the amount of slip is very small. It is probable that the major effect of the boundary may lie in some restriction of strain that it imposes. C. E. PEARSON: The viscous properties of extruded eutectic alloys of lead-tin and bismuth-tin. Elongations up to 2,000 per cent have been obtained in tensile tests employing prolonged loading. An apparatus designed to maintain a constant stress on the test-piece during extension shows that deformation takes place at a uniform rate which is greatest in freshly extruded rods and decreases with age pr on annealing. The viscosity is not that of simple liquids, but resembles that shown by some disperse systems in which the viscosity coefficient is a function of the stress causing flow. The locus of viscous flow is found to be at the inter-crystalline boundaries. E. W. FELL: A note on some formulae concerning viscous and plastic flow in soft metals. In particular, the flow of the metal in a prolonged ball-hardness test is compared with the flow in tensile test-pieces under a constant stress per unit area of cross-section. A. PORTE VIN and P. BASTIEN: Castability of ternary alloys. The ability of a molten metal or alloy to fill a mould completely is termed ‘castability’; it can be determined by ascertaining the length of a spiral cast-iron mould filled by the metal under predetermined casting conditions. The castability of a pure metal is a linear function of the difference between the pouring temperature 6 and the melting point F; the slopes of the castability (Q?F) curves vary with the viscosity of the metal. The castability of binary alloys varies with the solidification range and with the mode of crystallisation, being greater when polyhedral crystals separate than when the primary crystals are dendritic. Maximum castability occurs with the eutectic composition and minimum at the limit of solid solubility. The castability of ternary alloys generally varies inversely with the primary solidification range.