The study investigates turbulent mixing layers between two streams of different gases, focusing on the effects of density differences and large-scale structures. Experiments were conducted using a novel apparatus to study the mixing layer between nitrogen and helium. Spark shadow pictures and high-speed movies revealed that the mixing layer is dominated by large coherent structures, which convect at nearly constant speed and increase in size and spacing through amalgamation. Density fluctuations suggest that turbulent mixing and entrainment occur on the scale of these large structures. The results indicate that large changes in density ratio across the mixing layer have a relatively small effect on the spreading angle, suggesting that compressibility effects, not density effects, are responsible for the observed behavior in supersonic flows. The paper discusses the plane turbulent mixing layer, its characteristics, and the experimental setup used to study it. The apparatus was designed to achieve high density ratios and Reynolds numbers, using different gases to create density differences. The experiments revealed that the mixing layer spreads linearly, with the spreading rate depending on the velocity and density ratios. The results show that the spreading rate is not significantly affected by the density ratio, indicating that the effects of density are not as significant as previously thought. The study also compares the results of low-speed mixing layers with those of supersonic mixing layers. It is found that the observed differences may be related to the forms of the Reynolds equations for incompressible and supersonic flow. The pressure-velocity correlations are shown to account for the observed differences between shear layers in supersonic and incompressible flow. The results suggest that the spreading rate of the density and velocity profiles in incompressible flow can be described by an eddy-viscosity model only if the Schmidt number is much less than 1. The study concludes that the effects of density on turbulent mixing are not as significant as previously thought, and that compressibility effects are more important in supersonic flows.The study investigates turbulent mixing layers between two streams of different gases, focusing on the effects of density differences and large-scale structures. Experiments were conducted using a novel apparatus to study the mixing layer between nitrogen and helium. Spark shadow pictures and high-speed movies revealed that the mixing layer is dominated by large coherent structures, which convect at nearly constant speed and increase in size and spacing through amalgamation. Density fluctuations suggest that turbulent mixing and entrainment occur on the scale of these large structures. The results indicate that large changes in density ratio across the mixing layer have a relatively small effect on the spreading angle, suggesting that compressibility effects, not density effects, are responsible for the observed behavior in supersonic flows. The paper discusses the plane turbulent mixing layer, its characteristics, and the experimental setup used to study it. The apparatus was designed to achieve high density ratios and Reynolds numbers, using different gases to create density differences. The experiments revealed that the mixing layer spreads linearly, with the spreading rate depending on the velocity and density ratios. The results show that the spreading rate is not significantly affected by the density ratio, indicating that the effects of density are not as significant as previously thought. The study also compares the results of low-speed mixing layers with those of supersonic mixing layers. It is found that the observed differences may be related to the forms of the Reynolds equations for incompressible and supersonic flow. The pressure-velocity correlations are shown to account for the observed differences between shear layers in supersonic and incompressible flow. The results suggest that the spreading rate of the density and velocity profiles in incompressible flow can be described by an eddy-viscosity model only if the Schmidt number is much less than 1. The study concludes that the effects of density on turbulent mixing are not as significant as previously thought, and that compressibility effects are more important in supersonic flows.