Since the Annual Meeting of 1950, measurements of the heat conductivity of three mixtures of nitrogen-carbon dioxide have been continued to higher temperatures and the earlier reported measurements repeated. It is now established for the N2–CO2 mixtures at low pressure that the sign of the deviation of the observed heat conductivities, relative to the conductivities computed on the basis of the linear composition proportionality, changes, passing from negative at room temperature to zero at about 150 C, and becoming increasingly positive at higher temperatures. This behavior is unexpected and appears not to have been hitherto observed. The Enskog theory of the conductivity of mixtures does not predict the observed trend nor does the empirical formula of Lindsay and Bromley. The data have been formulated at each of the three constant concentrations of constituents using the same type of equation that has proved useful for the pure components. A new installation for low-temperature measurements has been completed and tested in the course of calibrating the copper-constantan thermocouples to be used for measuring the temperature gradient and absolute temperature from 30 to −180 C of the fluids whose heat-conductive properties are of interest. The N2–CO2 mixtures will be measured along with a series of pure gases and liquids. The measurement of the pressure effect for the heat conductivities of the N2–CO2 mixtures has been attempted but without satisfactory results above 50 C. The difficulty consists of failure of the fluid in the cell to arrive at a steady state above a comparatively low pressure. The cause of the phenomenon is believed, as expressed in last year’s report, to be directly related to the known tendency of the constituents of a mixture to separate in a temperature gradient.