Abstract
Recent microgravity experiments of Frohberg et al. indicate that diffusion coefficients in liquid metals
increase in proportion to the square of absoloute temperature. Such interesting results of the temperature dependence would appear to be in accord with the prediction due to the fluctuation model of diffusion proposed by Swalin about 30 years ago. However, in this theory an inadequate assumption was adopted; in fact, the characteristic time in Einstein's equation used there was given in terms of the absolute reaction rate theory. The time scale estimated in this way is too short to define a diffusion coefficient in view of later studies on the microscopic atomic transport in liquids. Some remarks are given on basic points indispensable for the discussion of the temperature dependence of transport coefficients in liquids.
