Combined mechanisms¹（组合机构） of transport have important applications—transport of nutrients² across cell membranes³（细胞膜） in plants and animals, the aeration⁴（通风，充气） of agricultural soils, performance of chemical reactors⁵, the design of membranes for desalting（淡化） brackish⁶ water（半咸水，淡盐水） , and the design of clay（泥土，粘土） membranes for retaining dangerous chemicals. Because mass transport of fluid constituents⁸ has important roles in biology, physics, and chemistry, one would assume that such transport would be well understood by the scientific community. However, transport of fluid constituents（成分） continues to be a source of confusion, particularly regarding models for combining transport by molecular⁹ diffusion¹⁰（分子扩散） and advection（平流，对流） . In a recent article in Vadose Zone Journal, A.T. Corey, W.D. Kemper (both of Colorado State Univ., Fort Collins), and J.H. Dane (Auburn Univ., Auburn, AL) show that the developers of popular models of diffusion have made invalid¹¹（无效的，有病的） assumptions. Currently popular models define diffusion of a particular constituent⁷ as a flux¹² relative to mass average flux（流量） so that diffusive¹³ flux of all constituents in a fluid mixture must sum to zero, and self-diffusion of a single-specie fluid cannot exist, contradicting experimental evidence previously¹⁴ reported in the literature. Research conducted and referenced by the authors shows that these assumptions and their models do not provide a satisfactory description of the flux taking place in media with small pores（气孔，毛穴） . The authors provide an improved analysis, based on the principle that driving forces (for both advection and diffusion) are each equal in magnitude (and opposite in direction) to the associated rate of change of momentum¹⁵. Mass average flux resulting from combined advection and diffusion is shown to be evaluated as the vector sum of advective and diffusive fluxes¹⁶, rather than diffusive flux being evaluated as a flux relative to a mass average flux. This procedure is necessary because "mean flux" cannot be determined¹⁷ independent of an evaluation¹⁸ of diffusive flux.

Corey et al. describe two experiments with transport of gas constituents through porous¹⁹ media（多孔介质，漏失通道） (providing data consistent with their revised model) that contradict widely accepted models. One of the experiments presents previously published data and the other describes new data obtained by the authors. Three additional experiments are presented (one representing new data) showing that self-diffusion of pure liquid water occurs in response to a temperature gradient（温度梯度） , contradicting theory that diffusion of a single-specie fluid cannot occur, and showing that the sum of diffusion fluxes do not sum to zero in the general case. The measured diffusion of water was proportional to the gradient of the vapor²⁰ pressure（蒸汽压） , which is a well-documented measure of the kinetic²¹ energy（动能） of water.