When pea (Pisum sativum) seeds were wetted, a sharp front separated the wet and dry portions, the seeds swelled, and the water content in the wetted portion continued to increase for a long time. A model was proposed and tested that takes into account these three characteristics and in particular does not postulate a constant diffusivity. The parameters of the model are simply the rate of penetration of the wetting front and a swelling factor.Recently the early water uptake and swelling of phloem has been examined. The uptake was interpreted in terms of both a constant diffusivity (5) and of a diffusivity that varied rapidly with water content (7). Although the variable diffusivity fitted the observations somewhat better, the total water uptake by the entire structure is insensitive to the effect of water content upon diffusivity, and the nature of diffusion in plant tissue can be solved only by observing the water profile in the tissue. The water content and profile vary smoothly when diffusivity is constant and abruptly when diffusivity increases rapidly with moisture. Because phloem or leaves are so thin that the water profile cannot be observed easily in them, we decided to observe water uptake and the profile of wetting in large seeds.As in any other living tissue, water uptake in dry seed is first controlled by permeation and then by growth (1). Since permeation and growth in seed occur on different time scales, they are easily separated. In addition, dead seeds, e.g., split peas (11), or swelling in water too cold for growth (4) can be observed.The equation so far used (8, 10) to model uptake by spherical seed was derived from three assumptions: (a) constant diffusivity, (b) no swelling, (c) constant water content near the surface after the first wetting. These are not borne out by observations; thus, a variable diffusivity was required to fit observations (8, 10). The seeds swelled to about double their size (8, 10, 11), and such swelling requires a different definition of coordinates (9). Finally, the swelling makes more room for water, the flux through the surface scarcely changes for a long time (11), and the water content near the surface increases for a long time (6).Our task is to observe the water profile in seed to show the inconstancy of the diffusivity and then to propose a model that does not require a constant diffusivity but does take into account swelling and the continuing change of water content as the surface expands. THEORY Fundamentally, our theory depends upon two things. First, see Figure 1, a location in the spherical seed is identified by its I This investigation of water movement into seeds is dedicated to Leon Bernstein whose investigations, scholarship, and editing have advanced knowledge of how plants get water and exchange salt.original distance r(o) from the center and by a swelling factor X: X = 8r(t)/0r(o)( 1) where Fr(o) and Mr(t) are the distances between nearby locations before and after wetting for time t. In general, X varies with water content and temperature....