The %C within the top sandy 0.15 m of a sodic Hydrosol under native trees consisted of a constant %C in uncharged organic matter and a %C in negatively charged organic matter decreasing linearly with depth, as did the specific volume of the soil. The kaolinitic clay present was strongly bonded together. In an adjoining canefield cleared 10 years earlier, incorporation of burnt cane residues to 0.35 m had more than doubled the CEC of the soil, but had not generated structural porosity. The clay in the top 0.15 m remained strongly bonded together. The rate of increase in the specific volume of the sandy soil under trees with %C was twice that reported for surface aggregates of a silty soil from rotation plots on a Chromosol, and of sectioned clay cores from a Ferrosol under softwood scrub. The rate of increase in the specific volume of pores [less than or equal to]30 [micro]m diameter with %C was measured by the water retention of aggregates at 10 kPa suction, and was 50% more for the sandy soil than for the silty soil. The difference is ascribed to the dominance of mycorrhizal fungi under trees compared with bacteria under grass. Both agents are presumed to link particles together through acidic polysaccharide gel. Subsequent air-drying then leaves pores stable to wetting and drying. It is suggested that the increase in the plastic limit of silty soils is mainly due to pores stabilised in this way. Pores in decomposing plant residues coated with inorganics could also contribute. Additional keywords: water retention, plastic limit, texture, CEC, fungi, polysaccharides.