Experiments are described in which the ‘once-used’ soils left over from previous culture experiments were employed as aggregates in the growth of vegetables. It was demonstrated that sand with full nutrients was superior to ‘once-used’ soil with water only, but that ‘once-used’ soil supplied with the full quota of soluble nitrogen was superior to the sand with full nutrients. The root of the cabbage in sand was an exception, and a possible explanation was that sand as aggregate favoured true root formation, while soil favoured the formation of tops. The ‘once-used’ soil in these experiments deteriorated in physical propertes owing to the conditions of the experiment, which included the use of nitrogen as sodium nitrate, possibly owing to the tendency to the formation of sodium clay; it is argued that a judicious choice of fertilizers and/or the use of flocculating electrolytes should overcome such tendencies.