Phototropic stimulation induces a spatial memory. This was inferred from experiments with maize (Zea mays L.) coleoptiles involving opposing blue-light pulses, separated by variable time intervals, and rotation on a horizontal clinostat (Nick and Schafer, 1988b, Planta 175, 380-388). In those experiments, individual seedlings either curved towards the first or towards the second pulse, or they remained straight. Bending, if it occurred, seemed to be an all-or-none response. Intermediates, i.e. plants, bending only weakly, were not observed. In the first part of the present study it was attempted to create such intermediates. For this purpose the strength of the first, inducing, and the second, opposing, pulse was varied. The result was complex: (i) Individual seedlings maintained the all-or-none expression of spatial memory. (ii) However, on the level of the whole population, the time intervals at which a given response type dominated depended on the fluence ratio. (iii) Furthermore, the final curvature was determined by the fluence ratio. These results are discussed in terms of a blue-light-induced transverse polarity. This polarity initiates from a labile precursor, which can be reoriented by an opposing stimulation (indicated by the strong bending towards the second pulse). The strong curvatures towards the first pulse over long time intervals reveal that, eventually, the blue-light-induced transverse polarity becomes stabilised and thus immune to the counterpulse. In the second part of the study, the relation between phototropic transduction and transverse polarity was characterised by a phenomenological approach involving the following points: (i) Sensory adaptation for induction of transverse polarity disappears with a time course similar to that for phototropic sensory adaptatation. (ii) The fluence response for induction of transverse polarity is a saturation curve and not bell-shaped like the curve for phototropism (iii) For strong counterpulses and long time intervals the clinostat-elicited nastic response (Nick and Schafer 1989, Planta 179, 123-131) becomes manifest and causes an "aiming error" towards the caryopsis. (iv) Temperature-sensitivity of polarity induction was high in the first 20 min after induction, then dropped sharply and rose again with the approach of polarity fixation. (v) Stimulus-summation experiments indicated that, for different inducing fluences, the actual fixation of polarity happened at about 2 h after induction. These experiments point towards an early separation of the transduction chains mediating phototropism and transverse polarity, possibly before phototrophic asymmetry is formed.