Rotation of unstimulated maize (Zea mays L.) seedlings on a horizontal clinostat is accompanied by a strong bending response of the coleoptiles towards the caryopsis, yielding curvatures exceding 100 degrees. The corresponding azimuthal distribution shows two peaks, each of which is displayed by 30 degrees from the symmetry axis connecting the shortest coleoptile and caryopsis cross sections. It is argued that this spatial pattern is not the result of two independent bending preferences, but caused by a one-peaked distribution encountering an obstacle in its central part and thus being split into the two subpeaks. The existence of one preferential direction justifies considering this response to be a nastic movement. Its time course consists of an early negative phase (coleoptiles bend away from the caryopsis) followed 2 h later by a long-lasting positive bending towards the caryopsis. In light-interaction experiments, fluence-response curves for different angles between blue light and the direction of the nastic response were measured. These experiments indicate that blue light interacts with the nastic response at two levels: (i) phototonic inhibition, and (ii) addition of nastic and phototropic curvatures. It is concluded that phototropic and phototonic transduction bifurcate before the formation of phototropic transverse polarity. The additivity of nastic and phototropic responses was followed at the population level. At the level of the individual seedling, one observes, in the case of phototropic induction opposing nastic movement, three distinct responses: either strong phototropism, or nastic bending, or an "avoidance" response which involves strong curvature perpendicular to the stimulation plane. With time the nastic bending becomes increasingly stable against opposing phototropic stimulation. This can be seen from a growing proportion of seedlings exhibiting nastic bending when light is applied at variable intervals after the onset of clinostat rotation. At the transition from instability to stability, this type of experiment produces a high percentage of seedlings displaying the "avoidance" response. However, no cancelling resulting in zero curvature can be observed. It is concluded that the endogenous polarity underlying the nastic response is different in its very nature from the blue-light-elicited stable transverse polarity described earlier (Nick and Schafer 1988b).