Another possible explanation to Demonstration 2 is that the edge displacement is simply due to the appearance of motion signals presented within the far half-pattern and their directions determine the direction of the edge displacement. However, this Demonstration ruled out this possibility by introducing a strong ‘speed imbalance’ into the expansion pattern, the speed of motion signals within the far half-pattern of the expansion was tripled, i.e. a ‘speed imbalance (faster far half-pattern of expansion)’ condition. If the role of motion signals within the far half-pattern is exactly as suggested above, we will expect to observe an edge displacement with the same direction but a larger magnitude as that in Demonstration 2. The reason is that we amplified the effect of the ‘special role’ of local motion signals within the far half-pattern of the expanding pattern by tripling the speed. However, here most observers perceived an edge displacement whose direction is the same as that in Demonstration 1 instead of Demonstration 2, i.e. the lower edge (defined by expanding translations) seems more close to the central fixation compared with the upper edge (defined by contracting translations), which suggests that the presentation of motion signals within the far half-pattern is not enough to induce the edge displacement pattern of Demonstration 2. A mechanism of processing imbalance, operating in a similar way to the ‘speed imbalance’ used here must be introduced to account for the edge displacement in Demonstration 2 whose direction is predicted by the ‘centripetal bias’.