No abstract

The effect on matched size of the oculomotor adjustments was determined by stimulation and relaxation of accommodation and convergence by means of spherical lenses. The normal coupling between accommodation and convergence was maintained by introducing the amount of convergence appropriate to the lens power and each S's interpupillary distance. Data indicate that the oculomotor adjustments are adequate to account for size constancy up to approximately 1 m, beyond which their effect progressively decreases. The actual accommodation in force was assessed by means of the laser scintillation technique. It was determined that the magnitude of accommodation responds accurately to the spherical lens introduced up to about 1 m observation distance, beyond which underaccommodation was noted. Examination of the matched size as a function of the actual accommodation distance reveals a very close correspondence to the size constancy prediction up to about 1 m.

We investigated amodal completion with changes in two parameters: figure orientation, and the shape of the occluding and occluded figures. First, Markovich's [2002, Visual Mathematics 4(1); http://www.mi.sanu.ac.rs/vismath/fila] figure was presented in vertical, horizontal, and diagonal orientations. The results supported Markovich's finding of amodal completion by smooth continuity and symmetry. Amodal completion was also affected by orientation: symmetrical shape tended to appear when the pattern had vertical axis whilst smooth continuity appeared in diagonal orientation. Next, the shape of the exposed side of the occluded figure was changed from straight to curved. Finally, the shape of the grey occluder was changed from a disk to a half-disk or a quarter-disk. The results then showed the influence of the visible contour on amodal completion. This shows that the contextual relation of the whole figure is significant in amodal completion.

We examined whether the homogeneity of the two profiles of Rubin's goblet affects figure/ground perception in infants. We modified the two profiles of Rubin's goblet in order to compare figure/ground perception under four test patterns: (1) two profiles painted with horizontal lines (horizontal-line condition), (2) two profiles painted middle gray (uni-color condition), (3) one profile painted light gray and the other dark gray (two-color condition), and (4) a goblet painted with concentric circles (concentric-circles condition). In the horizontal-line condition the homogeneity of the profile was strengthened, and in the two-color condition the homogeneity of the profile was weakened compared to the uni-color condition, which was an original Rubin's goblet. In the concentric-circles condition the homogeneity of the reversed areas of the horizontal-line were strengthened. After infants were familiarized with each Rubin's goblet, the infants were tested on their discrimination between the two profiles and the goblet in the post-familiarization test. In horizontal-line condition, uni-color condition and concentric-circles condition infants showed a novelty preference for the two profiles in the post-familiarization test. On the other hand, in the two-color condition no preference was observed in the post-familiarization test. This means that infants perceived the goblet as figure and the two profiles as ground in the horizontal-line condition, the uni-color condition and the concentric-circles condition. We found that infants could not perceive the goblet area as figure when the homogeneity of the two profiles was weakened. It can be said that figure/ground perception in infancy is not affected by strengthened homogeneity, but is affected by weakened homogeneity.

The influence of modified forms of the Ponzo-illusion figure and densitygradient patterns on the apparent size of lines and circles was investigated. Consistent with previous reports, the predictions of constancy scaling theory were not confirmed. The results are shown to be dependent both on visual size-comparison mechanisms and judgmental variables.