The present study was designed to investigate the phenomenon of serial order from the considerations of the reading as well as the writing habits of Japanese, i.e., from the perceptual-motor aspect. In order to select subjects equalized on the point, a psycho-motor pretest was administered for once a week fifteen times to 139 woman college freshman students in an introductory psychology course. Eased on the results of pretest nine groups of 5 to 10 Ss each were selected.Three different types of stimulus patterns were employed; digits, filled and open circles and Hiragana letters (a kind of Japanese ordinary alphabet). Each pattern contained six elements in a horizontal row. Patterns were matched in pairs as far as possible, and were divided into A-series and B-series. They were presented in random order at an exposure duration of 100msec. Ss were requested to report what they saw, in verbal responses (V. R.) to A-series, and in writing responses (W. R.) to B-series.Experiment I was conducted to discover how stimuli were recongnized when they were presented at the left fixation point (L. P.), or at the right fixation point (each, N=5). No significant difference of recognition scores was found between L. P. and R. P., nor between V. R. and W. R.. However, recognition characteristics were significantly different between L. P. and R. P. (See Fig. 1). Experiment II dealt with the effects of left-right and right-left writing directions on recognition at the center fixation point (C. P.). Recognition scores of two groups (each, NN=5) were almost equal. In left-right directions, ordering of letter stimuli was found less flexible than that of digit stimuli. Probably, letters were several times more redundant than digits (See Table 2 and 3). Experiment III was designed to study the effects of horizontal and vertical writing directions on recognition at C. P. (each, N=5) (Japanese have these two reading and writing habits; left-right and up-down). Both recognition scores and characteristics of two groups were almost the same (See Fig. 2).Experiment IV was an attempt to throw a light on the relation between the pretest results and the accurately recognized stimuli (each, N=5). Recognition scores of the superior group in the pretest were significantly higher than those of the inferior.An attempt was made to find out how stimuli were recognized by V. R. and W. R. groups (each, N=10) in Experiment V, where A-series patterns were presented to R. P. and B-series to L. P.. Though scores of the W. R. group were almost equal for L. P. and R. P., scores of the V. R. group for R. P. in digits and letters were significantly less than for L. P. (See Fig. 3).With the results of Experiment I and II, two retinal effects (the end effect and the fixation-point effect) and two central processes (the sliding process and the gearing process) were postulated. The end effect means that the primitive figure-ground segregation occurs more rapidly for stimuli located at both ends, because three sides of them were completely open. The fixation-point effect implies that any stimulus subtended about 2° visual angle is more sensitively received because it falls on fovea centralis. The sliding process is assumed to act rapidly, roughly and driftingly, while the gearing process slowly, steadily and in a fixed direction.A paradoxical phenomenon observed in previous studies that the right visual field dominates when stimuli are presented on either side, while the left field dominates when stimuli are shown on both sides, was interpreted by the microgenetic point of view discussed above. Based on the results of Experiment III, IV, and V, it was pointed out that the sliding process works mainly in visual areas, or rather spatially, while the gearing process works mainly in motor areas, or rather temporally. It was suggested that an autonomic integration of “Habit” is a spatiotemporal