The present paper was to analyze experimentally how the immediate reproduction of tactually perceived geometrical figures was influenced by some factors of touching such as (i) the movement of the hand, (ii) the starting point of touching, and (iii) the extent in tactual surface and others, and to examine what S's drawing process in reproducing the figures were, in order to find out the conditions necessary for the tactile apprehension of figures.In Exp. I, the two conditions of touching were set; (a) the condition of active touch in which S with closed eyes was asked to feel actively, moving his forefinger once along the contour of the figures, and (b) the condition of passive touch in which S was asked to hold constantly the finger on the same position and to feel once, passively, the contour of the figures which were moved by E. The four starting points of touching were set in each stimulus figure (Fig. 1). After touching a figure S was asked to reproduce it with open eyes. The results were: (1) The apprehension of figures in Active touch was better than in Passive touch (Table 2 and Fig. 2). (2) Point a was the most favorable starting point of touching in the scalene square. In the square and the lozenge, the starting points and the conditions of touching were interacting, and the favorable points for apprehension varied with the variation of stimulus figures (Table 1 and Fig. 2). All these favourable points, however, coincided with the starting points or the terminal points of open eyed drawing of the same figure (Table 6). Thus they were relevant to the drawing processes of figures in some respects. (3) In both Active and Passive touch, most of the S's reproduced the figures in uninterrupted lines, repeating the same course as in the touching condition (“continuous reproduction”). On the other hand, “dividing construction” which one employed ordinarily with open eyes in drawing the figures, scarecely appeared (Table 4).In Exp. II a four segment (Table 7) and in Exp. II b eight patterns (Table 8) were used as stimulus figures. Besides the conditions of Active and Passive touch, the two conditions as to the direction of touching were set in each figure (Table 7 and 8). The results in Exp. II a were: (1) Only the perception of the orientation of the segments in Active touch was better than that in Passive touch (Table 7). (2) For the vertical and horizontal segments in the active condition, the directions of touching which were favorable for perceiving tend to correspond to those of drawing of the lines (Fig. 4). In Exp. II b, on the other hand, it was observed that (3) the perception of the angles contained by two segments, that of the orientation of each segment, and that of the equivalence between a vertical segment and a horizontal one were all better under the active touch condition than were under the passive touch condition (Table 8 and 9), (4) In Active touch of the stimulus patterns consisting of a horizontal segment and a vertical one, the touching in the direction opposite to that of the drawing was more favorable for the correct perception than the touching in the same direction as that of drawing, in contrast with Exp. II a (Table 8). (5) In Exp. II a and II b most of Ss employed the “continuous reproduction”.In Exp. III Ss were allowed to feel a figure freely to their satisfaction, with one finger, one hand or both hands, respectively. The results were: (1) As the condition of touching shifted from “single touching” (Exp. I) to “free touching” with one finger (Exp. III), the perception of a figure gradually improved (Fig. 6). (2) The increase of tactual surface, however, did not necessary result in the improvement of the perception of a figure (Fig. 5). (3) “Dividing construction” remarkably increased in every condition of Exp. III (Table 13). (4) It was suggested that if