In three experiments, subjects viewed two lines (arms) of dots in which one dot was shared by both arms, then judged with which arm the shared dot appeared to group. Length of an arm, density of the dots in the arm, and angle between arms were manipulated. Increasing the density of an arm increased grouping with that arm regardless of the angle between arms. Increasing the length of an arm, however, increased grouping when arms met at 30° but not at 180°; intermediate angles produced intermediate effects. Advantages of the present method's simplification of the stimulus display are discussed.