thirty-five seconds, and during this time his hand on the automatograph moved from A to A'. At the latter point he was asked to count the oscillations of a pendulum; this entirely changed the movement, the hand at once moving rapidly toward the pendulum. The pendulum was a more attractive sense-impression than the colors; the special point of interest in this record is, that upon
Fig. 11.↦ I, Counting Metronome. Automatograph. Facing ↦. Time, 35 seconds.↦ II, Counting Pendulum. Automatograph. Facing ↦. Time, 25 seconds.
examination the subject's color-vision proved to be defective and thus explained the failure of the colors to hold his attention.
An important problem relates to the possible correlation of types of involuntary movements with age, sex, temperament, disease and the like. A few observations upon children are interesting in this regard. They reveal the limited control that children have over their muscles, and how difficult it is for them to fix the attention when and where desired. The movements they make are large, with great fluctuations, and irregularly toward the object of attention. Fig. 13 illustrates some of these points; in thirty-five seconds the child's hand moved by large steps seven
Fig. 12.— ↦ Facing ↦. Hand on Automatograph. From A to A', reading colors, 35 seconds. From A' on, counting pendulum, 25 seconds.
inches toward the pendulum, and the entire appearance of the outline is different from those obtained upon adults.
Much attention has recently been paid to automatic writing, or the unconscious indication of the nature, not the direction, of one's thoughts while the attention is elsewhere engaged. We attempted this upon the automatograph by asking the subject to view or think of some letter or geometric figure, and then search-
