keep it from moving. A large screen is interposed between the subject and the record to prevent him from indirectly seeing what is going on. On the wall facing him, some eight feet distant, are some small patches of color, the names of which he is asked to call out. The colors are small enough to necessitate close attention in their distinction, and the record of the hand, after the subject has been employed in this way for a minute or two, is usually quite significant. An average result is presented in Fig. 1. The hand 
Fig. 2. — Reading Colors arranged in Three Rows. Shows movement of the hand parallel with movement of the attention. Time of record, 90 seconds. The first line was read in this direction ↧, the second in this ↥, the third in this ↧. At the turn from the second to the third line the record is interrupted. moves clearly and directly toward the wall where the colors hang; the movement is at times halting and uncertain, but its general trend is unmistakable. Moreover, the result can not in general be anticipated, not alone because there are marked differences between individuals in the readiness with which they will manifest involuntary movements, but also because the intensity of the attention and the momentary condition of the subject are important and variable factors in the result. With very good subjects it becomes quite safe to predict the general nature of the result, and the different tracings of the same subject 1 )ear a family resemblance to one another.
A more unusual but very striking form of involuntary movement is shown in Fig. 2. As before, the subject's attention was fixed upon the colors on the wall, but these were arranged in three rows, the first being read from left to right, the second from right to left, and the third from left to right again. The record plainly indicates where the change of direction of reading took place; the correspondence between the movements of the hand and of the
