The process of securing a record is as follows: the subject, standing, places his hand upon the automatograph, with the arm nearly horizontal and not quite fully extended, and the elbow bent in a fairly comfortable posture; his attention is engaged by asking him to listen to and count the strokes of a metronome; to look at and count the oscillations of a pendulum; to read from a book; to call out the names of colors; to think of a given direction or locality, or the position of an object; and so on. He is instructed to think as little as possible of his hand, making a reasonable effort to keep it from moving. To cut off the apparatus from the subject's field of vision and attention, a large screen is interposed between him and the record, a curtain with a suitable opening for the arm forming part of the screen. The operator holds the glass pencil in his hand, and when all is in readiness allows it to slip through the glass tube and begin to write, removing it again after a definite interval or when the record seems completed.
[Illustration: FIG. 2.--READING COLORS. Time of record, 95 seconds. Position of colors →. Subject facing →. In all the figures A represents the beginning of the record, and Z the end. The arrows are used to indicate the direction in which the object attended to was situated, and also the direction in which the subject was facing. The tracings are permanently fixed by coating them with a weak solution of shellac in alcohol.]
We may now consider a few typical results. Fig. 2, an ordinary average result, was obtained while the subject was calling out the names of a series of small patches of color, displayed on the wall facing him, about eight feet distant. It will be observed that the movement (which in all the illustrations has its beginning marked by an A and its end by a Z) proceeds irregularly but decidedly _towards_ the object upon which the attention was fixed. As a rule the subject is unaware of the movement which his hand has made, and exercises no essential control over the results; indeed it is likely that he is considerably surprised when the results are first shown to him. At times he becomes conscious of the loss of equilibrium of the apparatus, but the indication is rarely sufficiently definite to inform him of the direction of the movement. Not infrequently, the movement is performed with complete unconsciousness, and is accompanied by a strong conviction that the apparatus has been stationary. In several cases an intentional simulation of the movements was produced for comparison with the involuntary records; the result was quite generally a very different and coarser type of movement, readily distinguishable from the involuntary writings. A prominent characteristic of practically all of the movements is their irregular and jerky character; the hand for a time oscillates about uncertainly, and then moves rather suddenly and quickly in a given direction; then another period of hesitation, again a more or less sharp advance, and so on. It is probable that it is these repeated brief movements of more vigorous indication of the direction of the subject's attention, that the muscle-reader waits for and utilizes.