(1) a reduced error and greater uniformity of result at all speeds where visual control is added, in the case of both mean and constant errors and for all subjects;
(2) the mean errors for visual-control records show a rise along a line whose equation is approximately y = px, or the equation of a straight line, where p is an undetermined constant.
On the other hand,
(3) the eyes-closed mean errors show no increase or decrease in value during the entire series;
(4) the constant errors for visual-control records show a drop from positive errors to negative errors, along a line whose equation is approximately y = qx or the equation of a straight line, where q is an unknown constant, somewhat less in value than p in the case of mean errors; the constant error becomes 0 at about 120 beats;
(5) the eyes-closed constant errors follow the same equation for left-handed subjects, using the left hand, but all other cases suggest a curve of the parabolic form, having 0 constant errors at 60 and 180 beats and being convex upward.
Considering individual records for 14 cm.
A general survey of the charts suggests certain irregularities that call for explanation, for there will be sudden large increases in errors, that are explicable on the hypothesis that the subject has temporarily lost control of the moving hand, that is, that fatigue is to be noted.
While the purpose of the investigation has been to allow no lines to be ruled while the subject was conscious of any such feeling, there being a pause of any desired length to permit time for rest, it is to be noted that a considerable amount of recorded data as to fatigue shows that it is an unconscious or subconscious phenomena.
Further, the series of records have been arranged to occur from 20 to 200 beats and never in the reverse order, because of subjective limitations, so it is reasonable to expect that during the period of twenty minutes to one half an hour required for a series of records, there will be lapses of volitional control entirely beyond the ken of the subjects. It is to this cause rather than to pure chance that the results will be attributed. With this exception, the individual records show close agreement with their average.