XXII. Hold up the forefinger and thumb of the disengaged hand, their points being at two and a half inches apart. Hold the odometer in the centre of a line which would join the points of the finger and thumb. Result—Oscillations transverse to the line indicated.
XXIII. Modify the preceding experiment by holding the odometer half an inch to one side of, and over, the middle of the line indicated. Result—Oscillations parallel to the said line.
XXIV. Modify experiment XXIII. by approximating the ends of the forefinger and thumb of the disengaged hand, so that they touch. Result—The odometer no longer moves.
XXV. Forefinger and thumb of the disengaged hand held upwards and apart, sustaining a short file longwise between them. Odometer then held over the last joint of the finger. Result—Odometer stationary. Odometer then held over the last joint of the thumb. Result—Odometer stationary.
XXVI. Odometer held over the northward pole of the magnetic needle, and its consequent rotatory motion in the direction of that of the hands of a watch established. Then advance the finger or the thumb of the other hand towards the odometer. (The odometer should be held in these experiments half an inch above, and a little wide of, or before, the apex of the needle.) The finger, or the thumb, is then to be brought as near to the odometer as is consistent with not touching it in its rotation. Result—Direction of the rotation reversed. Then join the finger and thumb, and hold the two thus brought into contact in the same proximity to the odometer. Result—The rotation returns to the former direction; that is, to the direction of the motion of the hands of a watch.
XXVII. Odometer held over the radial (or thumb) edge of the wrist. Result—The same as when held over thumb. Odometer held over the little-finger edge of the wrist. Result—The same as when held over either of the fingers. This difference in result extends a third the length of the fore-arm, over the middle of which the odometer becomes stationary.
XXVIII. A portion of rock-crystal five inches long, about two wide and deep, placed on the table with its long axis transverse to the operator. Glass odometer held over the middle of the upper plain surface. Result—Oscillations parallel to the axis of the crystal. Position of the crystal shifted, so as to make its axis point from the operator. Result—Oscillations as before parallel to the axis of the crystal, but longitudinal to the operator. Then the thumb applied to the forefinger. Result—Transverse oscillations.
XXIX. Glass odometer held suspended over one apex of the crystal. Result—Rotatory motion in the direction of the hands of a watch. Odometer held over the opposite end. Result—Rotation in the direction contrary to that of the hands of a watch.
XXX. The last experiment repeated. The forefinger of the operator’s unengaged hand brought near to the odometer in each of its two varieties. Result—The previous rotatory motion reversed. Then the point of the thumb brought into contact with the odometer finger. Result—The original rotatory motion re-established.
I will add in reference to the first and simplest experiments, that the interposition of several folds of silk between the Od-subject and the odometer renders the motions of the latter less brisk.