Having reached this point, I usually order the patient to practise all these with closed eyes. When he can do this, he begins to take one or two steps with shut eyes, first forward, then sideways, then backward. If he falter or move without freedom, he is kept at this until he does it confidently. Then exercises in following patterns traced on the floor are begun. In hospitals, or where bare floors are to be found, the patterns may be drawn with chalk. In carpeted rooms, which by the way are less suited for the work than plain boards or parquet floors, a piece of half-inch wide white tape may be laid in the required pattern, first in a straight line, later, as proficiency is gained, in curved, figure-of-eight, or angular patterns. The patient must be made to walk _on_ the line, putting one foot directly in front of the other, with the heel of the forward foot touching the toe of the one behind. Walking over obstacles is tried next. Wooden blocks measuring about six by twelve inches and two inches thick are stood on edge at intervals of eighteen inches and the patient walks over them, thus training several groups of muscles; the blocks are at first set in straight lines, then in curving patterns. An ordinary octavo book makes a good substitute for a block. If the trunk muscles are affected by the ataxia, further exercises are ordered for them, bending and twisting movements, picking up objects from the floor, etc. For the hands and arms, which, except in those very rare cases where the ataxia first shows itself in the upper extremities, seldom exhibit much incoördination in the primary and middle stages, the movements are the picking up of a series of different-shaped small articles, arranging objects like dominoes, marbles, or the kindergarten sticks in patterns, bringing the fingers of the two hands one after another together, or touching a finger to the ear or the nose, at first with open and then with shut eyes. With these methods, needing not more than twenty minutes three times a day, the ataxic symptoms sometimes rapidly diminish. In certain cases no other improvement will be observed, showing that what has taken place is of course not an alteration of the diseased nerve-tissues for the better, as no treatment can restore sclerotic spinal tissue to a normal state, but is merely a substitution of function, in which other and associated nerve-tracts have replaced in control the ones affected.