When the balance is slightly out of poise and the motion is exactly one and one-fourth turn during the twenty-four hours, this out of poise will not affect the isochronism. When the motion varies and reaches approximately one and one-half turn during the first few hours after winding and then drops to one and one-quarter turn and finally to one turn or less during the latter part of the twenty-four hours, the poise error will have considerable effect. This factor is not perceptible in the flat positions, but shows up to the full extent in the vertical positions and the variation differs according to the location of the point that is heavy. For example, if the balance is heavy on the lower side when at rest, the watch will lose during the hours that the arc of motion is over one and one-fourth turn and will gain when the motion drops to one turn or less.

Should the heavy point be on the top side of balance the result will be reversed and the watch will gain when the motion is over one and one-fourth turn and will lose when it drops to one turn or less.

Rules and Practice for Adjusting Watches — Part 28

The total variation may be either seconds or minutes, depending upon the extent of the poise error and experiments will prove that serious isochronal variations can be traced to the simple cause of lack of poise and irregular motion in more instances than to any other cause.

Rules and Practice for Adjusting Watches — Part 28

The arc of one and one-fourth turn is the ideal motion, as slight poise errors are neutralized at this point, but very few watches will maintain this motion for twenty-four hours, therefore the poise must be as nearly perfect as possible. The nearest approach to even motion of modern watches is found in the fine Swiss grades equipped with stop work, which causes only the best part of the mainspring to be utilized.

Rules and Practice for Adjusting Watches — Part 28

Such watches also receive the most expert attention as to gearings of wheels and pinions and the train wheels are specially rounded up on their respective staffs. This latter feature has been adopted by at least two of the American manufacturers of fine watches during the past few years with considerable benefit in producing even motion and the use of lighter mainsprings. It should be definitely understood that these tests refer to the vertical positions of the watch only and that the horizontal positions are not affected in the same way by lack of poise.