The next point of consideration may be the concentricity of the hairspring, and it is quite important that the spring be centered as nearly perfect as the trained eye can determine. Any unusual pressure of the spring in one direction will cause undue friction and a fast rate compared to the opposite direction.
There are several easy tests for determining as to how nearly the spring may be centered. One of these is to look straight down upon the spring and examine the space between the coils that extend beyond the circumference of the dome. This test may be made in three ways, one with the balance at rest, one with the coils of the spring wound up and the third with the coils unwound. With the balance at rest and the spring centered there will be the same space between the coils all around as though the spring were out of the watch entirely and laying on the bench.
If it is not properly centered there will be more space between the coils on one side than there will be on the opposite. The same conditions will be apparent when the spring is wound up, although the coils will all be nearer to each other than they were with the balance at rest, and when they are unwound the coils will all be farther apart with the same apparent difference on opposite sides when the centering is not correct.
The winding and unwinding of the spring is alternating and almost instantaneous, as the balance oscillates from one extreme to the other. For observation of the spring when it is wound or unwound it is necessary to stop the balance with the finger or camel's hair brush as it reaches its extreme arc of motion, then hold it stationary for a few seconds while the space between the coils is being examined. The balance should then be allowed to swing to the opposite extreme, when it should again be held for examination of the coils. In one of these extremes the coils will be wound and in the other they will be unwound and after a few experiments in stopping and starting the balance it will be found that the entire examination will not require over ten seconds' time.
When the spring is not properly centered the reason is of course found in some curve of the over coil and the most usual point at fault is the section or curve on which the regulator pins act. If the coils open too wide on the side where the regulator pins are located this section of the coil will be too near the center and should be moved outward, possibly equal to one-half or one full space of the coils. If the coils are too close on the side where the pins are it will probably be found that the section requires shifting toward the center slightly. The balance should be removed from the watch in either instance and the coil circled with the over-coiling tweezer, although experienced workmen can frequently make excellent corrections with a fine pointed tweezer without removing the balance.