Acceleration
This term in horology is applied to the steady gaining in the rate of a time-keeper, particularly to be observed in new movements. It is positively known to occur in marine chronometers, watches as a rule not being subjected to tests sufficiently accurate to detect it in them. There is but little doubt that the hairspring is the cause of acceleration. Old movements after being re-sprung sometimes accelerate, particularly if the overcoil is manipulated too much when timing. It is claimed that springs increase slightly in strength for some time after they are subjected to continuous action just as bells are found to alter a little in tone after use. It is said that the very best chronometers, after going for a year or two, will accelerate by about three or four seconds per day. M. Jacob attributes this acceleration to the fact that chronometers are exposed to heat oftener and for longer periods than to cold, and since the balance is thus more frequently contracted it follows that after a time the segments will not return exactly to their initial positions. There will therefore be necessarily a slight acceleration of the rate.
Dent believed that it was due to the combination of oxygen of the air with the steel hairspring, so that after a time its rigidity is increased. M. Villarceau attributed it to the influence of the escapement and that it arises from the fact that the impact communicating the impulse occurs before the balance has arrived at its neutral position.
M. H. Robert attributes it to the fact that the resistance opposed by oil at the pivots of the escape wheel differs from that at the pivots of the balance. Flat springs do not accelerate as much as those having overcoils. Palladium springs accelerate much less than hardened steel springs.