I have heard the about the polarity/bell tap issue enough over the years to know it must be true, and I have seen it. I never knew why it was, but recently I heard an explanation that makes sense [I think it may have been here in The Forum].
If you look at a standard double-gong ringer [C type, F type, etc.], you will notice that when the clapper is at rest, it is against a stop. It can really only move in one direction off the stop, and if the polarity of the inductive voltage spikes caused by pulse dialing move it in that direction, it can tap. Reverse the connections of the bell coil, and the spikes will pull the clapper towards the stop, and it can't actually move [or tap].
As Skip said, usually just tweaking the bias spring will cure most cases of bell tap. Also, Most WE rotary dials had capacitors and/or varistors across the dial contacts to limit the spikes caused by pulse dialing.
Jim
**************************************************
Speaking from a secure undisclosed location.
