Moving a wooden clock

Motivated by an oncoming winter storm -- the clock was blocking our generator -- I decided to move Clock #1 upstairs to our dining room.  I had heard of various troubles with moving wooden clocks, so I was ready to spend at least a little time adjusting things.  The clock ran for a few hours before becoming very temperamental.  I figured that either the humidity (drier) or the temperature (warmer) of the upstairs as opposed to the clock's previous home in the basement was to blame.  So I set about trying to figure out what changed.  Although I didn't know it, neither humidity nor temperature appear to be at fault.  Since I spent the better part of the day at this task, I figured I ought to record the eventual cause for posterity.

Long story short: the escape wheel arbor is a bit too short, and so it doesn't seat fully in both bushings.  It tends to sit in either the front or the back bushing.  If the clock is set slightly tipped forward (as it was in the basement, since the floor is not level), then it sits in the front bushing.  This keeps everything aligned and all is well; the clock runs.  This is especially important because the escape wheel pinion is not perfectly cut.  If the escape arbor sits in the back bushing, the escape pinion wanders around and infrequently fouls on the previous wheel in the train.  Worse, it only partially fouls, and so steals just enough energy to cause the clock to stop with the escape wheel in some other, later position.  This was quite maddening as there appeared to be no obvious pattern to where the train stopped!

So... when the clock was first installed (perfectly level), the arbor was evidently seated in the front pivot.  But since gravity was uncertain, the arbor eventually drifted to the back pivot, resulting in a stopped clock with no obvious cause.  It was only after I had retraced everything and realized that the clock was not level when it was happily running in the basement.