- 10 Mar 2013 [14:28]
- So here is a theory: If the temperature down at the pendulum bob is high then the temperature compensation is not dealing with change in air density (buoyancy). This effect is +30ms/day per degC so a 4C warmer bottom would give rise to a +120ms/day on going. This is not far from observed - note that the clock slower down overnight as the pendulum chute cooled down. Maybe I need to open up the chute to contrive that cool air can get down to the bottom. I might do that next weekend....
- 10 Mar 2013 [12:12]
- pendulum temperatures [15.0 14.3 14.3 11.1 9.9] and note that the minima overnight were [10.9 11.0 11.0 10.9 9.8] so it looks as if the pendulum bottom half cools down a lot overnight.