Does anybody have data on how Speedmasters performed in a zero-G environment? My Speedy (cal 1861) is notorious in my collection for losing 15 seconds dial side up and gaining 4 seconds on it’s side. Suffice to say, gravity has a profound effect on watch accuracies. Is it possible to achieve a +\- 0 second per day on a perfectly regulated watch in outer space!?

 

Possibly if there's no acceleration or other force which brings inertia into play. Then again that would mean hand winding only.
Perfect accuracy is unlikely in any case, but accuracy to less than one second per day has been achieved.

I suspect your Speedy would show no positional error in zero g though it might still gain or lose depending on wear and/or lubrication.

 

I would assume zero gravity would be the most accurate. The bigger issue that NASA tested for was the high and variable g’s during launch.

 

If resetting a watch midway of the journey using a radio transmitted time from the Naval Observatory the speed of the radio transmission must be taken into account. On a Mars journey the delay is significant.
The catch 22 is that unless you know the time exactly you can't calculate the distance.

 

Why only hand winding?

Gravity is far from being the only issue with regards to positional variation. Intertia and acceleration are at play when the balance is given an impulse by the pallet fork, and of course just from wearing the watch on the wrist as the wrist is moving around...

 

I was assuming he meant with the watch at rest. Moving the watch about to generate inertia in order to self wind would be user induced and might induce error.

In zero gee the balance would be unencumbered by changes in position just as a tourbillon is intended to operate.

 

The distance from the sun to the Earth was a great puzzle in the astronomical community for a few centuries. It's possible through geometry to calculate other solar system distances in terms of the distance from the sun to the Earth, but for a long time that distance was unknown in terms of terrestrial measurements. Eventually observing a transit of Venus from widely separated points on Earth, with some pretty hairy geometry, made that possible -- and then in the 1960s radar was bounced off Venus to confirm and extend the accuracy of the previous measurements.

 

From the earth you could determine the distance a spacecraft traveled and radio that information but it would not be so easy to determine from the spacecraft itself.

 

Pretty easy actually even if there was no accurate time device onboard. Ship pings earth for time gets time and subtracts delay divided by 2

 

I'll bet any error due to movement of the wearer wouldn't be any worse than on Earth, maybe less since astronauts in a tin can may not move their arms as much as someone on Earth. If the space user is reasonably active, it would be about the same. While positional error due to gravity would be eliminated, I think the watch might actually run LESS accurately since it was calibrated on Earth. However, for a well-calibrated 321/861/1861, I wouldn't think it would make a huge difference over week or two's worth of mission, particularly since the movement is non-hackable. (although I'm sure the astronauts knew the backward winding "trick" to hack the Speedy.)