It's time once again for our regular open thread. Talk about whatever you want, so long as it isn't culture war.
Overhauls are Understanding Hull Symbols, Auxiliaries Part 3, ICNW Part 5, the Arleigh Burke class, Standard Part 1, Evan's review of Turner Joy, and for 2025, Exotic Hulls Parts two, three and four, as well as Fighting the Last War.
Comments
This is a question inspired by my recent acquisition of a mechanical watch as a (early) birthday present. One difference between mechanical watches and quartz watches is that a mechanical watch is far less accurate than quartz. Whereas even cheap quartz watches are accurate to within a second a month or so, a mechanical chronometer will be off by several seconds even after a day. So given that the measurement of longitude required accurate timing of star sightings, how did mariners keep their clocks synchronized prior to the advent of electronic timepieces?
The short answer is marine chronometers are not nearly as constrained as wrist watches and are much better.
For example, the escapement used in a wrist watch is not orientation independent; it runs at a different rate depending on which way gravity goes. The watch is factory set with a rate that ought to correspond to average orientation experienced during casual wear, give or take. Nowadays balance wheels that help correct this deficiency (tourbillon) are reserved for extremely expensive watches mostly for marketing reasons. As an interesting consequence of this, there is a market for artificial wrists that move your watch collection around while not wearing them in order to simulate casual wear and reduce the runout. Similarly, wrist watches basically never contain temperature compensation. They are factor set assuming they are warmed by body temperature in moderate ambient conditions. When you take them off and they are no longer warmed by body temperature they run at very a different rate, whereas changes in ambient temperature make them run at modestly different rates. Said artificial wrist may include a heater.
By way of illustration, during the first sea trial of John Harrison's first sea watch (H4), the runout was about 5 seconds over 81 days. Corresponding to a longitude error of about 1 nautical mile. They only got better from there.
The short answer is marine chronometers are not nearly as constrained as wrist watches and are much better.
For example, the escapement used in a wrist watch is not orientation independent; it runs at a different rate depending on which way gravity goes. The watch is factory set with a rate that ought to correspond to average orientation experienced during casual wear, give or take. Nowadays balance wheels that help correct this deficiency (tourbillon) are reserved for extremely expensive watches mostly for marketing reasons. As an interesting consequence of this, there is a market for artificial wrists that move your watch collection around while not wearing them in order to simulate casual wear and reduce the runout. Similarly, wrist watches basically never contain temperature compensation. They are factor set assuming they are warmed by body temperature in moderate ambient conditions. When you take them off and they are no longer warmed by body temperature they run at very a different rate, whereas changes in ambient temperature make them run at modestly different rates. Said artificial wrist may include a heater.
By way of illustration, during the first sea trial of John Harrison's first sea watch (H4), the runout was about 5 seconds over 81 days. Corresponding to a longitude error of about 1 nautical mile. They only got better from there.
A mechanical chronometer for marine navigation is much more accurate than a mechanical watch, and the early ones came into use because they didn't need synchronisation. They're also precision bits of machinery. John Harrison's H-3 took about twenty years of research and development, and weights 27 kg. There's a good book "Longitude" by Sobel, Dava about the problem and how it was tackled.
I didn't realize that even early marine chronometers were that accurate.
Also, while I was looking into it, I found out that the US was transmitting time signals by radio as early as 1904, which allowed ships at sea to synchronize their chronometers at sea.
personofinterest:
A quartz watch could do that at the expense of power usage (for real accuracy atomic clock chips small enough to fit in a wristwatch are available, but even ordinary quartz watches are accurate enough that you could usually get away with only setting them when the cell is replaced) but could a mechanical watch do that and still fit on a wrist?
quanticle, consider reading Dava Sobel's Longitude, in case you haven't already read it. It's pretty wild looking back that some guy way out in the countryside managed to build a series of chronometers when the elite watchmakers were struggling to get anywhere near the required performance standard. The book is dramatised, but OTOH the story is of a total outsider doing the impossible so some drama is warranted.