It’s one of my favorite science fact and question to ask at parties. When I ask people when do people think humans thought for the first time light had a finite speed, the earliest date I got was 1950.
Speaking of favorite question, here’s a brain teaser. Give me a spot on earth from which you move south 30km, east 30km then north 30km and you end up in the same spot?
zodwick 1 hours ago [-]
I remember some vsauce (i think) video about how we dont know for sure if speed of light is constant in all directions. As a limitation of measurement we can only measure its speed in a round trip.
toast0 50 minutes ago [-]
It's not really possible to measure the time it takes light to move from point A to B. You can really only measure the time it takes for light to move from point A to B and back.
But you can do measure the difference between light going from point A to point B and light going from point A to point C to point B (and adding point C in the reverse direction as well) and I think you can use that to get reasonable confidence that light has constant speed in a given medium in all directions.
But this is totally relevant for trying to figure out asymmetric delivery delays in packet switched networks... You can't measure the delay directly, and if you use the network to synchronize your clocks, you can't use the clocks to measure the delay either, because most synchronization assumes symmetric delay. You need an externally synchronized clock, like a GPS reference etc.
function_seven 42 minutes ago [-]
Let's assume light goes faster in the "north" direction, and slower in the "south" direction.
Wouldn't the north- or south-component of the vectors to C cancel each other the same way they do when going from A–B vs. B–A? If the vector AB is on the north-south axis, and C is somewhere to the east, then the increase in speed from A–C–B would be offset by the decrease on the B–C–A return trip, giving you no information as to what direction is faster.
cyberax 12 minutes ago [-]
> It's not really possible to measure the time it takes light to move from point A to B. You can really only measure the time it takes for light to move from point A to B and back.
You absolutely can. Get two synchronized clocks, and start pulsing light at each tick. Then slowly move one clock away and observe how long it takes for the light pulse to arrive.
You don't even need crazy precision for that. A regular atomic clock would do.
Edit: if you didn't get the hint, that's exactly how GPS works.
glitchc 19 minutes ago [-]
While we cannot observe it because we lack an absolute frame of reference, we do know that the speed of light is constant in all directions due to the principle of super-position, provided the medium doesn't change between directions.
fooker 35 minutes ago [-]
I don't think it's just a limitation of measurement. It's a fundamental limitation of the universe, there's no absolute time so there's no accurate time delta between different frames of reference.
So your time measurement can only be valid for you.
A fun corollary of this sort of relative timing is that a dude casually walking by you when you're sitting still will observe the Andromeda Galaxy with a several day time difference.
Maxatar 42 minutes ago [-]
Not just light, it's impossible to know if the speed of anything is constant in all directions. The so called one-way speed of an object depends on a synchronization convention used for clocks at different locations. Different conventions can produce different one-way speeds while leaving all other observable physical phenomena unchanged.
The way I came to understand the speed of light, it is not so much that it is finite, but more so that time slows down the nearer you get and stops at light speed.
This means that going faster than light entails time travel into the past. While going just at the speed of light means you can travel any distance in exactly zero seconds. It's just that time passes for your surroundings.
newsicanuse 24 minutes ago [-]
[dead]
itchingsphynx 2 hours ago [-]
Even light, which travels so fast it takes most races thousands of years to realize that it travels at all, takes time to journey between the stars.
~Douglas Adams
Groxx 1 hours ago [-]
Bad news, however, travels far faster. Maybe instantaneously.
magicalhippo 1 hours ago [-]
Ridiculons[1] travels faster than the speed of light, so not unlikely bad news does too.
Yuk, please don't sully Adam's quotes with GenAI nonsense
magicalhippo 6 minutes ago [-]
It was posted on Usenet in 1997, as stated at the top of the Gist. I stumbled upon it around y2k.
Think it's fair to say it's not written by a generative AI.
I just linked to that since Google decided to wall in all old newsgroups postings.
16 minutes ago [-]
kstrauser 1 hours ago [-]
I don’t quite understand this. The length of the eclipse would be the same, right? Or was it more like that the eclipse ended 10 minutes earlier on the wall clock than expected? I guess what would surprise me about that is:
1. I didn’t know clocks were that good by then to make that reliably measurable.
2. I didn’t realize clocks of the time could measure the period of the eclipse that accurately. Like say one took 24 hours and 7 minutes. If you’re off by 30 seconds, and you’re making a prediction today of when the eclipse would happen six months from now, you could easily be off 90 minutes just based on cumulative error.
The discovery seems easy enough given today’s widespread access to atomic clocks, but extra freaking impressive with clocks from the 1600s.
dghlsakjg 34 minutes ago [-]
The eclipse is the same length. When an observer on Earth sees it will vary since our distance to Jupiter varies.
Think about lightning and thunder. The thunder happens the instant the lightning does (the lightning being the eclipse). When you hear the thunder depends on the you, the observer's, position. Imagine we know exactly where and when the lightning/eclipse happens. If hearing the thunder (seeing the eclipse) happens later than that time, we can then calculate the speed of sound based on our distance from the event, and the time we observed it. The final trick is that they didn't know when the eclipse would happen exactly, but they did know when it would be observed by an observer in a given location. If you then observe it from a second location that is farther away from the eclipse, you can determine the speed of light. They used the fact that the earth moves through space to change locations enough for the time differential to be super obvious.
Accurate clocks aren't super duper necessary. You just need one that can be accurate enough for a few hours (since you can fix your time based on celestial observations several times per day). Pendulum clocks were invented about 20 years before this discovery and were accurate to about 1 minute of drift per 24h.
adrian_b 1 hours ago [-]
The pendulum clock had been invented by Christiaan Huygens a couple of decades earlier.
It became quickly widespread, especially at astronomical observatories.
It is likely that the recent availability of accurate pendulum clocks was essential for Rømer's measurements, so such a result about the speed of light would have been impossible for the previous generation of astronomers.
Nonetheless, inside a fixed astronomical observatory, where there has been built a meridian circle with a telescope, and if there is available a good catalogue of stars, it is possible to measure the current "sidereal time" and use differences in sidereal times to compute the corresponding differences in time, even without a clock. The sidereal time is not a time, but the angle of rotation of the Earth vs. the stars. The rotation of the Earth vs. the stars is much more uniform than the apparent movement of the Sun on the sky, which determines the days.
So it would have been possible, but in a much more difficult way, to measure accurately the duration of some astronomical events even without clocks, but only with accurate instruments for measuring angles, used to determine the celestial coordinates of the stars, then compute the difference in the transit times of some stars whose transits coincided with the beginning and the end of the duration that must be measured, from the angles between them.
When good clocks are available, it is easier to measure the transit times of the stars and compute the angles between them from the times, than the reverse procedure, which would be needed without good clocks.
Epa095 55 minutes ago [-]
Io eclipses Jupiter with a certain frequency, let's say every 42.5 hour. This is easy enough to measure and verify. So if he noticed that at some point, when Jupiter is close to earth, that eclipse happens at some point during the night, it's trivial to calculate when during the day night months in the future the eclipse should happen.
The distance between the closest and furthest away Jupiter is from earth is ca 378 million km, which light uses ca 21 minute to traverse. So if he grounded his calculations from the time it was closest and then again observed it at it's furthest away (I think that's impossible as it's then behind the sun, but for arguments sake..) it will appear 21 minutes 'to late '.
The clock does not need to be able to keep time over many months, it just needs to be able to measure that its 12:21 and not 12:00.
canpan 1 hours ago [-]
To measure longer periods count the number of days. We have a built in clock which is relatively exact: The rotation of the earth. You only need a clock that measures time correctly during one day, much easier. Which existed at a time.
The orbital period is the same but there's phase drift in the occlusion times.
tome 41 minutes ago [-]
No, the apparent orbital period differs:
> there should be a difference of about 7 minutes between the periods of Io measured at the first quadrature and those measured at the second quadrature
Got it. I think it’s still very impressive that we could measure all that accurately enough to find a discrepancy, given the tech available then.
I think it’s the phrasing I’ve seen here and in similar articles that throws me:
> He said that on November 9, Io, one of Jupiter's moons, would emerge from the planet's shadow ten minutes later than expected.
What they meant was that the eclipse happened later than expected, both to start and to finish. From just that sentence, it reads to me like it started on time and finished later than expected. On further reading, that’s clearly not the case.
kazinator 2 minutes ago [-]
[delayed]
unrented7977 35 minutes ago [-]
HN should probably just blanket ban the word "this" in submission titles
wigster 37 minutes ago [-]
after the third popup i just bailed
cwnyth 28 minutes ago [-]
Firefox and ublock origin are life savers.
Terr_ 11 minutes ago [-]
Soon (if not already) it'll be that and "is this generated slop."
I wonder if we'll see a resurgence of "web of trust" stuff.
d-sanoj 1 hours ago [-]
It is just we are not able to comprehend it until someone pointed it out.
begueradj 2 hours ago [-]
>Journal des sçavans
Nowadays it's written savants.
goodmythical 1 hours ago [-]
According to [0], the journal has changed names twice, so "Journal des savants" would be referencing the modern journal that used to be called "Journal des sçavans"
https://a.co/d/0fZNXGxD
It’s one of my favorite science fact and question to ask at parties. When I ask people when do people think humans thought for the first time light had a finite speed, the earliest date I got was 1950.
Speaking of favorite question, here’s a brain teaser. Give me a spot on earth from which you move south 30km, east 30km then north 30km and you end up in the same spot?
But you can do measure the difference between light going from point A to point B and light going from point A to point C to point B (and adding point C in the reverse direction as well) and I think you can use that to get reasonable confidence that light has constant speed in a given medium in all directions.
But this is totally relevant for trying to figure out asymmetric delivery delays in packet switched networks... You can't measure the delay directly, and if you use the network to synchronize your clocks, you can't use the clocks to measure the delay either, because most synchronization assumes symmetric delay. You need an externally synchronized clock, like a GPS reference etc.
Wouldn't the north- or south-component of the vectors to C cancel each other the same way they do when going from A–B vs. B–A? If the vector AB is on the north-south axis, and C is somewhere to the east, then the increase in speed from A–C–B would be offset by the decrease on the B–C–A return trip, giving you no information as to what direction is faster.
You absolutely can. Get two synchronized clocks, and start pulsing light at each tick. Then slowly move one clock away and observe how long it takes for the light pulse to arrive.
You don't even need crazy precision for that. A regular atomic clock would do.
Edit: if you didn't get the hint, that's exactly how GPS works.
So your time measurement can only be valid for you.
A fun corollary of this sort of relative timing is that a dude casually walking by you when you're sitting still will observe the Andromeda Galaxy with a several day time difference.
https://en.wikipedia.org/wiki/Michelson%E2%80%93Morley_exper...
This means that going faster than light entails time travel into the past. While going just at the speed of light means you can travel any distance in exactly zero seconds. It's just that time passes for your surroundings.
~Douglas Adams
[1]: https://gist.github.com/pmarreck/7092d28f3d9bf8dff99faa61d01...
Think it's fair to say it's not written by a generative AI.
I just linked to that since Google decided to wall in all old newsgroups postings.
1. I didn’t know clocks were that good by then to make that reliably measurable.
2. I didn’t realize clocks of the time could measure the period of the eclipse that accurately. Like say one took 24 hours and 7 minutes. If you’re off by 30 seconds, and you’re making a prediction today of when the eclipse would happen six months from now, you could easily be off 90 minutes just based on cumulative error.
The discovery seems easy enough given today’s widespread access to atomic clocks, but extra freaking impressive with clocks from the 1600s.
Think about lightning and thunder. The thunder happens the instant the lightning does (the lightning being the eclipse). When you hear the thunder depends on the you, the observer's, position. Imagine we know exactly where and when the lightning/eclipse happens. If hearing the thunder (seeing the eclipse) happens later than that time, we can then calculate the speed of sound based on our distance from the event, and the time we observed it. The final trick is that they didn't know when the eclipse would happen exactly, but they did know when it would be observed by an observer in a given location. If you then observe it from a second location that is farther away from the eclipse, you can determine the speed of light. They used the fact that the earth moves through space to change locations enough for the time differential to be super obvious.
Accurate clocks aren't super duper necessary. You just need one that can be accurate enough for a few hours (since you can fix your time based on celestial observations several times per day). Pendulum clocks were invented about 20 years before this discovery and were accurate to about 1 minute of drift per 24h.
It became quickly widespread, especially at astronomical observatories.
It is likely that the recent availability of accurate pendulum clocks was essential for Rømer's measurements, so such a result about the speed of light would have been impossible for the previous generation of astronomers.
Nonetheless, inside a fixed astronomical observatory, where there has been built a meridian circle with a telescope, and if there is available a good catalogue of stars, it is possible to measure the current "sidereal time" and use differences in sidereal times to compute the corresponding differences in time, even without a clock. The sidereal time is not a time, but the angle of rotation of the Earth vs. the stars. The rotation of the Earth vs. the stars is much more uniform than the apparent movement of the Sun on the sky, which determines the days.
So it would have been possible, but in a much more difficult way, to measure accurately the duration of some astronomical events even without clocks, but only with accurate instruments for measuring angles, used to determine the celestial coordinates of the stars, then compute the difference in the transit times of some stars whose transits coincided with the beginning and the end of the duration that must be measured, from the angles between them.
When good clocks are available, it is easier to measure the transit times of the stars and compute the angles between them from the times, than the reverse procedure, which would be needed without good clocks.
The distance between the closest and furthest away Jupiter is from earth is ca 378 million km, which light uses ca 21 minute to traverse. So if he grounded his calculations from the time it was closest and then again observed it at it's furthest away (I think that's impossible as it's then behind the sun, but for arguments sake..) it will appear 21 minutes 'to late '.
The clock does not need to be able to keep time over many months, it just needs to be able to measure that its 12:21 and not 12:00.
For the other point, and how he did, i guess the article is badly written. This should be better: https://en.wikipedia.org/wiki/R%C3%B8mer%27s_determination_o... (he didnt need to measure months/years)
> there should be a difference of about 7 minutes between the periods of Io measured at the first quadrature and those measured at the second quadrature
https://en.wikipedia.org/wiki/R%C3%B8mer%27s_determination_o...
I think it’s the phrasing I’ve seen here and in similar articles that throws me:
> He said that on November 9, Io, one of Jupiter's moons, would emerge from the planet's shadow ten minutes later than expected.
What they meant was that the eclipse happened later than expected, both to start and to finish. From just that sentence, it reads to me like it started on time and finished later than expected. On further reading, that’s clearly not the case.
I wonder if we'll see a resurgence of "web of trust" stuff.
Nowadays it's written savants.
[0] https://fr.wikipedia.org/wiki/Journal_des_savants