How the Speed of Light was First Measured

Karl Smallwood 21
lightThe speed of light in a vacuum stands at “exactly 299,792,458 metres per second“. The reason today we can put an exact figure on it is because the speed of light in a vacuum is a universal constant that has been measured with lasers; and when an experiment involves lasers, it’s hard to argue with the results. As to why it comes out somewhat conspicuously as a whole number, this is no coincidence- the length of metre is defined using this constant: “the length of the path travelled by light in vacuum during a time interval of 1/299,792,458 of a second.”

Prior to a few hundred years ago, it was generally agreed or at least assumed that the speed of light was infinite, when in actuality it’s just really, really, really fast-  for reference, the speed of light is just slightly slower than the fastest thing in the known universe- a teenage girl’s response time if Justin Bieber were to say on Twitter, “The first to reply to this tweet will be my new girlfriend.”

The first known person to question the whole “speed of light is infinite” thing was the 5th century BC philosopher Empedocles.  Less than a century later, Aristotle would disagree with Empedocles and the argument continued for more than 2,000 years after.

One of the first prominent individuals to actually come up with a tangible experiment to test whether light had a speed was Dutch Scientist, Isaac Beeckman in 1629. Despite living in a time before lasers- which gives me the chills just thinking about- Beeckman understood that, lacking lasers, the basis of any good scientific experiment should always involve explosions of some kind; thus, his experiment involved detonating gunpowder.

Beeckman placed mirrors at various distances from the explosion and asked observers whether they could see any difference in when the flash of light reflected from each mirror reached their eyes. As you can probably guess, the experiment was “inconclusive”.

A similar more famous experiment that didn’t involve explosions was possibly conducted or at the very least proposed by Galileo Galilei just under a decade later in 1638. Galileo, like Beeckman also suspected that the speed of light wasn’t infinite and made passing references to an experiment involving lanterns in some of his work. His experiment (if he ever conducted it at all), involved placing two lanterns a mile apart and trying to see if there was any noticeable lag between the two; the results were inconclusive. The only thing Galileo could surmise was that if light wasn’t infinite, it was fast and that experiments on such a small scale were destined to fail.

It wasn’t until Danish Astronomer, Ole Römer entered the fray that measurements of the speed of light got serious. In an experiment that made Galileo flashing lanterns on a hill look like a primary school science fair project, Römer determined that, lacking lasers and explosions, an experiment should always involve outer space.  Thus, he based his observations on the movement of planets themselves, announcing his groundbreaking results on August 22, 1676.

Specifically, while studying one of Jupiter’s moons, Römer noticed that the time between eclipses would vary throughout the year (based on whether the Earth was moving towards Jupiter or away from it). Curious about this, Römer began taking careful notes about the time I0 (the moon he was observing) would come into view and how it correlated to the time it was usually expected. After a while, Römer noticed that as the Earth orbited the sun and in turn got further away from Jupiter, the time Io would come into view would lag behind the expected time written down in his notes. Römer (correctly) theorised that this was because the light reflected from Io wasn’t travelling instantaneously.

Unfortunately, the exact calculations he used were lost in the Copenhagen Fire of 1728, but we have a pretty good account of things from news stories covering his discovery and from other scientists around that time who used Römer’s numbers in their own work. The gist of it was that using a bunch of clever calculations involving the diameter of the Earth’s and Jupiter’s orbits, Römer was able to conclude that it took around 22 minutes for light to cross the diameter of Earth’s orbit around the Sun.  Christiaan Huygens later converted this to more commonplace numbers, showing that by Römer’s estimation, light traveled at about 220,000 kilometres per second.  This figure is a little off (about 27% off) from the figure noted in the first paragraph, but we’ll get to that in a moment.

When Römer’s colleagues almost universally expressed doubt in his theory about Io, Römer responded by calmly telling them that Io’s 9th of November eclipse in 1676 was going to be 10 minutes late. When the time came, the doubters stood flabbergasted as the movement of an entire celestial body lent credence to his conclusion.

Römer’s colleagues were right to be astounded in his estimation, as even today, his estimation of the speed of light is considered to be amazingly accurate, considering it was made 300 years before the existence of both lasers, the internet, and Conan O’Brien’s hair. Okay so it was 80,000 kilometres per second too slow, but given the state of science and technology at the time, that is remarkably impressive, particularly given he was primarily just working off a hunch to begin with.

What’s even more amazing is that the reason for Römer’s estimation being a little too slow is thought to have less to do with any mistake on his part and more to do with the fact that the commonly accepted diameter of the Earth’s and Jupiter’s orbits were off when Römer did his calculations. Meaning yes, Römer was only wrong because other people weren’t as awesome at science as he was. In fact, if you slot the correct orbit numbers into what is thought to be his original calculations from reports before his papers were destroyed in the aforementioned fire, his estimation is nearly spot on.

So even though he was technically wrong and even though James Bradley came up with a more accurate number in 1729, Römer will go down in history as the guy who first proved that the speed of light was not infinite and worked out a reasonably accurate ballpark figure on what the exact speed was by observing the movements of a speck orbiting a giant ball of gas positioned about 780 million kilometres away. That right there ladies and gentlemen is how a badass, lacking lasers, does science.

If you liked this article, you might also enjoy:

Bonus Facts:

  • The energy required to stop the Earth orbiting the sun is about 2.6478 × 10^33 joules or 7.3551 × 10^29 watt hours or 6.3285*10^17 megatons of TNT.  For reference, the largest nuclear explosion ever detonated (the Tsar Bomba by the Soviet Union) “only” produced 50 megatons of TNT worth of energy.  So it would take about 12,657,000,000,000,000 of those nuclear bombs detonated at the correct location to stop the Earth from orbiting the sun.
  • Aside from the debate over whether the speed of light was infinite or not, a common side debate throughout history was whether or not light originated in the eye itself or from something else.  Among the famous scientists to believe in the “light emitted from the eye” theory were Ptolemy and Euclid.  Most who thought this theory correct also thought the speed of light must be infinite, because the instant we open our eyes, we can see a vast number of stars in the night sky and that number does not increase the longer we look, unless of course we were previously looking at a bright light and our eyes are adjusting to darkness.

Expand for References

Enjoy this article? If so, get our FREE wildly popular Daily Knowledge and Weekly Wrap newsletters:

Subscribe Me To:  | 
Print Friendly
Check Out Our New Book!»

21 Comments »

  1. Wojtek August 14, 2013 at 3:51 am - Reply

    The reason why the speed of light has the value you mention has nothing to do with lasers.
    The speed of light is a constant. It is described in distance / time. The meter being described as the distance light travels in vacuum during 1 / the_value_you_mentioned seconds, the only free variable we have is the time.

    One second is defined as (apologies for the physicists who came up with that :)) “the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom”.
    In essence this means we measure the time it takes for an atom to get back to normal state after having been “excited” (we measure emitted radiation)

    • Wojtek August 14, 2013 at 3:53 am - Reply

      Well, the anti XSS mechanism removed my formatting…

      “It is described in /” should actually be “it is described in distance / time”

      “during 1/ seconds” should be “1 / the_value_you_mention seconds”

      • Daven Hiskey
        Daven Hiskey August 14, 2013 at 10:40 am - Reply

        @Wojtek: Yep. Dang filters. Fixed it for you in your original comment. :-)

    • Daven Hiskey
      Daven Hiskey August 14, 2013 at 10:38 am - Reply

      @Wojtek: Great comment. :-) On the lasers bit, I think Karl was just saying we’ve been able to very accurately measure the speed of light in a vacuum with lasers, as mentioned here “After 1970 the development of lasers with very high spectral stability and accurate caesium clocks made even better measurements possible.”

  2. LOVEPAREEK August 14, 2013 at 8:33 am - Reply

    THANKS FOR THE EXCELLENT SHARE, KARL. PLS KEEP POSTING LIKEWISE.

  3. Wojtek August 14, 2013 at 11:46 am - Reply

    @Daven: Sorry for not having been clear in my comment. You are correct with the lasers but the concept itself of the units changed.

    In the past the meter was defined as “many times the length of a very small wave”. The second was (and still is) defined as I mentioned: “many times a very small time”.
    So we had our meter and our second. With this we could measure the speed of light (v = s / t) — with a predictable and small uncertainty.

    Then things changed (I do not remember the year): the definition of the second remained as it but the meter was now defined as “the distance the light travels in vacuum during a given time” (the given time is 1/299,792,458 s).
    We assume we know exactly the speed of light and we derive other units from this. This is a Very, Very Bad Thing (C) because our units should be based on basic physical effects and be independent (the meter and the second are not independent as one is derived from the other).

    Bonus fact: the meter was once a piece of metal, it is today stored at Sèvres, near Paris. It is not possible to visit it (I tried :))

    • Daven Hiskey
      Daven Hiskey August 14, 2013 at 2:05 pm - Reply

      @Wojtek: Great comments. Thanks!

  4. Jumbybird August 14, 2013 at 6:35 pm - Reply

    You mentioned teenaged girls and a certain Canadian fool in a n effort to be funny, and I stopped reading there.

    • Daven Hiskey
      Daven Hiskey August 14, 2013 at 7:00 pm - Reply

      @Jumbybird: So you’re saying you started reading it, presumably interested in the topic from the title, then not only did you not care to learn the heart of the interesting fact you originally set out to because of one sentence in a 1400 word article, but you then felt inclined to spend time scrolling down and writing a comment to announce to the world that you didn’t think something was funny? I mean, everybody has their own sense of humor, but that just seems odd. :-)

      • Larry August 16, 2013 at 9:25 am - Reply

        Nice article. At first I was thinking of the Michaelson-Morely experiment, but I was pleased to learn of this early attempt. It reminds me of Eratosthenes’ measurement of the circumference of the Earth.

        As for jumbybird, I can think of several lakes into which that bird can jumbp … I think it was in a James Bond novel that one character said of another, “… he ain’t got no sensayuma.”

      • Tom August 25, 2013 at 8:24 am - Reply

        Very good article, thank you. It was so interesting that I can’t even remember the stuff that this bird guy complains of.

  5. Ali September 24, 2013 at 9:04 am - Reply

    The argument took place for over 2,000 years after Empedocles om the fifth century? Maybe I’m missing something, but last I checked it wasn’t the 2400′s.

    I’m finding most of the reporting on this site to be hurried and exaggerated. It seems a fine placed for a little entertainment, but maybe one’s own research is required before quoting, or citing it’s material.

    • Daven Hiskey
      Daven Hiskey September 24, 2013 at 11:13 am - Reply

      Ali: You’re missing the critical “BC” part of the “5th century”.
      .
      And to clarify, what we are experts here is on two things: research, and making grammatical typos. ;-) You will not find many factual errors in any of the articles here, and if you do, you should definitely let us know because we will then re-research the thing and correct any mistakes as necessary.
      .
      The guiding precept of this site is providing a place where people can go and know what they’re reading has been extremely well researched and triple checked on every point. Nobody bats a thousand, of course, not the OED, Encyclopedia Britannica, Snopes, the Oxford Dictionary of Quotations, the QI books (and show), Uncle John’s Bathroom Reader, The Straight Dope, Mental Floss, Neatorama, the Barnhart Concise Dictionary of Etymology, etc. etc. I’ve personally found several errors in each of those, and they are all among the best. But we still strive towards everything on this site being perfectly accurate. It’s the number 1 rule when I hire a new author- Accuracy is King. I care far more about that than if someone makes a typo or has an awkward sentence or the like. :-)

  6. Ali September 24, 2013 at 2:52 pm - Reply

    Thanks for claryfying.

Leave A Response »