Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Measuring the Speed of Light (Foucault Method)

  1. Dec 2, 2012 #1

    In this experiment, a beam of light from a laser was reflected off of a rotating mirror to a fixed mirror, and then reflected back to the rotating mirror. The returning light was focused to a point image in a microscope. Due to the continued rotation of the mirror while the light was in transit from the rotating mirror, to the fixed mirror, and back, the beam was reflected into the microscope at an altered angle, resulting in a displacement of the point image. Measurement of this displacement, along with distance parameters from our setup, known as the Foucault Method, allows for the speed of the light to be calculated. The mean speed of light calculated from this experiment was (2.9972 ± 0.0139) x 108 m/s.

    See PDF for full text.

    Attached Files:

  2. jcsd
  3. Dec 2, 2012 #2


    User Avatar
    Gold Member

    Hi, I did not check the details but
    looks weird to me. I'd write A=(0.2600 ± 0.0005) m, etc.
    In the table
    I'd write 2.84 ± 0.04, etc.
    Also in table 3 I get suspicious because the lesser is the calculated c value, the lesser the error you get. Using my intuition the opposite should happen.
    Maybe I was not taught the right way though and keep in mind that I'm just an undergraduate student.
  4. Dec 2, 2012 #3

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    I am not sure what your question is. Looking at the attached document, though, and I conclude that you have underestimated your uncertainties by a rather large factor: perhaps two or three.

    Also, Table 3 has a serious problem with significant figures.
  5. Dec 3, 2012 #4


    User Avatar
    Science Advisor

    If that is your estimate of the error, how do you explains statistical error being (2.9972 ± 0.1071) x 108 m/s?

    You should always check the statistical error on your data. It should never be higher than the error you claim. You have made a mistake somewhere by a whole order of magnitude. It doesn't affect your answer, but it affects how believable your answer is. If you claim an error of less than 0.5%, you need to show statistical variation of less than 0.5%. It's not. The fact that you didn't get 3.1 for an answer is entirely down to luck. Unless it's not even that. Did you keep all of the data points you took, or only the ten you liked?

    Just to illustrate this a little better I've attached a plot of your data. Red dots are the data from the table. Pale blue band is the probability distribution according to 2.9972 ± 0.0139 you claim, with solid blue dashes marking off the area of the band that should contain 90% of your data points according to that error. Doesn't look good, for your claim, does it?

    Attached Files:

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook