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

I White light thru prism: Why aren't separated colors actually mixed?

  1. Jul 12, 2016 #1
    Ok, here's the illustration of the experiment: http://media.web.britannica.com/eb-media/10/7710-050-36C066AC.jpg

    But if you trace the light path back from the prism and into the white light, it seems as though the white light is actually made up of STACKED colors of the rainbow, not a MIXTURE of the colors.

    So here is white light made of stacked colors: http://i.imgur.com/8PHHmoI.png

    Here is white light made of mixed colors (still has colors of rainbow, just not in order): http://i.imgur.com/3n6zQaa.png

    And here is the hypothesized mixed white light after coming out of prism: http://i.imgur.com/Azhc3Ge.png

    But if different colors refract at different angles, and the same colors refract at the same angles as one another, then that means the last picture would be observed. But since we don't observe that in real life, then white light is actually an ordered stack of colors?
     
  2. jcsd
  3. Jul 12, 2016 #2

    berkeman

    User Avatar

    Staff: Mentor

    The first image is the closest to real life. The other images are just silly.

    Use Google Images to see real pictures of a prism separating white light into the ROYGBIV rainbow of colors.
     
  4. Jul 12, 2016 #3

    berkeman

    User Avatar

    Staff: Mentor

  5. Jul 12, 2016 #4
    Sure, but the question still has to be answered: is white light a mixture of the rainbow or a stack of the rainbow?
     
  6. Jul 12, 2016 #5

    berkeman

    User Avatar

    Staff: Mentor

    Mixture. The stack thing seems quite erroneous. Where did you read that term?
     
  7. Jul 12, 2016 #6
    I made it up to describe what I was thinking about :D

    Also, I think I know my misunderstanding now. The rays I drew in the last picture, if extended, will result in light separation. I just didn't draw the rays long enough. So, then, I guess this means this light prism experiment doesn't show the rainbow below a certain distance because there wont be enough distance for the light rays to separate?
     
  8. Jul 12, 2016 #7

    Drakkith

    User Avatar

    Staff: Mentor

    Your last picture is incorrect. That is not the order of the colors when they emerge from the prism. Berk's picture shows this immediately.

    How so? If I trace the colors back I see a single beam of white light and no evidence of whatever this "stack" is supposed to be.

    Are you actually tracing light back through a real prism, or just doing a thought experiment?
     
  9. Jul 13, 2016 #8

    Merlin3189

    User Avatar
    Gold Member

    I think you are being mislead by the original Brittanica picture, which is wrong. They show the beam being split completely at the first face of the prism - as if the beam were as you conjecture - which is wrong.
    (Edit: I've looked for some better pictures, but most fall into the same error. I think it is very hard to draw the correct picture, either by hand or with most graphics software.)
    Say the original beam were 1mm wide, then the red light would come from the whole 1mm width and continue as a 1mm red beam producing 1mm wide band of red light on the screen.
    Equally the blue light would come from the whole 1mm beam width in a 1mm beam all the way to the screen and produce a 1mm wide band of blue light.
    Similarly for all the other colours - N.B. not just the nominal 7, but the infinite number of colour /wavelength gradations. Each produces its 1mm band on the screen, overlapping and mixing with the adjacent colours.
    This effect can be reduced by making the initial beam narrower, which dims the spectrum or by using lenses to focus a divergent beam from a narrow slit, which keeps it brighter.
    This is shown in SchoolPhysics . Their diagram shows only two representative rays at the sides of the beam, but you can see that both red and blue light (and all other colours) come from all parts of the beam.
    spectrum.png

    Edit: Maybe a better diagram of the same thing at Telecopes in OpenEd
    spectrum2.jpg
    The "polychromatic beam" is the equivalent of the white beam, but what they are looking at may not be white. (If you try this experiment, it might be interesting to use an LED or flourescent lamp as a "white" source. Their spectra are noticeably different from that of sunlight or incandescent light, even with simple equipment. &BTW, if you want to experiment and haven't got a suitable prism, a CD will do an adequate job as a diffraction grating.)
     
    Last edited: Jul 13, 2016
  10. Jul 13, 2016 #9

    Drakkith

    User Avatar

    Staff: Mentor

    Indeed. It's important to understand that the light being sent into the prism in most pictures is a thin beam or column. Without making the light into a beam or column you cannot see the dispersion. If you have a prism, cutting a thin rectangular slit in a piece of thick paper or cardboard should work just fine.
     
  11. Jul 13, 2016 #10
    Just a thought experiment based on the Brittanica picture.
    I see. So all the colors would be just as wide as the original white beam of light. Thanks everyone!
     
  12. Jul 14, 2016 #11

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Experimental limits will always cause spreading of any particular spectral line. Original beam width, chromatic aberration and diffraction will always give uncertainly of the wavelength measured in a spectrometer or spectroscope.
    It is precisely the same thing that happens with frequency measurement of other EM waves (or other types of oscillation). The analyser has a finite bandwidth so it will record the presence of frequencies on either side of the nominal measured frequency.
    The original naming of the spectral colours really didn't help either. A total arbitrary exercise, based on Numerology!
     
  13. Jul 14, 2016 #12

    jtbell

    User Avatar

    Staff: Mentor

    Except at the edges of the beam.

    [added] Oops, I now realize that Drak was assuming a beam in the first place. However, even without a narrow beam, you can still see dispersion effects at the edge of a broad "beam" e.g. when looking through a prism at some object. Or through a lens that is poorly corrected for dispersion, as in a cheap camera.

    I have to wear strong eyeglasses. I once had a pair that used glass with a high index of refraction to make the lenses thinner. That particular glass also had a lot of dispersion, so objects toward the edges of my field of view had nice colored fringes.
     
    Last edited: Jul 14, 2016
  14. Jul 14, 2016 #13

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    As ever, there are other issues. Making the slit narrower will restrict the amount of light getting through. So the pure spectrum display can get too dim to see!
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: White light thru prism: Why aren't separated colors actually mixed?
Loading...