I didn't believe that light slows down in water (part 1)

Supported by Screen Australia and Youtube through the Skip Ahead initiative.

Experiment:

Code:

References:

The Feynman lectures- “Ch 31: On the origin of refractive index” and “Ch 48: Beats”

Matter and Interactions 3rd Edition - The quote is from section 24-4 on page 1001. It’s a long passage but I tried to paraphrase it accurately. Here’s the full quote:

How might we measure the speed of propagation of an electromagnetic wave?
One can think of two different approaches:

(a) Follow a wave crest: If you watch one particular wave crest, you will see
that it travels a distance λ (one wavelength) in a time T (one period).
Therefore the speed of the crest is

v = λ / T

(b) Time the arrival of a radiative electric field: One could imagine a different way of measuring the speed of an electromagnetic wave. Suppose that you and a friend synchronize your clocks, then travel to locations that are a distance d apart. Your friend aims a laser at your location, and precisely at time t1, turns on the laser. You record the time t2 at which you first detect the radiative electric field. In the laser light, and knowing the distance between the locations and the elapsed time Δt = t2 - t1, you calculate the speed at which the laser light traveled toward you:

v = d / Δt

In a vacuum, these two ways of measuring the speed of a sinusoidal electromagnetic wave will give the same answer: 3 x 10^8 m/s. However, this will not necessarily be the case if part or all of the space through which the light wave travels is filled with a medium such as water, glass, or even air. In this case, method 2 (measuring the time required for information about a change in the electromagnetic field to travel a given distance) will still give 3 x 10^8 m/s. However, method 1 (timing the interval between crests in a steady state electromagnetic wave inside the medium) will give a different answer, which will almost always be less than 3 x 10^8 m/s.

[ModelThree]

I rarely comment on videos, but for this one, I feel I must. As someone who teaches physics and continues to study it in my free time, I feel like your video is such an honest and amazing look into the scientific method. You found something you were curious about, did research, came up with a very well argued hypothesis, tested your hypothesis experimentally, and found the result that you didn't expect, which you finally accepted. You should be really proud - I'm definitely going to be showing this video to my students every year.

[rosuav]

"Light's whole purpose is to harass charged particles." That means that playing with a laser pointer and a cat is fundamentally correct: a cat is just a really really BIG charged particle.

[SytRReD]

I've been following you for about the last ten years, and although you don't post often, your videos are of utmost quality, and I love them! So glad to have this tonight, plus it's a question that bugged me for years so I look forward to your discoveries and explanations!

[IBUCme]

I remember measuring the speed of light in my college lab 52 years ago. I am so blown away with how a person today can challenge physical concepts with commonly available devices.

[KitagumaIgen]

As an researcher with a experiment-observation job it pleases me to see a theoretician struggle with experimental problems - this is not just a little embarrassing but a lot embarrassing to admit. But the most important thing to remember: It is when we find out that we're wrong about something we have the opportunity to learn something new! Great job!

[pineapplegodguy]

just wow. the journey was crazy, and we've got a mouthful of emotions - somewhat of a recurrent theme of scientific research (and of your channel) is the expectations, the joy in anticipation of the results, the moment reality brings you back to earth with the experimental results. and we've got all of that filmed. incredible.

[INTO]

Some high frequency traders made a lot of money by building a radio connection between stock exchanges, since it was faster than the existing fiber optics connection (speed of light reduced to 2/3 compared to air).

[elinope]

3Blue1Brown recommended your channel to me, subscribed. That's a damned good recommendation, and you lived up to it.

[ghahrai]

I am so glad I found you. I will follow your enthusiasm from now on. Thank you

[Niohimself]

Please make more science videos like this one, this is amazing!

[peetiegonzalez]

I can't get get over how cute those anthropomorphized electrons are.
Excellent pair of videos, here. As usual. I haven't seen you in my recommended feed for quite a while so I'm very glad Grant pointed me in this direction!

[astrokevin]

I really loved this whole approach of thinking critically about the popular explanations, consulting multiple good-quality sources, putting your own understanding to the test, and even sharing when your experiment shows your understanding to be wrong. An excellent and honest demonstration of how science is supposed to work.

[dsteele]

I did a PhD in physics many years ago and thats a really excellent video you made. You demonstrate the scientific method, repeatedly testing theory against experiment (reality testing), revising your theories, whilst recognising the complexity of experiments that are not as simple as they might seem. The scientific method is so important because its the only way to bootstrap new knowledge (avoiding theorising in the absence of experiment, which is very tempting as it avoids the hard work of experiment, which is philosophy).

[kavehguilanpour]

Your videos are amazing. Please don’t stop making them. I started out as a physics undergrad years ago, but switched because I did not think my maths was good enough. Your videos have inspired me to start studying physics again for fun. From this video the things that blows my mind the most is the notion of a maximum speed of causality, and your example that the Earth would not ‘know’ for 8 mins if the sun disappeared. It completely upended my notion of the nature of the ‘force’ connecting the sun and the Earth.

[duggydo]

I like the way you made this video. Taking us through the whole process rather than editing it to show just the outcome is very genuine and I can appreciate that.

[muffinstots]

I just wanted to take a minute to say I absolutely love how this was presented. A personal journey of discovery that I felt like I was making with you. I'm glad you didn't shy away from showing your failures as it really helped me to understand the topic much better. Well done!

[nataliem]

this is some seriously inspiring science, it takes a lot of courage to post something like this! Looking forward to the next video!

[RealNovgorod]

The "original" wave does indeed travel at c through the medium. However, what comes out at the other side is NOT the original wave but a superposition of it with the (non-resonantly) stimulated dipole emitters in the medium. Every infinitesimal "layer" of the medium adds a phase delay to the emitted wave (because of the molecular harmonic potential, but that's not important at this point), and this delayed wave interferes with the original wave and shifts its phase by a little bit - but the resulting wave still travels at c. The thing to realize is that this tiny phase delay happens continuously at every single layer of the medium, so you're continuously adding a bit of delay to the original wave at some constant _rate_ which is the same as changing the (spatial) frequency. That explains the change in wavelength inside a medium in the 3b1b video.

In your case, you're dealing with short pulses of light and the TOF LIDAR measures when the peak of the pulse arrives back at the detector. This peak is slowed down inside the medium due to the same superposition mechanism with continuously phase-delayed versions of itself so it takes longer to pass through the medium. That's the definition of group velocity. What stays constant is the _phase velocity_ which is the oscillation speed of the E-field _underneath_ the pulse envelope. In a simulation you can see that the wave peaks underneath the envelope travel faster (i.e., at c) than the envelope itself (at c/n), but because of the interference you never detect these c-speed wave peaks at the output after the medium because they are destructively suppressed by the pulse envelope effectively "riding the wave backwards" (from the reference frame of the input wave). There's also broadening of the pulse due to group velocity dispersion in most media, but that's a story for another day...

[ArbitraryConstant]

The speed of light being slower in fiber optics is pretty important to understanding network latency, I assume similar principles apply there.

[kxs]

Really happy to have found this channel. Appreciate hearing the thought process, the hand-crafted animations, the honest reporting - and the fact you're adressing a topic that's niggled at the back of my mind.

Looks like the argument still continues in the comments section here...!