# Speed of water and relation to light

Fill a bowl with water, lay a piece of fabrics over the bowl, one end touching the water and the other end laying down on the table. Due to the capillarity effect the water will follow the fabrics and after a while it will be all wet. From this little experiment we can deduce that water travels, at a speed that depends of the type of fabrics, and has a maximum speed in each type of fabrics. We can even extrapolate that the speed of water in nylon can be something like 1 cm/minute. Silly isn't it? But its apparent speed is dictated by the medium that carries it.

That brings my questions related to light.
1. Could it be possible that light has an infinite speed but it is slowed down by the medium where it travels? Isn't there a vacuum permeability in the vacuum of space that would limit its speed?
2. (my favorite option) Is it possible that light doesn't move at all? Like the water in the previous analogy? That it just dissipates and follow the fabrics of space, even empty ones, as water in nylon? That would supply a different explanation of why light slow down in water and speed up when it get out.

Hi Serge, welcome to physicsforums!

The generally accepted model for light propagation in relativity is a wave model. However there are variations in the interpretations, as there are problems with details; in practice what people do is know how to calculate based on rules from wave mechanics. In that context it's easier to point out how your suggestions conflict or agree with currently used models:

- the speed of light in vacuum is a local constant determined by the vacuum and when light travels through water, it is effectively delayed due to interactions with the water molecules. See:

- concepts as "vacuum" and "fabric of space" can be used with relativity theory.

However, this forum is not meant for discussing personal possible new theories (see Rules at top of the page). As you can see in the linked threads, refraction is very well explained by current scattering theory.

Last edited:
mfb
Mentor
1. Could it be possible that light has an infinite speed but it is slowed down by the medium where it travels? Isn't there a vacuum permeability in the vacuum of space that would limit its speed?
It has a fixed, constant speed in vacuum, for all observers. This is not possible if the vacuum would be some medium and light would travel relative to that medium.

2. (my favorite option) Is it possible that light doesn't move at all?
It can travel from A to B, that is usually called "movement".

Like the water in the previous analogy?
Water moves in your (wrong) analogy.

It has a fixed, constant speed in vacuum, for all observers. This is not possible if the vacuum would be some medium and light would travel relative to that medium.[..].
In fact, it has been known since 1904/1905 that this is possible and some of the early relativists related to such a wave model*. The problems with specific wave models do not depend on relativity. To mention a basic problem: As far as I know, a spherical transverse light wave emitted from a light source is an impossibility.

*

Last edited:
Dale
Mentor
2021 Award
To mention a basic problem: As far as I know, a spherical transverse light wave emitted from a light source is an impossibility.
Why is that a problem? That is just another way of saying there is no monopolar EM radiation. Why is it a problem that there is no monopolar EM radiation?

Last edited:
mfb
Mentor
In fact, it has been known since 1904/1905 that this is possible and some of the early relativists related to such a wave model*. The problems with specific wave models do not depend on relativity. To mention a basic problem: As far as I know, a spherical transverse light wave emitted from a light source is an impossibility.
You can construct some co-moving ethers and whatever, but then you give up the concept of a preferred reference frame, and get the same predictions as SR (maybe just with a more complicated framework). Otherwise, you get predictions different from SR, which are inconsistent with experiments.

You can[..] give up the concept of a preferred reference frame, and get the same predictions as SR [..]
That's right; I simply corrected an erroneous claim. However, we were next both inaccurate. When I wrote that "the problems with specific wave models do not depend on relativity", I was only thinking of detailed mechanisms and not about what you call, "some co-moving ethers and whatever". While a Lorentz ether gives the same predictions, it is held that pluralistic "co-moving ethers" can't work as a model for SR (the prime argument was stellar aberration).

Why is that a problem? That is just another way of saying there is no monopolar EM radiation. Why is it a problem that there is no monopolar EM radiation?
I think that you mean something else. I merely referred to an idealized classical spherical uniform light source that supposedly emits transverse spherical waves around it. Did you ever try to draw that in 3D perspective? I did and if I'm not mistaken, then that is a geometric impossibility. However, if I am mistaken, then I'll be delighted to find out how it can be done! :tongue2:

Dale
Mentor
2021 Award
It is impossible, I just dont know why that is a problem according to you.

I wish to reformulate my question. Why is there a maximum speed of light in vacuum? is it because it won't go faster or because it can't?

mfb
Mentor
Light does not have a will, or "tries" to do something.
It has a specific speed (that is an observation), and you can derive this based on the Maxwell equations, for example.
The laws of physics are the same in all reference frames (another observation), this makes sure that the speed is the same for all frequencies and all directions.

I wish to reformulate my question. Why is there a maximum speed of light in vacuum? is it because it won't go faster or because it can't?
According to Maxwell's theory on which SR was based, it's a property of vacuum, so that light can only propagate at the speed of light - similar to sound which propagates at the speed of sound.
See: http://en.wikipedia.org/wiki/Vacuum_permittivity

It is impossible, I just dont know why that is a problem according to you.
Impossible means that such a basic classical model is plain wrong - simply due to self contradiction, without any need to introduce QM. I'm not aware of a more severe form of "problematic".

Dale
Mentor
2021 Award
Impossible means that such a basic classical model is plain wrong - simply due to self contradiction, without any need to introduce QM. I'm not aware of a more severe form of "problematic".
It is only a self contradiction if the model predicts it and it is impossible, but Maxwells equations do not predict a spherical transverse wave. The lowest order of radiation allowed by Maxwells equations is dipole.

Since it is not predicted the fact that it is impossible is not a problem.

According to Maxwell's theory on which SR was based, it's a property of vacuum, so that light can only propagate at the speed of light - similar to sound which propagates at the speed of sound.
See: http://en.wikipedia.org/wiki/Vacuum_permittivity

So i can suppose that if light doesn't travel faster, it is not because it won't, but because it can't. It is possible to assume that light would go faster if it wasn't of the permittivity of vacuum?

Dale
Mentor
2021 Award
So i can suppose that if light doesn't travel faster, it is not because it won't, but because it can't.
Can you think of an experiment which could distinguish between "it won't" and "it can't"? If not, then the concept is philosophy, not science, and doesn't belong here. If so, then please describe the experiment and we can analyze what SR would predict.

It is only a self contradiction if the model predicts it and it is impossible, but Maxwells equations do not predict a spherical transverse wave. The lowest order of radiation allowed by Maxwells equations is dipole. [..]
Good point! Assuming that that's correct, I'm afraid that this is not sufficiently recognized. A quick internet search produces on the one hand claims in courses such as
"light consists of [..] fields that travel through space as transverse waves", e.g.
http://ubpheno.physics.buffalo.edu/~dow/lectures/phy102/ch24_print.pdf ,
but also claims like "Spherical EM wave" and "The E.M. wave has a spherical wave front", e.g. both on slide 7 in http://www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px384/lecture_05.pdf
and not to forget the typical SR computation based on "consider a pulse of light that [..] propagates as a spherical wave" such as in
http://www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px384/lecture_05.pdf
and of course "let a spherical wave be emitted" in
http://www.fourmilab.ch/etexts/einstein/specrel/www/

[..] It is possible to assume that light would go faster if it wasn't of the permittivity of vacuum?
If vacuum permittivity and permeability are reduced, then also c is increased. In fact, as measured with a reference system on Earth, the speed of light is slightly reduced near the sun. That was predicted by GR and confirmed by experiments.
I found nice and clear explanations plus animations here:
- http://www.astro.ucla.edu/~wright/deflection-delay.html

Last edited:
Good point! Assuming that that's correct, I'm afraid that this is not sufficiently recognized. A quick internet search produces on the one hand claims in courses such as
"light consists of [..] fields that travel through space as transverse waves", e.g.
http://ubpheno.physics.buffalo.edu/~dow/lectures/phy102/ch24_print.pdf ,
but also claims like "Spherical EM wave" and "The E.M. wave has a spherical wave front", e.g. both on slide 7 in http://www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px384/lecture_05.pdf
and not to forget the typical SR computation based on "consider a pulse of light that [..] propagates as a spherical wave" such as in
http://www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px384/lecture_05.pdf
and of course "let a spherical wave be emitted" in
http://www.fourmilab.ch/etexts/einstein/specrel/www/

But aren't they actually talking about aggregate collections of photons isotropically emitted from a symmetric source. Not a single wave in any real sense.

But aren't they actually talking about aggregate collections of photons isotropically emitted from a symmetric source. Not a single wave in any real sense.
It looks to me that they are not talking about photons but about spherical waves just as they state; for sure Maxwell wasn't thinking of photons and neither was Einstein implying photons in his SR paper.

Last edited:
Good point! Assuming that that's correct, I'm afraid that this is not sufficiently recognized. A quick internet search produces on the one hand claims in courses such as
"light consists of [..] fields that travel through space as transverse waves", e.g.
http://ubpheno.physics.buffalo.edu/~dow/lectures/phy102/ch24_print.pdf ,
but also claims like "Spherical EM wave" and "The E.M. wave has a spherical wave front", e.g. both on slide 7 in http://www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px384/lecture_05.pdf
and not to forget the typical SR computation based on "consider a pulse of light that [..] propagates as a spherical wave" such as in
http://www2.warwick.ac.uk/fac/sci/physics/current/teach/module_home/px384/lecture_05.pdf
and of course "let a spherical wave be emitted" in
http://www.fourmilab.ch/etexts/einstein/specrel/www/

Harald - [..] You argue a lot about validity of Maxwell's eqn's in relation to SR [..]
Hi Q-reeus, I did not argue at all "about validity of Maxwell's eqn's in relation to SR" - quite the contrary, I mentioned that SR is based on it!
I mentioned in relation to the OP that we don't really understand what light exactly is but that for sure certain wave models can't be correct, such as spherical transverse waves which appear to be an impossibility. Next it was Dalespam who brought up Maxwell's equations by saying that those do not allow spherical transverse waves and I found that valuable input.
use any of the many good online resource [..]
Regretfully online university courses are supposed to be good online resources. Now, you seem to merely disagree with us about words and not about content:
Spherical simply means that the constant-phase wavefront from say a dipole radiator is spherical at large r from such a source, not that the amplitude of that wave is spherically symmetric - an impossibility.[..]
In my book a wavefront designates a constant phase, and I find two poles where this is impossible, so that the amplitude at those poles must be zero. Zero amplitude still implies no wavefront and no phase at those points. Are you really arguing about words? :uhh:

Last edited:
Hi Q-reeus, I did not argue at all "about validity of Maxwell's eqn's in relation to SR" - quite the contrary, I mentioned that SR is based on it!
I mentioned in relation to the OP that we don't really understand what light exactly is but that for sure certain wave models can't be correct, such as spherical transverse waves which IMHO are an impossibility. Next it was Dalespam who brought up Maxwell's equations by saying that those do not allow spherical transverse waves and I found that valuable input.

Now, you seem to merely disagree with us about words and not about content:
Sorry, I should have been a little more specific. Didn't mean you were arguing with or against compatibility of ME's and SR. Rather that you were arguing for their compatibility, but then getting into issues - e.g. of impossibility of spherically symmetric radiation from a single oscillator as somehow a problem for classical EM/ME's.
In my book a wavefront designates a constant phase, and I find two poles where this is impossible, so that the amplitude at those poles must be zero. Zero amplitude still implies no wavefront and no phase at those points. Are you really arguing about words? :uhh:
Well it gets down to agreed definitions I suppose. It is simply an agreed and universally adopted convention that 'spherical wave-front' in respect of radiation refers to just that - the phase. Mathematically, although the amplitude drops to zero at the poles of a dipole oscillator far-field, we still have that the phase is defined. Otherwise we have some kind of axial 'singularity crisis' is implied. Anyway, keep happy I say.

Sorry, I should have been a little more specific. Didn't mean you were arguing with or against compatibility of ME's and SR. Rather that you were arguing for their compatibility, but then getting into issues - e.g. of impossibility of spherically symmetric radiation from a single oscillator as somehow a problem for classical EM/ME's.
I don't think that I mixed up Maxwell's equations with descriptions of wavefronts in some of the literature; however it was Dalespam who stressed their incompatibility.
Well it gets down to agreed definitions I suppose. It is simply an agreed and universally adopted convention that 'spherical wave-front' in respect of radiation refers to just that - the phase. Mathematically, although the amplitude drops to zero at the poles of a dipole oscillator far-field, we still have that the phase is defined. Otherwise we have some kind of axial 'singularity crisis' is implied. Anyway, keep happy I say.
That was exactly my point, which means that this is probably not about definitions. Thus I repeat:
Did you ever try to draw a [edit: transverse] spherical wavefront (thus with defined phase everywhere) in 3D perspective? I did and if I'm not mistaken, then that is a geometric impossibility. However, if I am mistaken, then I'll be delighted to find out how it can be done! :tongue2:

Last edited:
I don't think that I mixed up Maxwell's equations with descriptions of wavefronts in some of the literature; however it was Dalespam who stressed their incompatibility.

That was exactly my point, which means that this is probably not about definitions. Thus I repeat:
Did you ever try to draw a spherical wavefront (thus with defined phase everywhere) in 3D perspective? I did and if I'm not mistaken, then that is a geometric impossibility. However, if I am mistaken, then I'll be delighted to find out how it can be done! :tongue2:
Odds are you will have visited and studied the Wiki site here In which case you have noted that even when θ and thus the sin(θ) part goes to zero, phase is a well defined function of radius for a dipole oscillator far-field. It all reminds me of a web site I came across where the individual there was hinging his/her entire case on that the field line of a bar-magnet that points exactly north or south 'never ends' - thus soundly 'refuting' the usually accepted position that lines of B always form closed loops. But I'm quietly confident you wouldn't subscribe to such an argument.