A New Limit on Photon Mass

50 views
Skip to first unread message

physics

unread,
Aug 14, 2010, 9:05:16 PM8/14/10
to
A New Limit on Photon Mass
A new limit on photon mass, less than 10-51 grams or 7 x 10-19
electron volts, has been established by an experiment in which light
is aimed at a sensitive torsion balance; if light had mass, the
rotating balance would suffer an additional tiny torque. This
represents a 20-fold improvement over previous limits on photon mass.
Photon mass is expected to be zero by most physicists, but this is an
assumption which must be checked experimentally. A nonzero mass would
make trouble for special relativity, Maxwell's equations, and for
Coulomb's inverse-square law for electrical attraction.
The work was carried out by Jun Luo and his colleagues at Huazhong
University of Science and Technology in Wuhan, China
(jun...@mail.hust.edu.cn, 86-27-8755-6653). They have also carried out
a measurement of the universal gravitational constant G (Luo et al.,
Physical Review D, 15 February 1999) and are currently measuring the
force of gravity at the sub-millimeter range (a departure from
Newton's inverse-square law might suggest the existence of extra
spatial dimensions) and are studying the Casimir force, a quantum
effect in which nearby parallel plates are drawn together. (Luo et
al., Physical Review Letters, 28 February 2003)

Arnold Neumaier

unread,
Aug 15, 2010, 1:20:52 PM8/15/10
to
physics wrote:
> A New Limit on Photon Mass
> A new limit on photon mass, less than 10-51 grams or 7 x 10-19
> electron volts, has been established by an experiment in which light
> is aimed at a sensitive torsion balance; if light had mass, the
> rotating balance would suffer an additional tiny torque. This
> represents a 20-fold improvement over previous limits on photon mass.
> Photon mass is expected to be zero by most physicists, but this is an
> assumption which must be checked experimentally. A nonzero mass would
> make trouble for special relativity,

Why for relativity?

QED with massive photons is still renormalizable and relativistic.


Arnold Neumaier

jw...@basicisp.net

unread,
Aug 15, 2010, 1:20:53 PM8/15/10
to
On Aug 14, 6:05 pm, physics <phy...@gmail.com> wrote:
> A New Limit on Photon Mass
> ...

> Photon mass is expected to be zero by most physicists, but this is an
> assumption which must be checked experimentally. A nonzero mass would
> make trouble for special relativity, Maxwell's equations, and for
> Coulomb's inverse-square law for electrical attraction.
> ...

This kind of experiment is a waste of time except as a way of training
the
experimenters in making careful measurements.

Special relativity is not in trouble; it is an observed fact that the
speed of
light in vacuum does not vary with reference frame. This could not
possibly be so if
photons had a nonzero mass or a proper time. The mass of a photon
must be exactly
zero; any rationale which postulates a transition region, a
continuity, from
zero to minute but nonzero mass is math fiction, not science.

One might as well remeasure the mass of the electron, to see whether
it was zero or not.
Think of the trouble that would cause for electronics!

Arnold Neumaier

unread,
Aug 15, 2010, 1:49:05 PM8/15/10
to
jw...@BasicISP.net wrote:
> On Aug 14, 6:05 pm, physics <phy...@gmail.com> wrote:
>> A New Limit on Photon Mass
>> ...
>> Photon mass is expected to be zero by most physicists, but this is an
>> assumption which must be checked experimentally. A nonzero mass would
>> make trouble for special relativity, Maxwell's equations, and for
>> Coulomb's inverse-square law for electrical attraction.
>> ...
>
> This kind of experiment is a waste of time except as a way of training
> the
> experimenters in making careful measurements.
>
> Special relativity is not in trouble; it is an observed fact that the
> speed of
> light in vacuum does not vary with reference frame.

This is demonstable only within a certain accuracy.


> This could not possibly be so if
> photons had a nonzero mass or a proper time.

If photons have mass, the relativity principle simply reads:

''There is a finite limit speed for particles, which is achieved if
the particle is massless, and which is asymptotically approached as
a massive particle acquires more and more momentum.''

Only this is required for a consistent standard model + relativitry.


Arnold Neumaier

Oliver Jennrich

unread,
Aug 15, 2010, 4:30:27 PM8/15/10
to
"jw...@BasicISP.net" <jw...@BasicISP.net> writes:

> On Aug 14, 6:05 pm, physics <phy...@gmail.com> wrote:
>> A New Limit on Photon Mass
>> ...
>> Photon mass is expected to be zero by most physicists, but this is an
>> assumption which must be checked experimentally. A nonzero mass would
>> make trouble for special relativity, Maxwell's equations, and for
>> Coulomb's inverse-square law for electrical attraction.
>> ...
>
> This kind of experiment is a waste of time except as a way of training
> the experimenters in making careful measurements.

It does that, for sure. Were there only coresponding activities that
trained the theoretians in making careful claims.

> Special relativity is not in trouble; it is an observed fact that the
> speed of light in vacuum does not vary with reference frame.

And those are the observations that give the current (or previous)
limits for the photon mass. How well do you think is the constancy of
the speed of light established?

> This could not possibly be so if photons had a nonzero mass or a
> proper time. The mass of a photon must be exactly zero; any rationale
> which postulates a transition region, a continuity, from zero to
> minute but nonzero mass is math fiction, not science.

Here is a training activity as alleged above: Devise an experiment (can
be a Gedankenexperiment, but real experiments are preferred) that allows
to distinguish through a measurement a photon mass of 1e-100 eV from 0
eV.

--
Space - The final frontier

eric gisse

unread,
Aug 16, 2010, 5:03:33 AM8/16/10
to
physics wrote:
[...]

Relativity would not care, as it is not a theory of light.

All that would happen is Maxwell --> Proca, and a decoupling between the
speed of light and the maximum speed on a Lorentzian manifold.

Joseph Warner

unread,
Aug 16, 2010, 9:19:01 AM8/16/10
to
<jw...@BasicISP.net> wrote in message
news:f6ee99d6-24dc-40ae...@t2g2000yqe.googlegroups.com...

> On Aug 14, 6:05 pm, physics <phy...@gmail.com> wrote:
>> A New Limit on Photon Mass
>> ...
>> Photon mass is expected to be zero by most physicists, but
>> this is an
>> assumption which must be checked experimentally. A nonzero
>> mass would
>> make trouble for special relativity, Maxwell's equations, and
>> for
>> Coulomb's inverse-square law for electrical attraction.
>> ...
>
> This kind of experiment is a waste of time except as a way of
> training
> the
> experimenters in making careful measurements.
>

No, it isn't. All postulation must be checked to the be method
of measurements. Only if the basis of physics is confirmed within
experimental limits can the theories based on those postulations
can be trusted. The more precise and the more accurate we can
measure fundamental physical constants the more confidence we
have in our explanation of world and universe.

This experiment is in the same league as measuring the decay rate
of a proton or the change in time of the gravitational constant.

Tom Roberts

unread,
Aug 16, 2010, 11:41:13 AM8/16/10
to
jw...@BasicISP.net wrote:
> On Aug 14, 6:05 pm, physics <phy...@gmail.com> wrote:
>> A New Limit on Photon Mass ... Photon mass is expected to be zero by most
>> physicists, but this is an assumption which must be checked experimentally.
>> A nonzero mass would make trouble for special relativity, Maxwell's
>> equations, and for Coulomb's inverse-square law for electrical attraction.
>> ...
>
> This kind of experiment is a waste of time except as a way of training the
> experimenters in making careful measurements.

No. It is always appropriate to test physical theories to improved accuracy. One
never knows where surprises might turn up (see neutrinos below).


> Special relativity is not in trouble; it is an observed fact that the speed
> of light in vacuum does not vary with reference frame. This could not
> possibly be so if photons had a nonzero mass or a proper time.

Observations of the constancy of the speed of light have a limit in their
accuracy, and that is ENORMOUSLY less accurate than these more direct
measurements of the photon mass. I don't know exactly, but in terms of a limit
on the photon mass it must be many billions of times greater (c.f. neutrinos below).


> The mass of
> a photon must be exactly zero; any rationale which postulates a transition
> region, a continuity, from zero to minute but nonzero mass is math fiction,
> not science.

Not true. There are viable physical theories in which the photon has a nonzero mass.

Sanity check: just a few years ago it was generally accepted that neutrinos had
zero mass. Today from oscillation measurements it is known that at least two of
them have nonzero mass, and it is suspected that all of them do. Note the lower
limits on neutrino masses are ~10^15 times larger than the limit on the photon
mass, yet neutrinos and light moved with the same speed to within ~10 parts per
billion from SN1987A -- that is comparable to the accuracy with which the speed
of light was known prior to the redefinition in 1983.


> One might as well remeasure the mass of the electron, to see whether it was
> zero or not. Think of the trouble that would cause for electronics!

This is not a valid analogy.


As I said before, Special Relativity would not be affected by a non-zero photon
mass, as Einstein's second postulate is not required in a modern derivation
(using group theory one obtains three related theories, two of which are solidly
refuted experimentally and the third is SR). So today's foundations of modern
physics would not be threatened.


Tom Roberts

jw...@basicisp.net

unread,
Aug 16, 2010, 12:06:52 PM8/16/10
to

==== Moderator's note ==================================================
The photon mass is determined to be 0 to a phantastic degree of
accuracy, but it's not absolutely known to be exactly 0. The upper
limit, according to the just appeared review of particle physics is
m<10^(-18) eV/c^2.
========================================================================

On Aug 16, 8:41 am, Tom Roberts <tjroberts...@sbcglobal.net> wrote:
> jw...@BasicISP.net wrote:
>

You are missing the point. The the photon mass HAS BEEN measured
with infinite precision already!

If a photon had ANY mass, it would by definition be associated
with an inertial rest frame. Therefore, the speed of propagation
of photons would vary with the velocity of the source, relative
to any other rest frame. The velocity of light would very obviously
add or subtract from the easily measured velocities of other objects,
such as satellites, planets, or stars.

But, the speed of light is known to be at least approximately constant
in all rest frames,
well within the margin of error which holds for objects moving
with various velocities. The mass can not be nonzero, whether
or not the speed of light is exactly the same in all frames.

harald

unread,
Aug 16, 2010, 1:15:30 PM8/16/10
to
On Aug 15, 7:20 pm, Arnold Neumaier <Arnold.Neuma...@univie.ac.at>
wrote:

Obviously that would invalidate the light postulate of SRT; but that
already happened with GRT. Such "trouble" is both real and small.

Harald

Oliver Jennrich

unread,
Aug 16, 2010, 3:07:07 PM8/16/10
to
"jw...@BasicISP.net" <jw...@BasicISP.net> writes:

> ==== Moderator's note ==================================================
> The photon mass is determined to be 0 to a phantastic degree of
> accuracy, but it's not absolutely known to be exactly 0. The upper
> limit, according to the just appeared review of particle physics is
> m<10^(-18) eV/c^2.
> ========================================================================
>
> On Aug 16, 8:41 am, Tom Roberts <tjroberts...@sbcglobal.net> wrote:
>> jw...@BasicISP.net wrote:
>>
>
> You are missing the point. The the photon mass HAS BEEN measured
> with infinite precision already!

If you could name the relevant publication, many of us would appreciate
it.

Oh No

unread,
Aug 17, 2010, 3:27:01 AM8/17/10
to
Thus spake "jw...@BasicISP.net" <jw...@BasicISP.net>

>
>==== Moderator's note ==================================================
>The photon mass is determined to be 0 to a phantastic degree of
>accuracy, but it's not absolutely known to be exactly 0. The upper
>limit, according to the just appeared review of particle physics is
>m<10^(-18) eV/c^2.
>========================================================================
>
>On Aug 16, 8:41 am, Tom Roberts <tjroberts...@sbcglobal.net> wrote:
>> jw...@BasicISP.net wrote:
>>
>
>You are missing the point. The the photon mass HAS BEEN measured
>with infinite precision already!
>
>If a photon had ANY mass, it would by definition be associated
>with an inertial rest frame.

This seems to be your misconception. It is true that Einstein originally
formulated sr wrt the speed of light. Since then it has been recognised
that sr should be formulated wrt a limiting speed, sometimes called the
maximal speed of information. When sr is formulated like this it still
holds but has no physical dependency on light and the photon does not
have to have zero mass.


Regards

--
Charles Francis
moderator sci.physics.foundations.
charles (dot) e (dot) h (dot) francis (at) googlemail.com (remove spaces and
braces)

http://www.rqgravity.net


J. J. Lodder

unread,
Aug 17, 2010, 9:56:24 PM8/17/10
to
physics <phy...@gmail.com> wrote:

> A New Limit on Photon Mass
> A new limit on photon mass, less than 10-51 grams or 7 x 10-19
> electron volts, has been established by an experiment in which light
> is aimed at a sensitive torsion balance; if light had mass, the
> rotating balance would suffer an additional tiny torque. This
> represents a 20-fold improvement over previous limits on photon mass.

The experiment has no value beyond a demonstration
of experimental skills.
It teaches us nothing new about the universe.

The limit corresponds (in more relevant units)
to a photon mass of about 2 mHz, or about 1 AU^-1

We already have far more stringent limits on the photon mass
from solar system physics,
(solar magnetopause, and interstellar magneto-hydrodynamics,
which tell us that the range of the EM force is far larger than 1 AU,
hence the photon mass is much smaller than that.

> Photon mass is expected to be zero by most physicists, but this is an
> assumption which must be checked experimentally.

Physists expect nothing of the kind.
They use the simplest adequate model,
which for the time being has zero photon mass.

> A nonzero mass would
> make trouble for special relativity, Maxwell's equations, and for
> Coulomb's inverse-square law for electrical attraction.

No problem for special relativity,
and we know how Maxwell's equations
and Coulombs law will be modified.

Jan

jw...@basicisp.net

unread,
Aug 17, 2010, 10:01:55 PM8/17/10
to
I'll try once more.

It has nothing to do with the exact speed of light, nor with
propagation of
information. We are not talking about information, but about light.

The infinite precision only depends upon the empirical fact that
the velocity of light does not add to the velocity of an object
sending or receiving light.

It also depends upon the assumption that light consists of photons.

If a photon had mass, then its velocity would depend on existence
of a rest frame for that photon. If such a rest frame existed,
then the velocity of light would add to that of the object at
its points of creating and annihilation. But this never is observed.

Surely, no one is questioning the idea of vector addition?

Therefore, photons have no rest frame, no proper time, nor
any mass.

The experiment described may indeed be valuable as training
for the researchers who carried it out, and it may be
valuable as a paradigm of precision, but it did not
reveal anythng new about the mass of the photon.

As for the information issue, one has to keep in mind
that the speed itself isn't the issue, but rather the mass
(or, value of the rest mass in some inertial frame).

Under reasonable assumptions, it eaasily can be proven
deductively tha all massless particles travel at the
same speed. If this is of interest, then you may
try reading http://www.siuc.edu/~pulfrich/Pulfrich_Pages/lit_nonp/phys_astro/2007_cSpeed/LimitingVelocity.html

Arnold Neumaier

unread,
Aug 18, 2010, 6:32:57 AM8/18/10
to
jw...@BasicISP.net wrote:
> I'll try once more.
>
> It has nothing to do with the exact speed of light, nor with
> propagation of
> information. We are not talking about information, but about light.
>
> The infinite precision only depends upon the empirical fact that
> the velocity of light does not add to the velocity of an object
> sending or receiving light.

As any empirical fact, this can be checked experimentally only up
to some tiny possible deviations.


> It also depends upon the assumption that light consists of photons.
>
> If a photon had mass, then its velocity would depend on existence
> of a rest frame for that photon. If such a rest frame existed,
> then the velocity of light would add to that of the object at
> its points of creating and annihilation. But this never is observed.

Observations are never exact, so sufficiently small deviations
(and hence a sufficiently small mass of the photon) cannot
be observed. The point of such experiments is precisely to
imporve the limit on how small these deviations must be.

If actual deviations would be found by an accurate enough
experiment, it would make a difference to physics.


> Surely, no one is questioning the idea of vector addition?

The question is how accurately Nature respects this vector addition.


> Therefore, photons have no rest frame, no proper time, nor
> any mass.

This only holds if you _define_ photons to be massless particles.
The question then would become whether these theoretical photons
describe Nature exactly. But the testing problem would remain the same.


> Under reasonable assumptions, it eaasily can be proven

> deductively that all massless particles travel at the
> same speed.

Yes, and currently, this speed is called the speed of light.
Would the photon be found to have a slight mass, this speed would
simply be renamed to ''limit speed'' or something like that,
photons could come to rest, but not much else would change.

Hendrik van Hees

unread,
Aug 18, 2010, 6:47:22 AM8/18/10
to
There is one more difference between a massless and a massive vector field,
and that's the fact that a massive field has three physical field degrees of
freedom. In the rest frame of a single particle (the usual choice of the
"standard momentum" for massive particles in the representation theory of
the Poincare group) these are the three spin components of the particle (the
little group is SO(3), and thus one has usual spin, defined by the behavior
of the states of particles at rest under the rotation group, i.e. (2s+1)
spin states).

For massless particles, however, the little group is generated by boosts in
direction of the light-like (!) standard momentum (usually the z axis) and
two independent "null rotations". This group is equivalent to ISO(2), the
affine group of the Euclidean 2D plane. In order not to have continuous
spin-like intrinsic degrees of freedom of massless particles, one assumes
that the null rotations are represented trivially, which implies that
massless vector fields (and such of higher spin) admit corresponding local
gauge transformations and are thus gauge fields. The physical degrees of
freedom are the two helicity states \pm s.

I've never thought about it carefully, but wouldn't indeed a finite mass of
photons, no matter how tiny it might be (the most recent PDB lists 10^-18 eV
as upper limit) change the thermodynamics of black-body radiation? At a
given temperature one would have a larger photon density by a factor of 3/2.
Shouldn't this have been observed then by now?

Arnold Neumaier wrote:

--
Hendrik van Hees
Justus-Liebig Universität Gießen
D-35392 Gießen
http://theorie.physik.uni-giessen.de/~hees/

Ralph Hartley

unread,
Aug 18, 2010, 3:24:01 PM8/18/10
to
On 08/18/2010 06:47 AM, Hendrik van Hees wrote:
> There is one more difference between a massless and a massive vector field,
> and that's the fact that a massive field has three physical field degrees of
> freedom.
...

> I've never thought about it carefully, but wouldn't indeed a finite mass of
> photons, no matter how tiny it might be (the most recent PDB lists 10^-18 eV
> as upper limit) change the thermodynamics of black-body radiation? At a
> given temperature one would have a larger photon density by a factor of 3/2.
> Shouldn't this have been observed then by now?

Must the Photon Mass be Zero?
L. Bass, E. Schrödinger
http://www.jstor.org/stable/99679

Discusses exactly this question. (answer:no)

The longitudinal degrees of freedom couple to matter so weakly, they can
never come to thermal equilibrium. Any practical cavity is effectively
transparent to to longitudinally polarized photons. In the m=0 limit
they don't interact at all, making them non-physical.

The effects of photon mass on black body radiation are (according to
http://link.aps.org/doi/10.1103/PhysRevA.32.623) much too small for any
reasonable experiment to detect (given existing upper bounds).

Ralph Hartley

Oh No

unread,
Aug 18, 2010, 7:06:53 PM8/18/10
to
Thus spake "jw...@BasicISP.net" <jw...@BasicISP.net>

>I'll try once more.
>
>It has nothing to do with the exact speed of light, nor with
>propagation of
>information. We are not talking about information, but about light.

Actually we are talking about the structure of spacetime, by which we
may mean a mathematical structure, i.e. R$, whose elements can be found
from a particular class of physical processes, namely, measurements of
time and position. Those processes only make sense given a maximal speed
of information. To the best of our knowledge this is equal to the speed
of light, but because all measurement has finite accuracy, we cannot say
with absolute certainty that the two are equal.


>
>The infinite precision only depends upon the empirical fact that
>the velocity of light does not add to the velocity of an object
>sending or receiving light.

That is an empirical fact known only with finite precision. Experiments
like the one describe improve the precision with which it is known.


>
>It also depends upon the assumption that light consists of photons.

This is quite another matter. We don't need an assumption here. Both the
theoretical and empirical evidence are conclusive.

Arnold Neumaier

unread,
Aug 18, 2010, 7:06:54 PM8/18/10
to
Arnold Neumaier wrote:
> jw...@BasicISP.net wrote:

>> If a photon had mass, then its velocity would depend on existence
>> of a rest frame for that photon. If such a rest frame existed,
>> then the velocity of light would add to that of the object at
>> its points of creating and annihilation. But this never is observed.
>
> Observations are never exact, so sufficiently small deviations
> (and hence a sufficiently small mass of the photon) cannot
> be observed. The point of such experiments is precisely to
> imporve the limit on how small these deviations must be.

A thorough analysis of the problems involved in determining the mass of
the photon is given in the paper
L.-C. Tu, J. Luo, and G.T. Gillies,
The mass of the photon
Rep. Prog. Phys. 68 (2005), 77--130.
See also
AS Goldhaber, MM Nieto
Photon and graviton mass limits
Reviews of Modern Physics, 2010
arxiv:0809.1003

Arnold Neumaier

unread,
Aug 18, 2010, 7:06:53 PM8/18/10
to
Hendrik van Hees wrote:
> There is one more difference between a massless and a massive vector field,
> and that's the fact that a massive field has three physical field degrees of
> freedom. In the rest frame of a single particle (the usual choice of the
> "standard momentum" for massive particles in the representation theory of
> the Poincare group) these are the three spin components of the particle [...]

>
> massless vector fields (and such of higher spin) admit corresponding local
> gauge transformations and are thus gauge fields. The physical degrees of
> freedom are the two helicity states \pm s.
>
> I've never thought about it carefully, but wouldn't indeed a finite mass of
> photons, no matter how tiny it might be (the most recent PDB lists 10^-18 eV
> as upper limit) change the thermodynamics of black-body radiation?

It would change it by a tiny amount only. There is no discontinuity.


> At a given temperature one would have a larger photon density by a factor of 3/2.

Why? This would assume that all modes are equivalent.
But the transverse state would be very little populated only.

The reason is that for almost massless particles, the transverse mode is
suppressed by a factor proportional to its mass. (Or even mass square?
I'd need to check; haven't done such calculations for a long time.)
This can be seen by writing the massive representation in the Jacob-Wick
helicity form, where the massless limit can be taken. The massless case
is just a conventional limit of the massive case, without any
discontinuity when the mass approaches zero.
Jacob M & Wick G C.
On the general theory of collisions for particles with spin.
Ann.Phys. 7:404-428, 1959.

Thus what is known about black body radiation gives just like any other
effect only an upper bound on the photon mass.


Arnold Neumaier


Ralph Hartley

unread,
Aug 19, 2010, 1:56:15 AM8/19/10
to
On 08/17/2010 10:01 PM, jw...@BasicISP.net wrote:

> If a photon had mass, then its velocity would depend on existence
> of a rest frame for that photon. If such a rest frame existed,
> then the velocity of light would add to that of the object at
> its points of creating and annihilation.

Also higher energy (blue) photons would go slightly faster than lower
energy (red) ones. Light from an approaching source would be faster, and
therefore appear bluer. The Doppler shift is about the same for zero
mass and very small mass.

> Surely, no one is questioning the idea of vector addition?

This remark makes me suspect that you may have forgotten that the
relativistic formula for adding velocities is s=(a+b)/(1+ab/c^2) not
s=a+b. If one of the two velocities is very close c (and the other is
not as close) then the sum will be very close to c as well.

Ralph Hartley

J. J. Lodder

unread,
Aug 19, 2010, 4:09:55 AM8/19/10
to
Arnold Neumaier <Arnold....@univie.ac.at> wrote:

[Moderator's note: Huge amount of quoted text snipped. -P.H.]

> > Therefore, photons have no rest frame, no proper time, nor
> > any mass.
>
> This only holds if you _define_ photons to be massless particles.
> The question then would become whether these theoretical photons
> describe Nature exactly. But the testing problem would remain the same.
>
> > Under reasonable assumptions, it eaasily can be proven
> > deductively that all massless particles travel at the
> > same speed.
>
> Yes, and currently, this speed is called the speed of light.
> Would the photon be found to have a slight mass, this speed would
> simply be renamed to ''limit speed'' or something like that,
> photons could come to rest, but not much else would change.

(if so) How would you go about decelerating a photon?

Jan

Tom Roberts

unread,
Aug 19, 2010, 4:10:26 AM8/19/10
to
jw...@BasicISP.net wrote:
> I'll try once more.
> It has nothing to do with the exact speed of light, nor with
> propagation of
> information. We are not talking about information, but about light.
> The infinite precision only depends upon the empirical fact that
> the velocity of light does not add to the velocity of an object
> sending or receiving light.

But turned into a limit on photon mass, this is at least some TRILLIONS
of times less stringent than the measurement being discussed. That comes
from estimates based on the neutrino/photon speed difference from
SN1987A and the lower limit on neutrino masses from oscillation
measurements.

If the light emitted from a source moving with velocity v toward the
observer has a speed c+kv in the observer's frame, then these
experiments place a limit on k. See
http://math.ucr.edu/home/baez/physics/Relativity/SR/experiments.html#moving-source_tests
for a list of experiments. The best limit on k is k < 2E-9.

That is by no means "infinite precision".

> It also depends upon the assumption that light consists of photons.
> If a photon had mass, then its velocity would depend on existence
> of a rest frame for that photon. If such a rest frame existed,
> then the velocity of light would add to that of the object at
> its points of creating and annihilation. But this never is observed.

You are ASSUMING that SR is valid. that does not constitute an
experimental test in any way.

> Surely, no one is questioning the idea of vector addition?

Yes, among other things we are. That's what experimental physicists do:
test theories. Your notion of "vector addition" is merely a theory of
how such things ought to behave.

BTW in SR velocity vectors do NOT add via vector addition. Rather,
they are composed via the Lorentz composition of velocities formula.

> Therefore, photons have no rest frame, no proper time, nor
> any mass.

You make quite strong assumptions not in evidence. Your claim is
worthless.

> Under reasonable assumptions, it eaasily can be proven
> deductively tha all massless particles travel at the
> same speed.

It is not "reasonable" to ASSUME that SR or GR is EXACTLY correct. This
is an experimental issue, not one to be solved from your armchair by
assuming some particular physical theory is true.

Tom Roberts

Bob_for_short

unread,
Aug 19, 2010, 7:11:01 AM8/19/10
to
On 19 août, 10:09, nos...@de-ster.demon.nl (J. J. Lodder) wrote:
>
> (if so) How would you go about decelerating a photon?

The only case of photon acceleration/deceleration known to me is the
photon propagation in a transparent non-uniform medium, i.e., in a
medium with real and variable n(r).

Arnold Neumaier

unread,
Aug 19, 2010, 3:35:13 PM8/19/10
to

Since it is uncharged, it is not easy. There is no conservation law,
and photons are easily absorbed, so it is likely to be destroyed long
before it has a small speed. The ''could'' was meant as a possibility
allowed by the theory, not as something we can actually do.
(Your question is like how would you go about picking an ounce of
solar matter from the center of the sun.)

Juan R. Gonzàlez-Àlvarez

unread,
Aug 20, 2010, 1:55:12 AM8/20/10
to
jw...@BasicISP.net wrote on Tue, 17 Aug 2010 22:01:55 -0400:

> I'll try once more.
>
> It has nothing to do with the exact speed of light, nor with
> propagation of information. We are not talking about information, but
> about light.
>
> The infinite precision only depends upon the empirical fact that the
> velocity of light does not add to the velocity of an object sending or
> receiving light.
>
> It also depends upon the assumption that light consists of photons.
>
> If a photon had mass, then its velocity would depend on existence of a
> rest frame for that photon. If such a rest frame existed, then the
> velocity of light would add to that of the object at its points of
> creating and annihilation. But this never is observed.
>
> Surely, no one is questioning the idea of vector addition?
>
> Therefore, photons have no rest frame, no proper time, nor any mass.

(...)

First, the idea of measuring something with "infinite precision" as
you claim is unreal, because it would mean recording infinite data
(0.000000000000000000000000000000000000...) among other issues.

Second, the statistics behind the current measurements of phton
properties is understood and the upper limit stated. The relevant
article is

Review of Particle Physics 2008: Physics Letters B 667(1-5), 1-6.
Amsler, C. et al. (Particle Data Group).

giving the upper limit for mass

m < 10^(-18) eV/c^2

and the next limit for the charge of photon

q < 10^(-35) e

--
http://www.canonicalscience.org/

BLOG:
http://www.canonicalscience.org/publications/canonicalsciencetoday/canonicalsciencetoday.html

Arnold Neumaier

unread,
Aug 20, 2010, 1:56:13 AM8/20/10
to
Bob_for_short wrote:

Yes, but this is due to changes in the nature of the photon.

Photons in matter are effective photons, dressed differently than
in vacuum.

Their lower speed has nothing to do with photon mass, since effective
photons are still massless.

Bob_for_short

unread,
Aug 21, 2010, 11:05:44 PM8/21/10
to
On 20 ao?t, 07:56, Arnold Neumaier <Arnold.Neuma...@univie.ac.at>
wrote:
> Bob_for_short wrote:

I don't think that the "nature" of photon may be changed in a
transparent medium, if we mean the photon electric and magnetic
fields. But in a medium we see that a photon is not a free "particle"
but a collective excitation mode. I think that "vacuum" for photon is
just a uniform medium and photon itself is its deviation from
uniformity.

Of course, slowly propagating photons are still massless.

J. J. Lodder

unread,
Aug 22, 2010, 3:16:57 PM8/22/10
to
Arnold Neumaier <Arnold....@univie.ac.at> wrote:

What you can do is run along with it
at a very relativistic speed. (Einstein's thought experiment)
This however also downshifts the photon frequency to zero.
So you would need a zero temperature container as well.
(or a perfectly reflecting one) [1]

My idea for building a photon cooler (without running so fast) would be
to reflect the photon a great many times from receding mirrors. Should
work, very much in principle,

Jan

[1] Since a Kelvin is about 20 GHz a microHertz photon will be
obliterated by noise at 10^{-16} K

jw...@basicisp.net

unread,
Aug 22, 2010, 3:17:28 PM8/22/10
to
On Aug 19, 4:11 am, Bob_for_short <vladimir.kalitvian...@wanadoo.fr>
wrote:

Although LIGHT can change velocity in a nonvacuum medium,
a photon can not. The speed of light declines in water or glass,
for example, because photons propagating in those media are
repeatedly absorbed by atoms; generally, for each such photon,
a new photon is reemitted with the same momentum. Of course,
sometimes the photon is absorbed and its momentum goes toward
modifying the momentum (and energy) of the atom.

The atom takes time to reemit a photon, causing the average
speed to decline below c. Photons always travel
only exactly at speed c; they can not change during propagation
because they have no proper time.

Photons have specific wavelengths, but they do not oscillate
during propagation -- their phase is effectively spatial, only,
and it differs at annihilation as a function of distance only,
not as a function of proper time.

Arnold Neumaier

unread,
Aug 22, 2010, 3:18:28 PM8/22/10
to
Bob_for_short wrote:
> On 20 ao?t, 07:56, Arnold Neumaier <Arnold.Neuma...@univie.ac.at>
> wrote:
>> Bob_for_short wrote:
>>> On 19 ao?t, 10:09, nos...@de-ster.demon.nl (J. J. Lodder) wrote:
>>>> (if so) How would you go about decelerating a photon?
>>> The only case of photon acceleration/deceleration known to me is the
>>> photon propagation in a transparent non-uniform medium, i.e., in a
>>> medium with real and variable n(r).
>> Yes, but this is due to changes in the nature of the photon.
>>
>> Photons in matter are effective photons, dressed differently than
>> in vacuum.
>>
>> Their lower speed has nothing to do with photon mass, since effective
>> photons are still massless.
>
> I don't think that the "nature" of photon may be changed in a
> transparent medium, if we mean the photon electric and magnetic
> fields.

I didn't assign any particular (or changing) ''nature'' to the photon.

For me, a photon is the lowest quantum excitiation of an
electromagnetic field. The electromagnetic field in a transparent
non-uniform medium is an effective field with different equations than
the free field; so it is also dressed differently than an e/m field
in vacuum.

> But in a medium we see that a photon is not a free "particle"
> but a collective excitation mode.

This makes the photon in a medium quite a different object than in
vacuum.

> I think that "vacuum" for photon is
> just a uniform medium and photon itself is its deviation from
> uniformity.

It is a quantum excitation of the medium.

> Of course, slowly propagating photons are still massless.

And that was the real point of my mail.

Lester Welch

unread,
Aug 22, 2010, 8:28:39 PM8/22/10
to
On Aug 22, 3:17 pm, "jw...@BasicISP.net" <jw...@BasicISP.net> wrote:
>
> Although LIGHT can change velocity in a nonvacuum medium,
> a photon can not.   The speed of light declines in water or glass,
> for example, because photons propagating in those media are
> repeatedly absorbed by atoms; generally, for each such photon,
> a new photon is reemitted with the same momentum.  

In this model, how is coherence maintained? I would think that the
time interval between absorption and emmission of the photon is a
quantum - i.e., random - process that coherence would be destroyed.

Bob_for_short

unread,
Aug 23, 2010, 4:16:57 PM8/23/10
to
On 22 August, 21:17, "jw...@BasicISP.net" <jw...@BasicISP.net> wrote:
> On Aug 19, 4:11 am, Bob_for_short <vladimir.kalitvian...@wanadoo.fr>
> wrote:
>
> > On 19 august, 10:09, nos...@de-ster.demon.nl (J. J. Lodder) wrote:
>
> > > (if so) How would you go about decelerating a photon?
>
> > The only case of photon acceleration/deceleration known to me is the
> > photon propagation in a transparent non-uniform medium, i.e., in a
> > medium with real and variable n(r).
>
> Although LIGHT can change velocity in a nonvacuum medium,
> a photon can not.

So, according to you, there is no time delay in photon propagation
through a medium layer. The why the LIGHT has this "retardation" if it
consists of non interaction photons?

> The speed of light declines in water or glass,
> for example, because photons propagating in those media are
> repeatedly absorbed by atoms; generally, for each such photon,

> a new photon is re-emitted with the same momentum.

I think it is plainly wrong about atoms. It is the medium (lots of
bound atoms) that has some EMF energy levels that take part in the
propagation. In particular, in the window of transparency nothing is
absorbed and re-emitted or even "interacts" with the medium. It is
better to think in terms of the medium proper EMF modes with somewhat
different from vacuum properties.

> The atom takes time to re-emit a photon, causing the average


> speed to decline below c.

Rubbish. Nothing can force an excited atom to re-emit a photon in the
same direction. Next, nothing can forme an excited atom to re-emit a
photon with the same delay. Delays are random: exp(-t/T). Finally, in
the theoretical description it is not atomic levels that are involved
but those of the medium. In other words, a medium is not a simple
collection of non interacting atoms.

Arnold Neumaier wrote:

>> Yes, but this is due to changes in the nature of the photon.

> I didn't assign any particular (or changing) ''nature'' to the photon.

Choose one of the statements to be self-consistent.

>> Photons in matter are effective photons, dressed differently than
>> in vacuum.

And in vacuum they are ineffective or what?

> For me, a photon is the lowest quantum excitation of an
> electromagnetic field.

Better, it is the first excited mode of the EMF in a given situation.

> The electromagnetic field in a transparent
> non-uniform medium is an effective field with different equations than
> the free field; so it is also dressed differently than an e/m field
> in vacuum.

Dressing means that there is something undressed, bare. But bare
photon is a purely theoretical bla-bla involved in the renormalization
ideology and it has nothing to do with always physical photons in
physics. Everybody agrees that EMF equations in a medium are different
(involve n(r) =/= 1) but the nature of equations remains practically
the same. As a result, the space frequency of a photon changes but not
the frequency omega in exp(-i*omega*t) (in a still medium, of course).
The omega determines the photon energy.

Concerning the photon mass, there is no such a thing in a theory even
in resonators with the frequency spectrum starting at finite omega > 0.

Oh No

unread,
Aug 23, 2010, 6:01:45 PM8/23/10
to
Thus spake Bob_for_short <vladimir.k...@wanadoo.fr>

>> The atom takes time to re-emit a photon, causing the average
>> speed to decline below c.
>
>Rubbish. Nothing can force an excited atom to re-emit a photon in the
>same direction.

Actually conservation laws can do exactly that. Remember this is not an
excitation mode of the atom.

> Next, nothing can forme an excited atom to re-emit a photon with the
>same delay.

No one suggested it did.

>Delays are random: exp(-t/T). Finally, in the theoretical description
>it is not atomic levels that are involved but those of the medium. In
>other words, a medium is not a simple collection of non interacting
>atoms.

No, clearly it is a complex collection of interacting atoms. It is atoms
nonetheless, and must get its behaviour from the behaviour of atoms and
their constitituents.

Oh No

unread,
Aug 23, 2010, 6:02:01 PM8/23/10
to
Thus spake Lester Welch <lester...@gmail.com>
The explanation requires that one understands wave functions as
conveyors of information, not as physical waves. We have no way of
knowing, even in principle, when a photon is absorbed or by which
electron. The wave function does not then describe the behaviour of an
individual photon or photons, but rather the net effect.

jw...@basicisp.net

unread,
Aug 24, 2010, 3:06:59 AM8/24/10
to
Here's my call on that:

LIGHT coherence is maintained because in a transparent medium such as
glass, all the atoms are about the same, to the extent that almost none
are absorbing (without reemitting). Because it is transparent,
reemission of photons preserves their momentum p, and with time delay
about the same for each atom.

It is true that absorption and reemission of a single photon is an event
requiring quantum-mechanical analysis. However, coherence of light is
determined by expectancies, <p>, not p of individual photons.
Transmission of light through any medium always introduces some
dispersion of phase, some loss of coherence. If you want, you can view
this as the Second Law in action.

Also, one can not have "coherence" except for light of a given
wavelength. This means that momentary absorption and reemission always
takes place in the same energy eigenstate of all atoms involved. It is
the eigenstate that keeps the delay identical for all transmission
through the medium, thus preserving coherence (at a reduced wavelength,
frequency being constant, while the light is in the medium).

jw...@basicisp.net

unread,
Aug 25, 2010, 12:35:30 PM8/25/10
to
There is an alternate explanation which does not assume
interaction of the photons with atoms of a transparent
medium:

Photons have no inertial rest frame; therefore they can not
be accelerated relative to anything; thus, they always travel
at speed c. However, c might vary in a transparent medium
such as water or glass:

We know that c = sqrt(epsilon_0*mu_0)^-1 in vacuum;
suppose that the permittivity and/or permeability of the
transparent medium was higer than in vacuum, which
would make c lower. Then, we could assume that a
photon leaving a transparent medium was identical to
one which entered it and just never interacted at all.

The reason I favor brief absorption and reemission
over change in c, is that I know that any atom
can be manipulated by "laser tweezers" -- shining
a laser on a surface to which an atom adheres
exerts a force on the atom and can move it. This
is because of interaction of the photons with the
electric fields of the atom's electrons. So, I
don't see how a photon could get through a
sheet of glass with billions of atoms, all
bound by electric charges, without intercting.

Also, why does adding a dense substance such as lead
to glass increase its refractive index? As density
goes up, generally refractive index goes up.
The greater density is from the atomic nuclei, not
its electrons. The electrons in a higher-Z material
are on the average farther from its nucleus than for
a lower=Z atom; thus, speeds of electronic relaxation
would average lower; therefore, reemission after
absorption would be slower, increasing the refractive
index.

Finally, a mirror reflects light back; this can't
possibly be because of a change in c but rather involves
transfer of momentum to the mirror. The reflected
photons can't possibly be the same as the ones
incident on the mirror.

glird

unread,
Sep 4, 2010, 5:03:47 PM9/4/10
to
[ Mod. note: Note to poster, please include distinguishing marks around
quoted text, as is done in most other posts, to avoid confusion of
attribution. Quotation marks were restored. -ik ]

On Aug 25, 12:35 pm, jw wrote:
>
> There is an alternate explanation which does not assume interaction of the photons with atoms of a transparent medium:
> Photons have no inertial rest frame; therefore they can not be
> accelerated relative to anything; thus, they always travel
> at speed c.   However, c might vary in a transparent medium such as
> water or glass:
> We know that c = sqrt(epsilon_0*mu_0)^-1 in vacuum; suppose that the

> permittivity and/or permeability of the transparent medium was higher


> than in vacuum, which
> would make c lower.   Then, we could assume that a photon leaving a
> transparent medium was identical to one which entered it and just
> never interacted at all.

> The reason I favor brief absorption and re-emission over change in c,


> is that I know that any atom can be manipulated by "laser tweezers" --
> shining a laser on a surface to which an atom adheres exerts a force
> on the atom and can move it.   This is because of interaction of the
> photons with the electric fields of the atom's electrons.  So, I
> don't see how a photon could get through a
> sheet of glass with billions of atoms, all

> bound by electric charges, without interacting. >

All of the above assumes that a photon is a particle that exists and
transmits through the empty space in a vacuum. That assumption stems
from Einstein's theory, not from Max Planck. In Planck's theory there
are two different quanta;
1. A quantum of action, h = 2pirmc'; and
2. A quantum of energy, e_0.
Though different from each other, they are related by the equation,
e_0 = hf,
in which f, the frequency of the waves of h,
depends on the state of motion of the observer.
The meaning of h = 2pirmc' is this:
Let there be a resonator in a black box. In Planck's paper, he took
the resonator as stationary, thus at absolute zero degrees K.
Even so, inside such an atom a circulating wave system called an
"electron" exists. It moves therein at c' = cFs, where Fs is the fine
structure constant and c is the speed of light in a vacuum.
If such a wave system and its trapping material medium escapes from
an atom -- say during a quantum emission -- the remaining weight of
the atom is m_grams less than it had been. This "mass" (actually
weight) is attributed to the electron.
Given that the orbital path of such an electron is 2pir long, in
which r is the average length of the radius of an atom under the
specified conditions, then the quantity of action, h, is equal to
2pir *m * (c' = cFs).

When such a quantity of matter with its trapped wave system escapes
from an atom, it is an increase in the local density of the
surrounding material field. Such a +grad d causes a local increase in
pressure, thus a =grad s exists. Since a pressure disequilibrium
cannot remain present in the compressible material filling the local
field, it will radiate away in all directions, until grad s = 0 for
any value of grad d.

> Also, why does adding a dense substance such as lead to glass increase its refractive index?  As density goes up, generally refractive index goes up.
> The greater density is from the atomic nuclei, not its electrons.
>  The electrons in a higher-Z material are on the average farther from
> its nucleus than for a lower=Z atom; thus, speeds of electronic

> relaxation would average lower; therefore, re-emission after


> absorption would be slower, increasing the refractive index. >

Light is not made of separate particles, it is waves of pressure
transmitted by the matter in the local field. the speed of each wave
depends on how one elects to measure it. (If you use the usual
method, you measure how many feet -- or miles or km -- it travels per
second. Since that is a function of the density of the traversed
medium, the results will be variable.) However, if you measure the
quantity of matter traversed per second, the result will be a constant
regardless of the variable density of the luminiferous medium.
Even though there are zillions of tiny atoms per cc in a material
field, between any group of them there is always a space filled by
continuous matter. If its density is less than theirs, light waves
will move faster through that space than through the atom-filled
places. If they can get through it, the medium will be transparent. If
you add lead to glass, thereby increasing the local density, that will
increase the refractive index.

> Finally, a mirror reflects light back; this can't possibly be because of a change in c but rather involves transfer of momentum to the mirror.  The reflected photons can't possibly be the same as the ones incident on the mirror. >

Momentum is mass times velocity. If a light wave had any rest mass
at all, it would be infinitely great when moving at c; thus would
eradicate anything it hit. Since that never happens, there is no such
thing as the "mass" of a quantum of energy; which Einstein called "a
photon". For reasons I won't get into here, Planck was right;
although a quantity of energy e_0 IS emitted and/or absorbed by an
atom, no such quantum exists during transmission.

glird

Reply all
Reply to author
Forward
0 new messages