Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Sagnac & relativity

24 views
Skip to first unread message

Mark Samokhvalov

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to

Sagnac & relativity

In this group I recently found statements to the effect that the Sagnac
experiment was explained and even predicted (sic!) by SR. I have perused
many books on relativity dating back to 1920, including Einstein's, and saw
no mention of Sagnac, until I came across a textbook by M.-A.Tonnella on
magnetism and relativity ('55?), where a GR explanation of it was
presented.
I have the following questions to relativity fans:
1. Do you agree, that, if two waves radiated at the same point return to
that point with an accumulated phase difference, this means that there is a
difference in either their:
a) paths, or
b) phase velocities?
Or, perhaps, you can suggest another possibility?
2. When, as in Sagnac experiment, two waves are made to travel in opposite
directions along the same path (typically, an optical filament), does this
path in some mysterious way become, in the course of the ring's rotation,
different for the clock-wise and the anti-clockwise waves?
3. If not, how, with the phase velocity being equal for both waves, can a
phase difference accumulate?
Thanks for the answers.

Dennis McCarthy

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
MS: >2. When, as in Sagnac experiment, two waves are made to travel in opposite

>directions along the same path (typically, an optical filament), does this
>path in some mysterious way become, in the course of the ring's rotation,
>different for the clock-wise and the anti-clockwise waves?

Dennis: You may not believe this, but that is precisely the relativist
argument. Heck, people accepted the Copenhagen interpretation for decades, why
not that a path is longer clockwise than counter-clockwise? There is simply no
limit or boundary to what people are willing to believe in defense of certain
theories.


Dennis McCarthy


Stephen

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
In article <8a30d6$2j19$1...@gavrilo.mtu.ru>, "Mark Samokhvalov"
<samokh...@mtu-net.ru> wrote:

> Sagnac & relativity
>
> In this group I recently found statements to the effect that the Sagnac
> experiment was explained and even predicted (sic!) by SR. I have perused
> many books on relativity dating back to 1920, including Einstein's, and saw
> no mention of Sagnac, until I came across a textbook by M.-A.Tonnella on
> magnetism and relativity ('55?), where a GR explanation of it was
> presented.
> I have the following questions to relativity fans:
> 1. Do you agree, that, if two waves radiated at the same point return to
> that point with an accumulated phase difference, this means that there is a
> difference in either their:
> a) paths, or
> b) phase velocities?
> Or, perhaps, you can suggest another possibility?

> 2. When, as in Sagnac experiment, two waves are made to travel in opposite
> directions along the same path (typically, an optical filament), does this
> path in some mysterious way become, in the course of the ring's rotation,
> different for the clock-wise and the anti-clockwise waves?

If you look at the path in an inertial frame (in which the speed of
light going round the ring is c) it's clear that, as the ring is rotating,
a light beam travelling in the direction of rotating has to travel further
than one travelling against the direction of rotation in order to return
to the point on the ring from which they were emitted. I would not
describe that as 'mysterious' but rather as 'obvious'.

--
Felix qui potuit rerum cognoscere causas - Virgil.

sh...@my-deja.com

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
Mark Samokhvalov wrote:

> Sagnac & relativity
>
> In this group I recently found statements to the effect that the
> Sagnac experiment was explained and even predicted (sic!) by SR. I
> have perused many books on relativity dating back to 1920, including
> Einstein's, and saw no mention of Sagnac, until I came across a
> textbook by M.-A.Tonnella on magnetism and relativity ('55?), where
> a GR explanation of it was presented.
> I have the following questions to relativity fans:

> 1. Do you agree, that, if two waves radiated at the same point return
> to that point with an accumulated phase difference, this means that
> there is a difference in either their:
> a) paths, or
> b) phase velocities?
> Or, perhaps, you can suggest another possibility?

> 2. When, as in Sagnac experiment, two waves are made to travel in
> opposite directions along the same path (typically, an optical
> filament), does this path in some mysterious way become, in the
> course of the ring's rotation, different for the clock-wise and
> the anti-clockwise waves?

> 3. If not, how, with the phase velocity being equal for both waves,


> can a phase difference accumulate?
> Thanks for the answers.

SR predicts the Sagnac effect quite easily, and here's a beautiful
derivation and explanation:

http://mathpages.com/rr/s2-07/2-07.htm

1a) yes, different path lengths

2) no mystery needed, a rotating frame is non-inertial


---Tim Shuba---

Sent via Deja.com http://www.deja.com/
Before you buy.

sh...@my-deja.com

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
Dennis McSophist wrote:

> > When, as in Sagnac experiment, two waves are made to travel in
> > opposite directions along the same path (typically, an optical
> > filament), does this path in some mysterious way become, in the
> > course of the ring's rotation, different for the clock-wise and
> > the anti-clockwise waves?
>

> Dennis: You may not believe this, but that is precisely the
> relativist argument. Heck, people accepted the Copenhagen
> interpretation for decades, why not that a path is longer
> clockwise than counter-clockwise? There is simply no
> limit or boundary to what people are willing to believe
> in defense of certain theories.

No limit? Yes indeed Mr Sophist. How many dozens of posts have
you made about Sagnac? SR's explanation is simple and consistent.

See http://mathpages.com/rr/s2-07/2-07.htm for an explanation
that includes the exact reason why the Sagnac effect is perfect
fodder for your sophistry. Note especially the last paragraph.

Dennis McCarthy

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to

Dennis: Unfortunately, here's the problem:
"Still, the doubter imagines that we can transform this bias away, by treating
the situation with respect to coordinates that are rigidly attached to and
rotating along with the device. Surely (he reasons) we can regard the paths of
the two light beams as spatially congruent and equal as seen from this system
of reference, and so the asymmetric travel times must imply anisotropic light
speed with respect to these coordinates. This is actually true, in the sense
that it's possible to define a system of coordinates in terms of which the
positions of the points on the disk are independent of the time coordinate, but
of course such coordinates are necessarily accelerating (viz., rotating), and
special relativity does not assume light speed to be isotropic with respect to
non-inertial coordinates."

That's called the "naive" SR explanation. Allowing anisotropy wrt the
non-inertial rim observer results in the Selleri paradox.
If you want the correct SR explanation, check Rizzi and Tartaglia's.
You'll find it's not so simple. And you'll find that the explanation does in
fact involve that " this path in some mysterious way become, in the course of


the ring's rotation, different for the clock-wise and

the anti-clockwise waves"--even for the rim observer.
Dennis McCarthy


Dennis McCarthy

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
>
>
>In article <8a30d6$2j19$1...@gavrilo.mtu.ru>, "Mark Samokhvalov"
><samokh...@mtu-net.ru> wrote:
>
>> Sagnac & relativity
>>
>> In this group I recently found statements to the effect that the Sagnac
>> experiment was explained and even predicted (sic!) by SR. I have perused
>> many books on relativity dating back to 1920, including Einstein's, and saw
>> no mention of Sagnac, until I came across a textbook by M.-A.Tonnella on
>> magnetism and relativity ('55?), where a GR explanation of it was
>> presented.
>> I have the following questions to relativity fans:
>> 1. Do you agree, that, if two waves radiated at the same point return to
>> that point with an accumulated phase difference, this means that there is a
>> difference in either their:
>> a) paths, or
>> b) phase velocities?
>> Or, perhaps, you can suggest another possibility?
>> 2. When, as in Sagnac experiment, two waves are made to travel in opposite

>> directions along the same path (typically, an optical filament), does this
>> path in some mysterious way become, in the course of the ring's rotation,
>> different for the clock-wise and the anti-clockwise waves?
>
Wells: If you look at the path in an inertial frame (in which the speed of

>light going round the ring is c) it's clear that, as the ring is rotating,
>a light beam travelling in the direction of rotating has to travel further
>than one travelling against the direction of rotation in order to return
>to the point on the ring from which they were emitted. I would not
>describe that as 'mysterious' but rather as 'obvious'.

Dennis: Ahh, so you think it is "obvious" that when you move at a light beam,
it approaches you faster than when you move away from it?
I agree that that is obvious, but you wouldn't believe that some people try to
deny it.


Dennis McCarthy


Dennis McCarthy

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
>SR predicts the Sagnac effect quite easily, and here's a beautiful
>derivation and explanation:
>
>http://mathpages.com/rr/s2-07/2-07.htm
>
>1a) yes, different path lengths
>
>2) no mystery needed, a rotating frame is non-inertial

D: Sorry. 1) The path lengths aren't different to a person on the rim.
2) The non-inertial fall back has been rejected by mainstream physicists.
Check Rizzi and Tartaglia's explanation for the correct SR version.
Dennis McCarthy


Paul Stowe

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
In <stephenwells-0...@mac009.joh.cam.ac.uk>
> If you look at the path in an inertial frame (in which the speed of
>light going round the ring is c) it's clear that, as the ring is
rotating,
>a light beam travelling in the direction of rotating has to travel
further
>than one travelling against the direction of rotation in order to
return
>to the point on the ring from which they were emitted. I would not
>describe that as 'mysterious' but rather as 'obvious'.

Geez, we agree. What a shocker...

Paul Stowe

Standeven

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to

Mark Samokhvalov wrote:

> Sagnac & relativity
>
> In this group I recently found statements to the effect that the Sagnac
> experiment was explained and even predicted (sic!) by SR. I have perused
> many books on relativity dating back to 1920, including Einstein's, and saw
> no mention of Sagnac, until I came across a textbook by M.-A.Tonnella on
> magnetism and relativity ('55?), where a GR explanation of it was
> presented.
> I have the following questions to relativity fans:
> 1. Do you agree, that, if two waves radiated at the same point return to
> that point with an accumulated phase difference, this means that there is a
> difference in either their:
> a) paths, or
> b) phase velocities?
> Or, perhaps, you can suggest another possibility?

It could be their frequencies which differ.


> 2. When, as in Sagnac experiment, two waves are made to travel in opposite
> directions along the same path (typically, an optical filament), does this
> path in some mysterious way become, in the course of the ring's rotation,
> different for the clock-wise and the anti-clockwise waves?

> 3. If not, how, with the phase velocity being equal for both waves, can a
> phase difference accumulate?

What makes you think that the phase velocity is equal for the two waves?
You're not in an inertial frame...


Tom Roberts

unread,
Mar 7, 2000, 3:00:00 AM3/7/00
to
Mark Samokhvalov wrote:
> In this group I recently found statements to the effect that the Sagnac
> experiment was explained and even predicted (sic!) by SR.

It is easily explained by SR. I do not think it was predicted by
any SR advocate before Sagnac performed his experiment (but I am
not an expert on such historical questions).


> 1. Do you agree, that, if two waves radiated at the same point return to
> that point with an accumulated phase difference, this means that there is a
> difference in either their:
> a) paths, or
> b) phase velocities?
> Or, perhaps, you can suggest another possibility?

There could be a difference in the gravitational potential
integrated along their paths. There could be different optical
media in their paths, or media moving differently (but are these
differences in path?). There are probably other possibilities....


> 2. When, as in Sagnac experiment, two waves are made to travel in opposite
> directions along the same path (typically, an optical filament), does this
> path in some mysterious way become, in the course of the ring's rotation,
> different for the clock-wise and the anti-clockwise waves?

This depends upon what coordinates (aka reference frame) you use to
describe it. Note that both speed and spatial path length are
inherently coordinate dependent.

Let me assume the center of rotation of the ring interferometer is
at rest in an inertial frame. In this frame SR predicts the speed
of light is isotropically c; in this frame it is clear that the
co-rotating and counter-rotating light rays travel different
distances. So there is no mystery -- the light rays travel
different distances at the same speed and therefore accumulate a
phase difference.

One might also describe this in coordinates rotating with the
ring interferometer. In these coordinates the _average_ speed
of light _going_all_the_way_around_the_ring_ is different for
the two directions, but the distance is the same for both rays.
Again there is no mystery -- the light rays travel the same
distance at different average speeds and therefore accumulate
a phase difference.

Note that it is tricky to define "speed of light"
in these coordinates, but there is indeed a well-defined
_average_ speed of light _going_all_the_way_around_.
This is intimitely related to the difficulty of
synchronizing clocks in a rotating system....

One could also describe this in the inertial frame which is
co-moving with the emitter/detector when the rays are emitted
(think of very short pulses of light). Note that the emitter/
detector moves (slightly) in this frame while the rays are in
transit. In this frame the speed of light is isotropically c,
and the two paths the rays follow are distorted circles which
are not congruent, and which have different lengths. Again
there is no mystery -- the light rays travel different
distances at the same speed and therefore accumulate a phase
difference.

One could also choose any other inertial frame. Again the
conclusion is that the light rays travel different distances
at the same speed and therefore accumulate a phase difference.


> 3. If not, how, with the phase velocity being equal for both waves, can a
> phase difference accumulate?

See above. Like so many things, one can consider it in several
different ways, and one's description/explanation will be
different for different viewpoints. And as always, one must
use a single viewpoint (coordinate system) for the entire
analysis.

This latter seems to be the problem in your question, in
which you appear to switch between the co-rotating frame
and an inertial frame. You seem to (implicitly) think
that SR implies that the speed of light in the rotating
system will be isotropically c -- that is _NOT_ true.
Your "same path" is true only in the rotating system,
and you clearly mean same _spatial_ path (in spacetime
there is no way to consider the two rays as following
the same path -- they are helices of opposite twist
and different path length).


Tom Roberts tjro...@lucent.com

Robert Chan

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
On 07 Mar 2000 23:01:21 GMT, djm...@aol.com (Dennis McCarthy) wrote:
>>See http://mathpages.com/rr/s2-07/2-07.htm for an explanation
>>that includes the exact reason why the Sagnac effect is perfect
>>fodder for your sophistry. Note especially the last paragraph.
>Dennis: Unfortunately, here's the problem:
>
>"Still, the doubter imagines... isotropic with respect to
>non-inertial coordinates."
>
>That's called the "naive" SR explanation...

As usual, Dennis is quite wrong about this. I checked the referenced
web page, and the very next paragraph after the one Dennis quotes says

"There are, however, some interesting issues raised by
accelerating coordinate systems, but we'll defer discussion
of those until Section 4.5. For the moment, let's just confine
our attention to inertial coordinates, and show how a Sagnac
device appears in terms of the co-moving inertial frame of
an arbitrary point on the perimeter."

It then goes on to describe the Sagnac effect for a range of inertial
coordinates, and then in Section 4.5 discusses the same thing with
respect to accelerated (rotating) coordinates.

On 07 Mar 2000 23:01:21 GMT, djm...@aol.com (Dennis McCarthy) wrote:
> Allowing anisotropy wrt the non-inertial rim observer results
> in the Selleri paradox.

Just so no one is misled by this, Dennis is referring to a silly
mistake by Selleri that was thoroughly debunked in this newsgroup
a year or two ago. Check deja news. In summary, the Sagnac effect
remains proportional to the amount of ANGULAR travel, i.e., rotation,
during the transit of the light, even in the limit as the radius R
of the device goes to infinity.

On 07 Mar 2000 23:01:21 GMT, djm...@aol.com (Dennis McCarthy) wrote:
>If you want the correct SR explanation, check Rizzi and Tartaglia's.
>You'll find it's not so simple.

In terms of accelerating coordinates, very few physical phenomena
can be described simply - which is why we seldom use accelerating
coordinates when we can avoid it (as we certainly can in this case).
The explanation given by Rizzi and Tartaglia for accelerating
coordinates is similar to the explanation given in Section 4.5
of the web page referenced above.

>And you'll find that the explanation does in fact involve that

>"this path in some mysterious way becomes, in the course of


>the ring's rotation, different for the clock-wise and the

>anti-clockwise waves"--even for the rim observer.

Again, just so no one is misled, Dennis is completely incorrect in
his assessment. There is nothing mysterious about accelerating
coordinate systems, at least not to anyone who understands them.

Mark Samokhvalov

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to

Tom Roberts пишет в сообщении <38C5C242...@chicago.avenew.com> ...

>Mark Samokhvalov wrote:
>> In this group I recently found statements to the effect that the Sagnac
>> experiment was explained and even predicted (sic!) by SR.
>
>It is easily explained by SR. I do not think it was predicted by
>any SR advocate before Sagnac performed his experiment (but I am
>not an expert on such historical questions).
>
>
>> 1. Do you agree, that, if two waves radiated at the same point return to
>> that point with an accumulated phase difference, this means that there is
a
>> difference in either their:
>> a) paths, or
>> b) phase velocities?
>> Or, perhaps, you can suggest another possibility?
>
>There could be a difference in the gravitational potential
>integrated along their paths. There could be different optical
>media in their paths, or media moving differently (but are these
>differences in path?).

Yes, they are differences in optical paths.

>There are probably other possibilities..

Are there?


.
>
>
>> 2. When, as in Sagnac experiment, two waves are made to travel in
opposite
>> directions along the same path (typically, an optical filament), does

this


>> path in some mysterious way become, in the course of the ring's rotation,
>> different for the clock-wise and the anti-clockwise waves?
>
>This depends upon what coordinates (aka reference frame) you use to
>describe it. Note that both speed and spatial path length are
>inherently coordinate dependent.

That was my slip. I thought, those informed knew, the problem existed only
in the rotating RF. In a RF at rest wrt ether, SR = LET.


>
>Let me assume the center of rotation of the ring interferometer is
>at rest in an inertial frame. In this frame SR predicts the speed
>of light is isotropically c; in this frame it is clear that the
>co-rotating and counter-rotating light rays travel different
>distances. So there is no mystery -- the light rays travel
>different distances at the same speed and therefore accumulate a
>phase difference.


Exactly. Here SR = LET.

>One might also describe this in coordinates rotating with the
>ring interferometer. In these coordinates the _average_ speed
>of light _going_all_the_way_around_the_ring_ is different for
>the two directions, but the distance is the same for both rays.
>Again there is no mystery -- the light rays travel the same
>distance at different average speeds and therefore accumulate
>a phase difference.

You are a poor relativist: GR insists: in ALL RF the locally measured
velocity of light is isotropic and = c.


>
> Note that it is tricky to define "speed of light"
> in these coordinates, but there is indeed a well-defined
> _average_ speed of light _going_all_the_way_around_.
> This is intimitely related to the difficulty of
> synchronizing clocks in a rotating system....


There's no such problem in this case with the point of emission coinciding
with the point of reception.

The whole story reminds me of a Russian anecdote: Contributions were invited
to a conference on new methods in surgery, and one was submitted by an
Armenian team entitled "A New Method of Tonsilectomy". "Could anything be
new in that traditional method?" - "Ah, but you don't know, how we perform
it - we do it thtough the ass hole!" (Armenians in Russia are believed to
have homosexual preferemces.)

J.Bielawski

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to

Dennis McCarthy <djm...@aol.com> wrote in message
news:20000307180426...@ng-cb1.aol.com...

> >SR predicts the Sagnac effect quite easily, and here's a beautiful
> >derivation and explanation:
> >
> >http://mathpages.com/rr/s2-07/2-07.htm
> >
> >1a) yes, different path lengths
> >
> >2) no mystery needed, a rotating frame is non-inertial
>
> D: Sorry. 1) The path lengths aren't different to a person on the rim.

The previous poster was referring to the inertial lab system.

> 2) The non-inertial fall back has been rejected by mainstream physicists.

Nonsense.

> Check Rizzi and Tartaglia's explanation for the correct SR version.

They didn't investigate the matter fully because their main concern
was Selleri's paradox. In fact noninertial frames work here just fine.

Luc Bourhis

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
Mark Samokhvalov wrote:
>
> Sagnac & relativity
>
> In this group I recently found statements to the effect that the Sagnac
> experiment was explained and even predicted (sic!) by SR. I have perused
> many books on relativity dating back to 1920, including Einstein's, and saw
> no mention of Sagnac, until I came across a textbook by M.-A.Tonnella on
> magnetism and relativity ('55?), where a GR explanation of it was
> presented.

A very good reference which analyzes many aspects of the problem,
considering relativistic as well as non-relativistic Sagnac, is [1]. A
pedagogical treatment can also be found in [2] for light Sagnac. A
presentation of the experiments based on light can be found in [3]. For
the experiments using neutrons, see [5].

> I have the following questions to relativity fans:

> 1. Do you agree, that, if two waves radiated at the same point return to
> that point with an accumulated phase difference, this means that there is a
> difference in either their:
> a) paths, or
> b) phase velocities?
> Or, perhaps, you can suggest another possibility?

> 2. When, as in Sagnac experiment, two waves are made to travel in opposite
> directions along the same path (typically, an optical filament), does this
> path in some mysterious way become, in the course of the ring's rotation,
> different for the clock-wise and the anti-clockwise waves?

> 3. If not, how, with the phase velocity being equal for both waves, can a
> phase difference accumulate?

As many people, you consider that defining the length of co- and
counter-rotating light rays or the length of the rim itself is a trivial
question. It is not the case at all. Indeed the naive notion of length
require your being able to consider the whole rim or the whole rays at a
given instant. But it can be proved that it is impossible to realize
coherently such a "space + time splitting" because it is impossible to
devise an unambiguous global synchronisation procedure on the whole
ring. I will not give this demonstration because it is too complicated
to expose it in ASCII. The interesting reader is refered to [4] and
references therein.

However, just to give you a flavour of the issue, let's consider the
following a priori straightforward situation: two particles are sent
from some point A on the rim, in opposite directions, with the *same
speed V with respect to the rim*. According to a clock at A, SR predicts
that they do not come back at the same time ! This time delay T between
the two arrivals (which does not depend on V) is observed experimentally
in neutron experiments as a phase shift [5] between the wave functions
of these particles (the experimental configuration is not exactly the
one I described but it is close enough for my example to be relevant).

Now consider two slow moving clocks C+ and C-, respectively co- and
counter-rotating. If they are synchronized with a clock at point A, when
they come back to A, C+ and C- are respectively late and ahead with
respect to A and the difference between their reading is exactly the
above time delay T.

This shows clearly that there is much more than a speed anisotropy at
work in Sagnac experiments, even if an etherist might believe at first
that it is a possible explanation for Sagnac experiments based on light
propagation.


[1] J. Anandan, Phys. Rev. D, Vol. 24, N. 2, (1981) 338
[2] G. Rizzi, A. Tartaglia, gr-qc/9805089
[3] E.J. Post, Rev. Mod. Phys., Vol. 39, N. 2 (1967) 475
[4] V. Cantoni, Il Nuovo Cimento, Vol. 57 B, N. 1 (1968) 220
[5] S.A. Werner et al, Phys. Rev. A 21 (1980) 1419
--
Luc Bourhis

Center for Particle Theory
University of Durham, UK

Dennis McCarthy

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
>
>On 07 Mar 2000 23:01:21 GMT, djm...@aol.com (Dennis McCarthy) wrote:
>>>See http://mathpages.com/rr/s2-07/2-07.htm for an explanation
>>>that includes the exact reason why the Sagnac effect is perfect
>>>fodder for your sophistry. Note especially the last paragraph.
>>Dennis: Unfortunately, here's the problem:
>>
>>"Still, the doubter imagines... isotropic with respect to
>>non-inertial coordinates."
>>
>>That's called the "naive" SR explanation...
>
Rchan: >As usual, Dennis is quite wrong about this.

Dennis: Hmmm. First post to me and already sets the tone. Couldn't you have
been more polite?

Rchan: I checked the referenced


>web page, and the very next paragraph after the one Dennis quotes says
>
> "There are, however, some interesting issues raised by
> accelerating coordinate systems, but we'll defer discussion
> of those until Section 4.5. For the moment, let's just confine
> our attention to inertial coordinates,

Dennis: That paragraph I quoted does not confine its "attention to inertial
coordinates." Instead, it states "but


of course such coordinates are necessarily accelerating (viz., rotating), and

special relativity does not assume light speed to be isotropic with respect to
non-inertial coordinates."
Are you arguing that it is theoretically acceptable for the SR explanation to
assume that the speed of light is non-isotropic for the rim observer?


>On 07 Mar 2000 23:01:21 GMT, djm...@aol.com (Dennis McCarthy) wrote:
>> Allowing anisotropy wrt the non-inertial rim observer results
>> in the Selleri paradox.
>

Rchan: >Just so no one is misled by this, Dennis is referring to a silly


>mistake by Selleri that was thoroughly debunked in this newsgroup
>a year or two ago.

Dennis: Um, the only "mistake" of Selleri that was *alleged* by R&T was that
the paradox negates the SR view. However, R&T use the Selleri Paradox to
actually *refute* the naive SR position which states that the velocity of light
is anisotropic wrt the rim observer. Instead, they retrieve an explanation for
SR by *avoiding* the Selleri Paradox.

>On 07 Mar 2000 23:01:21 GMT, djm...@aol.com (Dennis McCarthy) wrote:
>>If you want the correct SR explanation, check Rizzi and Tartaglia's.
>>You'll find it's not so simple.
>
>In terms of accelerating coordinates, very few physical phenomena
>can be described simply

Dennis: Sagnac and Michelson did it in a few lines of algebra. According to
the rim observer, the speed of light is c+/-v. Not too hard.

D:>>And you'll find that the explanation does in fact involve that
>>"this path in some mysterious way becomes, in the course of


>>the ring's rotation, different for the clock-wise and the

>>anti-clockwise waves"--even for the rim observer.
>

Rchan: >Again, just so no one is misled, Dennis is completely incorrect in

>his assessment. There is nothing mysterious about accelerating
>coordinate systems, at least not to anyone who understands them.

Dennis: Are you trying to argue that R&T's position regarding the Sagnac effect
was *not* that the "path becomes, in the course of the ring's rotation,


different for the clock-wise and the

anti-clockwise waves"--even for the rim observer?


Dennis McCarthy


Dennis McCarthy

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
>> 2) The non-inertial fall back has been rejected by mainstream physicists.
>
JBiel: >Nonsense.

Dennis: Well, you know yourself that R&T do argue that the notion that the
speed of light is anisotropic wrt a rim observer is a problem for SR.

>> Check Rizzi and Tartaglia's explanation for the correct SR version.
>

JB: >They didn't investigate the matter fully because their main concern


>was Selleri's paradox. In fact noninertial frames work here just fine.

Dennis: R&T figured it out for the rim observer, i.e., a non-inertial frame.
And you are right that such frames "work just fine"--if you manage to deduce
the fact that the length of the path is different clockwise than
counter-clockwise wrt a rim observer. To me, that seems somewhat artificial.
I would rather trust the rulers of the rim observer, wouldn't you?


Dennis McCarthy


Dennis McCarthy

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
Roberts: > In these coordinates the _average_ speed
>of light _going_all_the_way_around_the_ring_ is different for
>the two directions, but the distance is the same for both rays.
>Again there is no mystery -- the light rays travel the same
>distance at different average speeds and therefore accumulate
>a phase difference.

Dennis: Interestingly, the average speed of light for the trip is not the
average of all the speeds for each small segment of the trip. Anyway, R&T
rejected such explanations via Selleri years ago.
Allegedly, according to the rim observer, the speed of light is still
anisotropic--despite what his rulers and clock reads.

Roberts: Note that it is tricky to define "speed of light"
> in these coordinates,

Dennis: Not for etherists.

but there is indeed a well-defined
> _average_ speed of light _going_all_the_way_around_.
> This is intimitely related to the difficulty of
> synchronizing clocks in a rotating system....

Dennis: No difficulty for etherists.


Dennis McCarthy


Tom Roberts

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
Mark Samokhvalov wrote:

> Tom Roberts wrote:
> >One might also describe this in coordinates rotating with the
> >ring interferometer. In these coordinates the _average_ speed
> >of light _going_all_the_way_around_the_ring_ is different for
> >the two directions, but the distance is the same for both rays.
> >Again there is no mystery -- the light rays travel the same
> >distance at different average speeds and therefore accumulate
> >a phase difference.
> GR insists: in ALL RF the locally measured
> velocity of light is isotropic and = c.

Yes, _LOCALLY_. That's why I explicitly said "_AVERAGE_ speed".
Going all the way around the ring in these rotating cooridnates is
not a local measurement, and there is no constraint on the speed
of light for such non-local measurements.


> > Note that it is tricky to define "speed of light"
> > in these coordinates, but there is indeed a well-defined
> > _average_ speed of light _going_all_the_way_around_.
> > This is intimitely related to the difficulty of
> > synchronizing clocks in a rotating system....
> There's no such problem in this case with the point of emission coinciding
> with the point of reception.

Yes. As I was at pains to point out, the _AVERAGE_ speed _GOING_
_ALL_THE_WAY_AROUND_ is well defined. This is so because one can
measure this average speed with a single clock, and the difficulties
of clock synchronization do not arise. But for any other situation
(other than average speed around a closed path) one does indeed need
to synchronize clocks to measure the speed of light, and those
difficulties arise....


Tom Roberts tjro...@lucent.com

Tom Roberts

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
Dennis McCarthy wrote:
> Dennis: Interestingly, the average speed of light for the trip is not the
> average of all the speeds for each small segment of the trip.

Yes. Because global properties can be different from local properties.
Note that the procedures used to measure the speed of light on those
individual small segments are _different_ from both each other and
from the procedure used to measure the average, so it's not too
surprising that the results are different: each segment must
synchronize clocks differently, but the average needs no clock
synchronization at all (it's a single-clock measurement).


> Anyway, R&T
> rejected such explanations via Selleri years ago.

Selleri made a different mistake -- he claimed (incorrectly) that
in the limit R->infinity his rotating example showed an anisotropic
local speed of light. His error was that to obtain this result he
used clocks synchronized in a differrent frame from the one in which
he measured the distance travelled. That's a common undergraduate
mistake....

To the mystified onlookers: "R&T" refers to Rizzi and
Tartaglia's preprints gr-qc/9805089 and gr-qc/9904028.
In essence they show that in the rotating system the sum of
the rulers laid around the rotating rim does not measure the
_spatial_ distance around the rim, it measures distance
along a helix which does not lie in any purely spatial
3-space; when one projects this onto a purely spatial
3-space, one obtains isotropy for the speed of light in the
rotating system.

This is directly related to my comments elsewhere that this
system is _ambiguous_ -- one can select many different but
individually self-consistent coordinates to describe this.
I discussed 4 different coordinates in my post, R&T use
still different coordinates....


> Allegedly, according to the rim observer, the speed of light is still
> anisotropic--despite what his rulers and clock reads.

When you insist on making such poorly-worded statements it's no
surprise that your statements contain internal contradictions, and
you get confused.

First, I suspect you intended to say "isotropic" rather than
"anisotropic" there. But it really doesn't matter. The thing you are
missing is the distinction between a local measurement and a non-local
measurement, and the different procedures used to make them in this
case.

One can take the sum of the rulers laid around the rotating rim and
divide it by the time taken for a light pulse to circumnavigate the
rim (measured on a single clock). This is clearly an _average_
mesurement. But as R&T showed, this is not really "distance"/"time",
this is really "distance plus an admixture of time"/"time", and
calling it "speed" is really a misnomer.


Tom Roberts tjro...@lucent.com

Tom Roberts

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to Luc Bourhis
Luc Bourhis wrote:
> [1] J. Anandan, Phys. Rev. D, Vol. 24, N. 2, (1981) 338
> [2] G. Rizzi, A. Tartaglia, gr-qc/9805089
> [3] E.J. Post, Rev. Mod. Phys., Vol. 39, N. 2 (1967) 475
> [4] V. Cantoni, Il Nuovo Cimento, Vol. 57 B, N. 1 (1968) 220
> [5] S.A. Werner et al, Phys. Rev. A 21 (1980) 1419

More recent experiments:
Anderson et al, Am. J. Phys. 62#11 (1994), p975.
A more recent review, and description of a much more accurate
ring interferometer.
Hasselbach and Nicklaus, Phys. Rev. A 48#1 (1993), p143.
The Sagnac effect using electrons.
Allan et al, Science, 228 (1985), p69.
They observed the Sagnac effect using GPS satellite signals
observed simultaneously at multiple locations around the world.

Other possibly interesting articles (theory) are:
Gron, "Relativistic description of a Rotating Disk", AJP 43#10 (1975),
p869.
Mainwaring and Stedman, "Accelerated Clock Principles in Special
Relativity:, Phys. Rev. A47#5 (1993), p3611.
Berenda, "The Problem of the Rotating Disk", Phys. Rev. 62 (1942), p280.
Ashtekar and Magnon, "The Sagnac Effect in General Relativity", J. Math.
Phys. 16#2 (1975), p341.

BTW Sagnac's original papers are:
Sagnac, C.R.A.S 157 (1913), p708, p1410; J. Phys. Radium, 5th Ser.
4 (1914), p177.


Tom Roberts tjro...@lucent.com

Dennis McCarthy

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
>
>Dennis McCarthy wrote:
>> Dennis: Interestingly, the average speed of light for the trip is not the
>> average of all the speeds for each small segment of the trip.
>
Roberts: >Yes. Because global properties can be different from local
properties.

Dennis: But what's astounding is that the global properties (the whole) does
not equal the sum of the local properties (the parts.) This, of course, is a
newly invented, complicated, counter-intuitive and bizarre principle--that must
be believed on faith--in order to maintain allegiance to SR.
The sound Sagnac explanation and the medium Sagnac explanation in general has
no such problems.

Roberts: >Note that the procedures used to measure the speed of light on those


>individual small segments are _different_ from both each other

Dennis: You are claiming that c as the local speed of light is the *reality* of
the underlying situation. The real speed of the local speeds of light does not
add up to average global speed--according to you.
If you are arguing that a measured local speed of c is just an artifact of the
method of measurement (ie, Einstein synch procedures) then we are agreed.

>> Anyway, R&T
>> rejected such explanations via Selleri years ago.
>

Roberts: >Selleri made a different mistake -- he claimed (incorrectly) that


>in the limit R->infinity his rotating example showed an anisotropic
>local speed of light.
>

> To the mystified onlookers: "R&T" refers to Rizzi and
> Tartaglia's preprints gr-qc/9805089 and gr-qc/9904028.
> In essence they show that in the rotating system the sum of
> the rulers laid around the rotating rim does not measure the
> _spatial_ distance around the rim, it measures distance
> along a helix which does not lie in any purely spatial
> 3-space; when one projects this onto a purely spatial
> 3-space, one obtains isotropy for the speed of light in the
> rotating system.

Dennis: Exactly. So the speed of light is isotropic for the rim
observer--despite what he measures. Hmm. And it appears some other sage just


wrote: "One might also describe this in coordinates rotating with the ring
interferometer. In these coordinates the _average_ speed of light
_going_all_the_way_around_the_ring_ is different for the two directions, but
the distance is the same for both rays.
Again there is no mystery -- the light rays travel the same distance at
different average speeds and therefore accumulate a phase difference."

So apparently there is a bit of a problem with this explantion.

Roberts:> This is directly related to my comments elsewhere that this
> system is _ambiguous_

Dennis: Not in ether theory.

Roberts: -- one can select many different but


> individually self-consistent coordinates to describe this.
> I discussed 4 different coordinates in my post, R&T use
> still different coordinates....

Dennis: And you get completely different results for what the speed of light is
wrt the rim observer. Such complications don't occur in ether theory.

D:>> Allegedly, according to the rim observer, the speed of light is still
>>[now corrected:] isotropic--despite what his rulers and clock reads.
>
Roberts:

>First, I suspect you intended to say "isotropic" rather than
>"anisotropic" there. But it really doesn't matter. The thing you are
>missing is the distinction between a local measurement and a non-local
>measurement,

Dennis: Sorry. R&T comes up with a global measurement of
isotropy--contradicting your claim above.

Roberts: and the different procedures used to make them in this

>case.
>
>One can take the sum of the rulers laid around the rotating rim and
>divide it by the time taken for a light pulse to circumnavigate the
>rim (measured on a single clock). This is clearly an _average_
>mesurement. But as R&T showed, this is not really "distance"/"time",
>this is really "distance plus an admixture of time"/"time", and
>calling it "speed" is really a misnomer.

Dennis: LOL. Ahh, so what your stationary rulers read is not necessarily
distance. (Forget your rulers we have this theory you see.) Anyway, let me
again quote a sage who just recently wrote something very different. Notice
his use of the word "speed" for "distance/time" as measured by the rim
observer:

Roberts wrote: "One might also describe this in coordinates rotating with the
ring interferometer. In these coordinates the _average_ speed of light
_going_all_the_way_around_the_ring_ is different for the two directions, but
the distance is the same for both rays.
Again there is no mystery -- the light rays travel the same distance at
different average speeds and therefore accumulate
a phase difference."

That seems pretty clear: "the light rays travel the same distance at different
average speeds." You appeared to be using a "misnomer" there.


Dennis McCarthy


Dennis McCarthy

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
Bourhis:
>As many people, you consider that defining the length of co- and
>counter-rotating light rays or the length of the rim itself is a trivial
>question.

Dennis: By "A trivial question," Bourhis means using stationary rulers to
measure a length of a stationary object. Bourhis feels you can't do that.
By "many people", Bourhis means all engineers, scientists, and physicsts--who
aren't addressing the Sagnac question--*ever* who measure length in the normal
way. In fact, this was the method of measuring length for Sagnac which
**Bourhis himself agreed to** within the last few weeks--before he researched
the subject--because it's somewhat embarrassing to deny it. And in fact, I defy
Bourhis to find any experiment in the history of science where length was
determined in the way he advocates. This new anti--empirical method has been
trotted out simply for the Sagnac effect--and that's it. And its reason is to
save the special theory of relativity. It is a length that can only be
determined based on theoretical assumptions, i.e., by putting pen to paper--and
it is a length that *contradicts* what is actually observed by real
experimenters using real stationary rulers. "Forget what you measure, we have
this theory, you see. So the length around a table is longer counter-clockwise
than clockwise. I know that's not what you measure, but trust us." Geez,
that's not an extra-complicating factor, is it?

B: It is not the case at all. Indeed the naive notion of length


>require your being able to consider the whole rim or the whole rays at a
>given instant. But it can be proved that it is impossible to realize
>coherently such a "space + time splitting" because it is impossible to
>devise an unambiguous global synchronisation procedure on the whole
>ring.

Dennis: That's not only goofy but flat out false. All earth clocks and and all
satellite clocks are absolutely synchronized wrt each other in the GPS. And
the Earth and satellites are rotating. If they weren't synched, planes would
not land as safely as they do.


Bourhis:>However, just to give you a flavour of the issue, let's consider the


>following a priori straightforward situation: two particles are sent
>from some point A on the rim, in opposite directions, with the *same
>speed V with respect to the rim*. According to a clock at A, SR predicts
>that they do not come back at the same time ! This time delay T between
>the two arrivals (which does not depend on V) is observed experimentally
>in neutron experiments as a phase shift [5] between the wave functions
>of these particles (the experimental configuration is not exactly the
>one I described but it is close enough for my example to be relevant).
>
>Now consider two slow moving clocks C+ and C-, respectively co- and
>counter-rotating. If they are synchronized with a clock at point A, when
>they come back to A, C+ and C- are respectively late and ahead with
>respect to A and the difference between their reading is exactly the
>above time delay T.
>
>This shows clearly that there is much more than a speed anisotropy at
>work in Sagnac experiments,

Dennis: How on Earth does it show that? Do you still not understand that the
same effect occurs for sound--and for sound clocks? There's a sound Sagnac
effect--and if you send two sound clocks around the rim--"moving clocks C+ and


C-, respectively co- and
counter-rotating. If they are synchronized with a clock at point A, when
they come back to A, C+ and C- are respectively late and ahead with
respect to A and the difference between their reading is exactly the

above time delay T"???? All with sound. Does that also "show clearly that
there is much more than a speed anisotropy at work in [sound] Sagnac
experiments"?
Tell me Bourhis how do you measure length in the sound Sagnac experiment? Same
way? What's your explanation for sound Sagnac? Or do you use a different one?

Dennis McCarthy


pmb

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
There is also a couple of papers by Robert Klauber on rotating systems which
mention the Sagnc effect

It was Early 1999 in the Am. J. Phys. as I recall

See references therein too

Pete

Tom Roberts <tjro...@lucent.com> wrote in message
news:38C68429...@lucent.com...

Mark Samokhvalov

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to

Luc Bourhis пишет в сообщении <38C63528...@durham.ac.uk> ...
this

>> path in some mysterious way become, in the course of the ring's rotation,
>> different for the clock-wise and the anti-clockwise waves?
>> 3. If not, how, with the phase velocity being equal for both waves, can a
>> phase difference accumulate?
>
>As many people, you consider that defining the length of co- and
>counter-rotating light rays or the length of the rim itself is a trivial
>question. It is not the case at all. Indeed the naive notion of length

>require your being able to consider the whole rim or the whole rays at a
>given instant. But it can be proved that it is impossible to realize
>coherently such a "space + time splitting" because it is impossible to
>devise an unambiguous global synchronisation procedure on the whole
>ring. I will not give this demonstration because it is too complicated
>to expose it in ASCII. The interesting reader is refered to [4] and
>references therein.

An excellent example ot high-brow relativistic blabber just to save the,
apparently, initially false constant c assumption. There are no problems in
defining lengths, if it's the same for both rays, if not, there's again no
need to define them - just to cite a PHYSICAL cause which makes two lengths
out of one when the ring rotates wrt stars, and certainly no sync problems.
For an etherist, the cause is simple and one for all RFs - it's constant
speed of light in ether at rest wrt stars. A relativist shrouds the problem
in so many words and formulae, and still fails to come up with an answer.

>
>However, just to give you a flavour of the issue, let's consider the
>following a priori straightforward situation: two particles are sent
>from some point A on the rim, in opposite directions, with the *same
>speed V with respect to the rim*. According to a clock at A, SR predicts
>that they do not come back at the same time !

A clock has nothing to do here - a coincidence indicator does the job. What
d'you mean by "the same speed"? D'you postulate it? If so, it's exactly the
same unanswered question (2).

This time delay T between
>the two arrivals (which does not depend on V) is observed experimentally
>in neutron experiments as a phase shift [5] between the wave functions
>of these particles (the experimental configuration is not exactly the
>one I described but it is close enough for my example to be relevant).

In experiment, postulating equal V's is not enough. You must make them
equal - most probably, it was assumed that opposite equal energy particles
have equal V's wrt to the rotating ring - a wrong assumption from the
etherist viewpoint because m(V) is wrt ether.


>
>Now consider two slow moving clocks C+ and C-, respectively co- and
>counter-rotating. If they are synchronized with a clock at point A, when
>they come back to A, C+ and C- are respectively late and ahead with
>respect to A and the difference between their reading is exactly the
>above time delay T.

This, again is not surprising, since time dilation in clocks moving wrt
ether is an experimental fact attributable to ether wind action on the
frequency standard. The ether approach would, in addition, expect a
dependence on the clock type - a clock with a directional frequency standard
(e.g., a piezoquartz resonator) could change its ticking rate with a change
in orientation. By the way, there's sufficient experimental evidence on the
ticking rate of satellite clocks to belie the equivalence pinciple - it's
different on different satellites, whereas conditions inside them should be
indistinguishable.


>
>This shows clearly that there is much more than a speed anisotropy at

>work in Sagnac experiments, even if an etherist might believe at first
>that it is a possible explanation for Sagnac experiments based on light
propagation.

By no means. For an etherist, there's nothing more than speed anisotropy in
Sagnac's experiment, but in other experiments, there are other ether wind
effects

[1] J. Anandan, Phys. Rev. D, Vol. 24, N. 2, (1981) 338
>[2] G. Rizzi, A. Tartaglia, gr-qc/9805089
>[3] E.J. Post, Rev. Mod. Phys., Vol. 39, N. 2 (1967) 475
>[4] V. Cantoni, Il Nuovo Cimento, Vol. 57 B, N. 1 (1968) 220
>[5] S.A. Werner et al, Phys. Rev. A 21 (1980) 1419

Luc Bourhis

unread,
Mar 8, 2000, 3:00:00 AM3/8/00
to
On Wed, 8 Mar 2000 17:49:06 +0000, Dennis McCarthy wrote
(in message <20000308124906...@ng-fm1.aol.com>):

>> This shows clearly that there is much more than a speed anisotropy at
>> work in Sagnac experiments,
>

> Dennis: How on Earth does it show that? Do you still not understand that the
> same effect occurs for sound--and for sound clocks?

What I have explained is treated extensively and precisely in the references
I have given. So it is now your turn: please show your math or at least some
references to them and please do not forget to explain the effects for the
two particles moving with the same speed wrt the rim: no speed anisotropy by
definition there.

--
Luc Bourhis
Center of Particle Physics/University of Durham


Luc Bourhis

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
On Wed, 8 Mar 2000 19:15:16 +0000, Mark Samokhvalov wrote
(in message <8a68pl$utl$1...@gavrilo.mtu.ru>):

>> As many people, you consider that defining the length of co- and
>> counter-rotating light rays or the length of the rim itself is a trivial
>> question. It is not the case at all. Indeed the naive notion of length
>> require your being able to consider the whole rim or the whole rays at a
>> given instant. But it can be proved that it is impossible to realize
>> coherently such a "space + time splitting" because it is impossible to
>> devise an unambiguous global synchronisation procedure on the whole
>> ring. I will not give this demonstration because it is too complicated
>> to expose it in ASCII. The interesting reader is refered to [4] and
>> references therein.
>
> An excellent example ot high-brow relativistic blabber just to save the,
> apparently, initially false constant c assumption. There are no problems in
> defining lengths, if it's the same for both rays, if not, there's again no
> need to define them - just to cite a PHYSICAL cause which makes two lengths
> out of one when the ring rotates wrt stars, and certainly no sync problems.
> For an etherist, the cause is simple and one for all RFs - it's constant
> speed of light in ether at rest wrt stars. A relativist shrouds the problem
> in so many words and formulae, and still fails to come up with an answer.

Before defining you as an etherist, you have to precise clearly what ether
theory you want to use to model a Sagnac experiment. Be careful because in
order to escape the difficulties encountered with SR, you have to avoid
LET-like theories which have Lorentz transforms embedded in them. But then
you are most surely in trouble with Michelson & Morley, Kennedy & Thorndike
or Doppler because an Ether theory compatible with these experiments is
equivalent to SR as long as we are concerned with observable predictions. So
I would say that this is the typical answer of an etherist on this forum:
taught about what is the situation in pedagogical terms and referred to
serious references to obtain a deeper knowledge, he comes with a ghost
rebutal not supported by any scientific argument. I am disappointed to see
that you have not better prepared your troll.

>> However, just to give you a flavour of the issue, let's consider the
>> following a priori straightforward situation: two particles are sent
>> from some point A on the rim, in opposite directions, with the *same
>> speed V with respect to the rim*. According to a clock at A, SR predicts
>> that they do not come back at the same time !
>
> A clock has nothing to do here

It could have if one was measuring the arrival times ....

> - a coincidence indicator does the job.

... but indeed one performs quantum experiment and observe the interferences
between the particles going in opposite directions. However the phase shift
is directly proportional to the time delay.

> What d'you mean by "the same speed"? D'you postulate it?

The same speed. I can not be clearer. It is imposed by the experimental
configuration. See below.

> If so, it's exactly the same unanswered question (2).

The 3 following objects have different length:
- the trajectory of a co-rotating particle
- the trajectory of a counter-rotating particle
- the rim
if one works in the non-inertial frame whose world lines are defined by the
following parametric equations in some inertial frame:
x0 = t
x1 = a cos(w t) - b sin(w t)
x2 = a sin(w t) + b cos(w t)
x3 = 0,
with (a^2+b^2)^(1/2) being the radius of the rim and w its rotation speed.

>> This time delay T between
>> the two arrivals (which does not depend on V) is observed experimentally
>> in neutron experiments as a phase shift [5] between the wave functions
>> of these particles (the experimental configuration is not exactly the
>> one I described but it is close enough for my example to be relevant).
>
> In experiment, postulating equal V's is not enough. You must make them
> equal - most probably, it was assumed that opposite equal energy particles
> have equal V's wrt to the rotating ring - a wrong assumption from the
> etherist viewpoint because m(V) is wrt ether.

Again which Ether theory ? Anyway for the non-relativistic particles used
here, de Broglie law relating the wave-length lambda, and therefore the
energy of the particles, their mass m and their speed v,
lambda = h/(m v)
is exceptionaly well verified. Then the two beams are obtained from one
incoming beam by a Bragg diffraction through a layer of silicon. The optical
nature of the phenomena guaranties that the wave-length of the two outgoing
beams is the same. Therefore the situation is the one I have described and
you have only forced me to explain the details related to my remark between
parentheses. Note that the same kind of experiments have been made with
electrons [7] (thanks to Tom Roberts for the reference).

So now how can you explain the experimental results ?

>> Now consider two slow moving clocks C+ and C-, respectively co- and
>> counter-rotating. If they are synchronized with a clock at point A, when
>> they come back to A, C+ and C- are respectively late and ahead with
>> respect to A and the difference between their reading is exactly the
>> above time delay T.
>
> This, again is not surprising, since time dilation in clocks moving wrt
> ether is an experimental fact attributable to ether wind action on the
> frequency standard.

What is tremendously interesting however is that one finds the same T for
this shift between clocks and for the delay between the arrival time of
particles rotating in opposite direction even if a priori there is no links
between these two configurations. If one remembers that the same geometry of
spacetime is at work, everything becomes clear and indeed the important point
is of course that SR gives the correct predictions for all these effects. And
Ether theories ?

> The ether approach would, in addition, expect a
> dependence on the clock type - a clock with a directional frequency standard
> (e.g., a piezoquartz resonator) could change its ticking rate with a change
> in orientation. By the way, there's sufficient experimental evidence on the
> ticking rate of satellite clocks to belie the equivalence pinciple - it's
> different on different satellites, whereas conditions inside them should be
> indistinguishable.

I do not understand : do you say the the equivalence principle can be
believed or that it is ruled out ? I do not know any experimental data
supporting the latter.

[4] V. Cantoni, Il Nuovo Cimento, Vol. 57 B, N. 1 (1968) 220
[5] S.A. Werner et al, Phys. Rev. A 21 (1980) 1419

[6] C.M. Will, gr-qc/9811036
[7] Allan et al, Science, 228 (1985), p69.

Dennis McCarthy

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
>
>On Wed, 8 Mar 2000 17:49:06 +0000, Dennis McCarthy wrote
>(in message <20000308124906...@ng-fm1.aol.com>):
>
>>> This shows clearly that there is much more than a speed anisotropy at
>>> work in Sagnac experiments,
>>
>> Dennis: How on Earth does it show that? Do you still not understand that
>the
>> same effect occurs for sound--and for sound clocks?
>
Bourhis: >What I have explained is treated extensively and precisely in the

references
>I have given. > So it is now your turn: please show your math or at least some

>references to them

Dennis: You have given no evidence for your claim above. That's just a strange
Bourhis conclusion.
Moreover, when you first came to this thread I had to show you the precise
mathematical description of sound Sagnac, ether Sagnac and sound clocks. It's
an obvious physical fact that no one denies--and takes three lines of algebra.
Also, it took me two weeks of getting past all of your ridiculous
objections--but you finally understood it (sound and ether Sagnac.) I'm not
going through it again. If you don't believe sound Sagnac occurs or just want
to deny obvious physical facts that you find inconvenient and irritating, then
this discussion is not worth my time and you are not on an objective search for
the truth.
Also, you gave references for the interesting argument that the length of a
rim is not the same CW as CCW for a rim observer. Very strange to say the
least.
You haven't replied to the following:

Bourhis:

>As many people, you consider that defining the length of co- and
>counter-rotating light rays or the length of the rim itself is a trivial
>question.

Dennis: By "A trivial question," Bourhis means using stationary rulers to
measure a length of a stationary object. Bourhis feels you can't do that.
By "many people", Bourhis means all engineers, scientists, and physicsts--who
aren't addressing the Sagnac question--*ever* who measure length in the normal
way. In fact, this was the method of measuring length for Sagnac which
**Bourhis himself agreed to** within the last few weeks--before he researched
the subject--because it's somewhat embarrassing to deny it. And in fact, I defy
Bourhis to find any experiment in the history of science where length was
determined in the way he advocates. This new anti--empirical method has been
trotted out simply for the Sagnac effect--and that's it. And its reason is to
save the special theory of relativity. It is a length that can only be
determined based on theoretical assumptions, i.e., by putting pen to paper--and
it is a length that *contradicts* what is actually observed by real
experimenters using real stationary rulers. "Forget what you measure, we have
this theory, you see. So the length around a table is longer counter-clockwise
than clockwise. I know that's not what you measure, but trust us." Geez,
that's not an extra-complicating factor, is it?

Dennis McCarthy


Dennis McCarthy

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
>
>On Wed, 8 Mar 2000 19:15:16 +0000, Mark Samokhvalov wrote
>(in message <8a68pl$utl$1...@gavrilo.mtu.ru>):
>
>>> As many people, you consider that defining the length of co- and
>>> counter-rotating light rays or the length of the rim itself is a trivial
>>> question. It is not the case at all. Indeed the naive notion of length
>>> require your being able to consider the whole rim or the whole rays at a
>>> given instant. But it can be proved that it is impossible to realize
>>> coherently such a "space + time splitting" because it is impossible to
>>> devise an unambiguous global synchronisation procedure on the whole
>>> ring. I will not give this demonstration because it is too complicated
>>> to expose it in ASCII. The interesting reader is refered to [4] and
>>> references therein.
>>
>> An excellent example ot high-brow relativistic blabber just to save the,
>> apparently, initially false constant c assumption. There are no problems in
>> defining lengths, if it's the same for both rays, if not, there's again no
>> need to define them - just to cite a PHYSICAL cause which makes two lengths
>> out of one when the ring rotates wrt stars, and certainly no sync problems.
>> For an etherist, the cause is simple and one for all RFs - it's constant
>> speed of light in ether at rest wrt stars. A relativist shrouds the problem
>> in so many words and formulae, and still fails to come up with an answer.
>
B:>Before defining you as an etherist, you have to precise clearly what ether
>theory you want to use to model a Sagnac experiment. Be careful because in
>order to escape the difficulties encountered with SR, you have to avoid
>LET-like theories which have Lorentz transforms embedded in them.

Dennis: You are coming close to outright dishonesty. The only person in the
world who follows the type of LET that you use to confuse situations and
complicate Sagnac is you. All other known etherists (Michelson, Sagnac, Ives,
etc., and those on these boards and all whom I know personally), no matter what
the ether theory, derive the Sagnac effect quite simply with the Galilean
analysis.
Now, unfortunately, the following involves physical reasoning, but I'll still
try. In the Sagnac effect, the Lorentz contraction can be disregarded by
Lorentzians because there is only one rim and so LC has an equal effect on both
path lengths. Also, no clock movements or synch procedures need to be made.
Thus, the Galilean nature of light is clearly recovered in the
experiment--because there are no relevant deformations of the measuring devices
to counteract this Galilean nature. And **all** etherists whom I know believe
in the underlying Galilean nature of light. Lorentzians apply the Lorentzian
corrections only when measuring devices are altered in a relevant way. If not,
then the Galilean nature is recovered, and as you know, I hope, it takes 3
lines of algebra to produce Sagnac.

B: But then

>you are most surely in trouble with Michelson & Morley, Kennedy & Thorndike
>or Doppler because an Ether theory compatible with these experiments is
>equivalent to SR as long as we are concerned with observable predictions.

Dennis: That's false. Obviously. IGS theory, for example, fits the known
experiments. And Lorentz contraction does not need to be assumed.

B:So

>I would say that this is the typical answer of an etherist on this forum:
>taught about what is the situation in pedagogical terms and referred to
>serious references to obtain a deeper knowledge, he comes with a ghost
>rebutal not supported by any scientific argument.

Dennis: The etherist argument was put forth by Sagnac in 1913 in the original
paper on the subject. The Galilean analysis that predicts the result came
centuries before that. It is unfortunate that you still pretend to be unaware
of it.

Dennis McCarthy


Standeven

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to

Mark Samokhvalov wrote:

> Luc Bourhis пишет в сообщении <38C63528...@durham.ac.uk> ...
> >

> >However, just to give you a flavour of the issue, let's consider the
> >following a priori straightforward situation: two particles are sent
> >from some point A on the rim, in opposite directions, with the *same
> >speed V with respect to the rim*. According to a clock at A, SR predicts
> >that they do not come back at the same time !
>
> A clock has nothing to do here - a coincidence indicator does the job. What
> d'you mean by "the same speed"? D'you postulate it? If so, it's exactly the
> same unanswered question (2).

No, our observers measure it. They, for some strange reason, are under the
impression that the speeds they measure with their clocks and rulers are
the actual speeds in their frame.

I suppose _you're_ going to say that their measurements are invalid for
some reason now, aren't you...


Robert Chan

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
On 08 Mar 2000 djm...@aol.com (Dennis McCarthy) wrote:
>That paragraph I quoted does not confine its "attention to
>inertial coordinates." Instead, it states "but of course such
>coordinates are necessarily accelerating (viz., rotating), and
>special relativity does not assume light speed to be isotropic
>with respect to non-inertial coordinates."

That web page mentions analyzing the Sagnac device in terms of
coordinates rigidly fixed to the ring, notes that these are
necessarily accelerating coordinates, notes that lightspeed need
not be c with respect to such coordinates, comments on how one
might imagine that this leads to a contradiction, since it IS
possible to define a coordinate system "in terms of which the
position of a point fixed on the disk is independent of time", and
then defers further discussion of accelerated coordinates until
later, saying "For the moment, let's just confine our attention
to inertial coordinates...", following which appears an analysis
of the Sagnac device in terms of the co-moving inertial coordinates
of a fixed point on the ring. In the subsequent sections it
describes the helical locus of local simultaneity implied by the
the sequence of co-moving inertial coordinates, and then gives an
account in terms of a single accelerating coordinate system and
the resulting pseudo-gravitational potentials. In summary, you
misread, misunderstood, and misrepresented the reference (IMHO).

On 08 Mar 2000 djm...@aol.com (Dennis McCarthy) wrote:
>Are you arguing that it is theoretically acceptable for the

>SR explanation to assume that the speed of light is non-
>isotropic for the rim observer?

Fundamentally the Sagnac effect from an SR perspective is due
to the difference in optical path lengths, but it's possible to
define spatial and temporal parameters in terms of which this
optical path lengh difference can be expressed as an equivalent
anisotropy in the "speed of light" at various points around
the rim. Of course, this involves defining "speed" in terms
of distances and times that are not components of a single
inertial coordinate system. The reference web page describes
in detail one way of doing this.

You have to be careful not to confuse the different approaches
that can be taken to describe the operation of a Sagnac device.
First, since the Sagnac effect is not essentially relativistic,
it's possible to get the right answer by taking a Galilean approach.
Second, there is the drop-dead simple SR method in terms of a single
system of inertial coordinates, which is how anyone in their right
mind approaches the problem. Needless to say, the speed of light
is isotropic with respect to this parameterization.

Now, if we get bored doing things the sensible way, there are (at
least) two ways that we can describe "how this looks" in terms of
coordinates in which an observer fixed on the rim is stationary.
The first way is to use momentarily co-moving inertial coordinates,
which vary as the disk rotates. This is still an "SR explanation"
in the sense that no pseudo-gravitational potentials are invoked,
but it's really a composite view over a sequence of coordinate
systems, each of which is inertial in itself, but the union of
which does not constitute a single inertial frame. Instead it
yields a helical locus of simultaneity, and of course a difference
in the effective path lengths in the two directions around the
rim. In this case, as mentioned above, it's possible to define
space and time parameters in such a way that we can express the
difference in path lengths as an equivalent difference in light
"speed", on the understanding that this speed is not defined in
terms of components of an inertial coordinate system. (There's
not much reason to actually DO this, other than to show that
it's possible.)

Another approach is to use a single accelerating coordinate
system, with respect to which a given point on the rim is
*permanently* stationary. This requires the introduction of
pseudo-gravitational potentials to account for the accelerations
to which the various points on the rim are subjected with respect
to these accelerated coordinates. Some people would call this
a "GR" approach, but since the spacetime is still flat, this is
debatable. In any case, the key here is to properly account
for ALL the accelerations, including the Coriolis acceleration,
which turns out to be the dominant factor in establishing the
pseudo-gravitational potential corresponding to the differential
progress of light around the rim in the two directions, because
the Coriolis effect is different for the two directions.

This is all standard stuff, and has been explained many times
before. For example, there's a nice overview in Ciufolini and
Wheeler's "Gravitation and Inertia".

On 08 Mar 2000 djm...@aol.com (Dennis McCarthy) wrote:
>...the only "mistake" of Selleri that was *alleged* by R&T was

>that the paradox negates the SR view.

Selleri claimed (in a self-published note) that Sagnac was
inconsistent with special relativity. Rizzi and Tartaglia
published a rebutal, pointing out that Selleri was mistaken (as
if this wasn't obvious), and they highlighted one particular
flaw in Selleri's reasoning, sufficient to invalidate his claim.
However, R&T did not detect (or at least didn't mention) ALL of
the errors in Selleri's reasoning. (This would have made for a
MUCH longer paper.) They focused only on the most interesting
error in his reasoning, one that has been committed by other
people as well, namely, the assumption that we can always define
a single-valued foliation of simultaneity for an accelerating
observer. Of course, this assumption is false, and has always
been known to be false, i.e., this was not a discovery of R&T.
They simply published their note to rebut Selleri and as an
informational service to highlight an interesting aspect of
special relativity.

One of Selleri's errors that R&T failed to mention in their
little rebutal note was his simple failure to correctly evaluate
the "anisotropic speed ratio" in the limit of increasing radius,
even according to his own (flawed) premises. He asserted that
this ratio remains constant as the radius increases to infinity,
leading to (he claimed) a conflict with special relativity in
the limit of *linear* light paths. The correct analysis shows
that the deviation from 1 of the "anisotropic ratio" is always
exactly proportional to the *angular* travel of the device
during the transit of light, and this remains true in the
limit as the radius increases to infinity.

On 08 Mar 2000 djm...@aol.com (Dennis McCarthy) wrote:
>RChan: Again, just so no one is misled, Dennis is completely

>incorrect in his assessment. There is nothing mysterious about
>accelerating coordinate systems, at least not to anyone who
>understands them.
>
>Dennis: Are you trying to argue that R&T's position regarding

>the Sagnac effect was *not* that the "path becomes, in the

>course of the ring's rotation, different for the clock-wise

>and the anti-clockwise waves"--even for the rim observer?

Why on Earth would I be trying to argue that? Let's see now,
you're saying that R&T are saying that the path lengths are
different due to the absolute rotation of the ring. Gosh,
what an insight. How mysterious! Actually, that's what I
(and everyone else, as far as I can tell) have been saying
all along, including the reference web page, which says "Notice
that this anisotropy ... for any inertial frame is actually
in the distance travelled, not the speed of travel", and then
goes on to describe the helical locus of simultaneity, etc.
What I WAS arguing is that "There is nothing mysterious about

Mark Samokhvalov

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to

Luc Bourhis пишет в сообщении
<01HW.B4ECA87F0...@news.freeserve.net> ...

To my knowledge, ALL experiments performed to detect the effect of the
Earth's orbital velocity on physical processes, including my own to detect
the anisotropy of electrophysical parameters induced by it, have produced
nil results. Together with the existence of star aberration, this could be
explained by ether drag by the Earth's gravitational field and verified by
any of the aforementioned experiments on board satellites, or on the Earth
by such experiments sensitive enough to detect the effects of Earth's
rotation velocity. Perhaps, such experiments have already been performed -
any information?


Clearly, a case when sollipsism comes in conflict with reality - different
results obtained by reasoning with no physical cause. SR was originally
constructed not as a physical, but as a metric theory - had another signal
carrier (e.g., fast particles) been used, the results would have been
different. Had different signal carriers been used jointly, all ambiguity
(=relativity), would have been removed.
Indeed, during the XX-th century triumphs of theoretical physics elated the
intellectual elite aleready influenced by sollipsist philosophy to the
extent that they thought they could construct reality. It the same with
quantum physics: all sorts of weird hypotheses have been advanced to explain
the interference of isolated quanta instead of aknowledging the simple fact
of their division and propagation in the form of sub-quantum oscillations.


>>> This time delay T between
>>> the two arrivals (which does not depend on V) is observed experimentally
>>> in neutron experiments as a phase shift [5] between the wave functions
>>> of these particles (the experimental configuration is not exactly the
>>> one I described but it is close enough for my example to be relevant).
>>
>> In experiment, postulating equal V's is not enough. You must make them
>> equal - most probably, it was assumed that opposite equal energy
particles
>> have equal V's wrt to the rotating ring - a wrong assumption from the
>> etherist viewpoint because m(V) is wrt ether.
>
>Again which Ether theory ? Anyway for the non-relativistic particles used
>here, de Broglie law relating the wave-length lambda, and therefore the
>energy of the particles, their mass m and their speed v,
> lambda = h/(m v) is exceptionaly well verified. Then the two beams are
obtained from one
>incoming beam by a Bragg diffraction through a layer of silicon. The
optical
>nature of the phenomena guaranties that the wave-length of the two outgoing
>beams is the same. Therefore the situation is the one I have described and
>you have only forced me to explain the details related to my remark between
>parentheses. Note that the same kind of experiments have been made with
>electrons [7] (thanks to Tom Roberts for the reference).
>
>So now how can you explain the experimental results ?

>The experimental fact of inertial mass vs V dependence explained/predicted
by several theories, includung SR. With the ether concept, V is measured wrt
ether, and therefore in the ring's RF it will be different for co- and
counter-rotating equal energy particles.

>>> Now consider two slow moving clocks C+ and C-, respectively co- and
>>> counter-rotating. If they are synchronized with a clock at point A, when
>>> they come back to A, C+ and C- are respectively late and ahead with
>>> respect to A and the difference between their reading is exactly the
>>> above time delay T.
>>
>> This, again is not surprising, since time dilation in clocks moving wrt
>> ether is an experimental fact attributable to ether wind action on the
>> frequency standard.
>
>What is tremendously interesting however is that one finds the same T for
>this shift between clocks and for the delay between the arrival time of
>particles rotating in opposite direction even if a priori there is no links
>between these two configurations. If one remembers that the same geometry
of
>spacetime is at work, everything becomes clear and indeed the important
point
>is of course that SR gives the correct predictions for all these effects.
And
>Ether theories ?

No space-time needed. Suffice it that both effects are described by small
terms quadratic in V. Integration over travel time produces the result.


>
>> The ether approach would, in addition, expect a
>> dependence on the clock type - a clock with a directional frequency
standard
>> (e.g., a piezoquartz resonator) could change its ticking rate with a
change
>> in orientation. By the way, there's sufficient experimental evidence on
the
>> ticking rate of satellite clocks to belie the equivalence pinciple - it's
>> different on different satellites, whereas conditions inside them should
be
>> indistinguishable.
>
>I do not understand : do you say the the equivalence principle can be
>believed or that it is ruled out ? I do not know any experimental data
>supporting the latter.
>

I mean, it doesn't work, because the clocks on satellites at different
altitudes tick differently.

Luc Bourhis

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
Dennis McCarthy wrote:
>
> >
> >On Wed, 8 Mar 2000 17:49:06 +0000, Dennis McCarthy wrote
> >(in message <20000308124906...@ng-fm1.aol.com>):
> >
> >>> This shows clearly that there is much more than a speed anisotropy at
> >>> work in Sagnac experiments,
> >>
> >> Dennis: How on Earth does it show that? Do you still not understand that the
> >> same effect occurs for sound--and for sound clocks?
> >
> >What I have explained is treated extensively and precisely in the references
> >I have given. So it is now your turn: please show your math or at least some
> >references to them
>
> Dennis: You have given no evidence for your claim above. That's just a strange
> Bourhis conclusion.

For the moment I have not seen either Tom Roberts or Steve Carlip
contradicting me. Usually they do not hesitate to point out errors in
whatever messages they are made. Do you think you know better Special
and General Relativity than all of us ?

> Moreover, when you first came to this thread I had to show you the precise
> mathematical description of sound Sagnac, ether Sagnac and sound clocks. It's
> an obvious physical fact that no one denies--and takes three lines of algebra.

For Sagnac experiment, you have shown only a mere Galilean analysis
which does give the correct result for light only because terms in
(v/c)^2 are experimentally not observable. However this same kind of
analysis predicts no delay between the comeback of two particles sent
simultaneously in opposite direction with the same speed with respect to
the rim. SR predicts on the contrary a delay and this effects is
observed for neutron and electrons [1,2]. So how do you reconciliate
your point of view with experimental results ? Surely not with clocks.

Moreover your sound clocks are the analogs of Einstein light clocks :
waves bouncing between two walls. How do you manage to model mechanical
clocks, atomic clocks, molecular clocks, instable particles decaying
through weak interaction, instable particles decaying through strong
interaction with such a kindergarten *non-quantum* model ? Is it that
you have to postulate that this simple analogy can be extended to treat
any kind of periodical phenomena ? But are you not using the Postulate
of Relativity then ? If fact you have advocated a long time ago what you
called PLET (Poincarre LET), i.e. LET + PoR. Well it was before your
infatuation for IGS.

> Also, you gave references for the interesting argument that the length of a
> rim is not the same CW as CCW for a rim observer. Very strange to say the
> least.

Strange is not by itself a scientific comment. Furthermore you are
mistaken. Co- and counter-rotating light rays do not have the same
lengths and these lengths are different from the one of the rim.
Precisely (I have to be careful: Tom and Steve are around in this
thread) this is true in the non-inertial frame whose world lines are
paramametrized by:
x0 = t
x1 = R cos(w t)
x2 = R sin(w t)
x3 = 0
where R is the radius of the rim and w its angular speed (as you see
this frame is intuitively the rim). This comes from the fact that
neither the rim nor the light rays are well defined geometrical objects
entirely embedded in space at a given instant.

This is a definitive result of SR and/or GR, and of course of LET -- the
theory exposed by Lorentz in 1904 can indeed be extended to deal with
non-inertial frame as SR is and this results in the same framework since
SR and LET are equivalent for inertial motions.

> You haven't replied to the following:

Sun-Tzu said : "When involved in a guerilla, launch first the attacks
which do not cost you anything".

> >As many people, you consider that defining the length of co- and
> >counter-rotating light rays or the length of the rim itself is a trivial
> >question.
>

> Dennis: By "A trivial question," Bourhis means using stationary rulers to
> measure a length of a stationary object. Bourhis feels you can't do that.
> By "many people", Bourhis means all engineers, scientists, and physicsts--who
> aren't addressing the Sagnac question--*ever* who measure length in the normal
> way.

That is to say in situations where relativistic effects are completely
negligeable. That is to say by working with a Galilean framework which
was indeed a good approximation.

> In fact, this was the method of measuring length for Sagnac which
> **Bourhis himself agreed to** within the last few weeks--before he researched
> the subject--because it's somewhat embarrassing to deny it. And in fact, I defy
> Bourhis to find any experiment in the history of science where length was
> determined in the way he advocates.

The definition of length used in [4] and precisely explained in [3] is
used routinely in GR for a very long time. Moller [5] described it in a
book published in the 60's and it is therefore even older. It is a very
general definition because it is equivalent locally to the kind of
intuitive method you advocate which uses rulers -- semi-precisely, it
transforms the measurement of the length of a 1-dimensional body in a
frame R into an infinite succession of length measurements at successive
points P by inertial observers moving with the same speed as R at this
point P. So this means naively that for infinitesimal lengths we are
back to our usual rulers.

Difficulties do only occur when one considers global properties, the
whole rim for example. These difficulties are seen only for a particular
class of non-inertial frames (see [5] p. 248 for a complete analysis)
and rotating frames are in this class. And yes this is a new effect
predicted by Special Relativity and LET but not by Galilean Relativity.
But how is it possible to be surprised by this fact ?

> This new anti--empirical method has been
> trotted out simply for the Sagnac effect--and that's it. And its reason is to
> save the special theory of relativity.

It is blatantly wrong as explained above. Before criticizing modern
physic one is supposed to learn it. You know you can not get this
knowledge solely from this newsgroup. Contributors like me do not have
the time or the energy to give the best pedagogical explanations and we
have to be often too concise or sketchy.

> It is a length that can only be
> determined based on theoretical assumptions, i.e., by putting pen to paper--and
> it is a length that *contradicts* what is actually observed by real
> experimenters using real stationary rulers.

You are again completely confused. An observer moving infinitely slowly
along the rim will measure the perimeter:
L = 2 pi R/(1-(R Omega)^2)^(1/2)
in units where c=1. But this is not the length of a light ray
propagating along the rim in the non-inertial frame I have defined
precisely above. Now how do you propose to measure with rulers the
length of a light ray ?


[1] S.A. Werner et al, Phys. Rev. A 21 (1980) 1419
[2] Hasselbach and Nicklaus, Phys. Rev. A 48#1 (1993), p143.
[3] J. Anandan, Phys. Rev. D, Vol. 24, N. 2, (1981) 338
[4] G. Rizzi, A. Tartaglia, gr-qc/9805089
[5] C. Moller, The Theory of Relativity (Oxford 1952)

Dennis McCarthy

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
>
>>Dennis: Are you trying to argue that R&T's position regarding
>>the Sagnac effect was *not* that the "path becomes, in the
>>course of the ring's rotation, different for the clock-wise
>>and the anti-clockwise waves"--even for the rim observer?
>
Chan: Why on Earth would I be trying to argue that? Let's see now,

>you're saying that R&T are saying that the path lengths are
>different due to the absolute rotation of the ring. Gosh,
>what an insight. How mysterious! Actually, that's what I
>(and everyone else, as far as I can tell) have been saying
>all along, including the reference web page, which says "Notice
>that this anisotropy ... for any inertial frame is actually
>in the distance travelled, not the speed of travel",

Dennis: Perhaps, you should read the paper again? R&T are arguing the distance
is different according to the frame **for the rim observer**--for the frame of
the rim. That means the desk your computer is on would be longer CCW than
CW--according to *you*--who is stationary wrt the desk despite what you
measure.
And apparently that result is "mysterious" --as you didn't understand it.
I have snipped your other mistakes as they all relate to that point.



Dennis McCarthy


Luc Bourhis

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
Dennis McCarthy wrote:
>
> >
> >On Wed, 8 Mar 2000 19:15:16 +0000, Mark Samokhvalov wrote
> >(in message <8a68pl$utl$1...@gavrilo.mtu.ru>):
> >
> >Before defining you as an etherist, you have to precise clearly what ether
> >theory you want to use to model a Sagnac experiment. Be careful because in
> >order to escape the difficulties encountered with SR, you have to avoid
> >LET-like theories which have Lorentz transforms embedded in them.
>
> Dennis: You are coming close to outright dishonesty. The only person in the
> world who follows the type of LET that you use to confuse situations and
> complicate Sagnac is you. All other known etherists (Michelson, Sagnac, Ives,
> etc., and those on these boards and all whom I know personally), no matter what
> the ether theory, derive the Sagnac effect quite simply with the Galilean
> analysis.

And you are then using incoherent ad hoc solution ruled out
experimentaly if you advocate a mere Galilean analysis of the Sagnac
effect. Indeed,
1) Are you really ready to promote LET for Michelson & Morley, Kennedy &
Thorndike, Trouton & Noble, ..... while you advocate a Galilean analysis
as the fundamental theory for Sagnac ?
2) Or is it that a Galilean analysis is a good approximation of LET when
it comes to compare the theoretical predictions with the experimental
results of light Sagnac experiments ?
3) But then what about the experiments made with massive particles for
which this Galilean analysis is completely wrong ? How do you treat this
case ?

> Now, unfortunately, the following involves physical reasoning, but I'll still
> try. In the Sagnac effect, the Lorentz contraction can be disregarded by
> Lorentzians because there is only one rim and so LC has an equal effect on both
> path lengths. Also, no clock movements or synch procedures need to be made.
> Thus, the Galilean nature of light is clearly recovered in the
> experiment--because there are no relevant deformations of the measuring devices
> to counteract this Galilean nature. And **all** etherists whom I know believe
> in the underlying Galilean nature of light.

Obviously not Lorentz.

> Lorentzians apply the Lorentzian
> corrections only when measuring devices are altered in a relevant way. If not,
> then the Galilean nature is recovered, and as you know, I hope, it takes 3
> lines of algebra to produce Sagnac.

But you end up with a theory which has to treat differently light
propagation, motion of clocks and motion of massive particles around the
rim. On the contrary SR treats all these effects exactly in the same way
and the results can also be derived in a few lines of algebra if one
works in an inertial frame in which the rim is rotating. And more
importantly, this SR treats also with the same tools every experiments
involving inertial frames, and in many cases (all MMX-line experiments),
the result is a mere tautology.

> >But then
> >you are most surely in trouble with Michelson & Morley, Kennedy & Thorndike
> >or Doppler because an Ether theory compatible with these experiments is
> >equivalent to SR as long as we are concerned with observable predictions.
>

> Dennis: That's false. Obviously. IGS theory, for example, fits the known
> experiments. And Lorentz contraction does not need to be assumed.

IGS postulate that the Earth is corotating with a vortex of Ether and
therefore that the speed of a lab on Earth wrt Ether is tiny. But then
you have to assume the consequences of this hypothesis: a vortex around
the Earth corotating with the Moon, a vortex around Saturn corotating
with its satellites. But wait the trajectory of these satellites are not
in the same planes and their periods of rotation are very different. So
does this leads to some observable effects related to the propagation of
the light coming from these objects ? Or do you need one vortex for each
satellite ? Same problem with Jupiter. And what about comets ?

And more importantly where is the complete exhaustive quantitative
comparison of the prediction of IGS with the available relevant
experimental data ? GR and SR predictions and their confrontation to
experiments are reported in hundreds of papers published in widely
available journals. There are even summaries of the current situation,
cf [1] for example. So if you want to convince mainstream science you
have to produce a work with the same quality. Can you do that ?


[1] C.M. Will, gr-qc/9811036 and references therein

Dennis McCarthy

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
>
>
>Dennis McCarthy wrote:
>>
>> >
>> >On Wed, 8 Mar 2000 17:49:06 +0000, Dennis McCarthy wrote
>> >(in message <20000308124906...@ng-fm1.aol.com>):
>> >
>> >>> This shows clearly that there is much more than a speed anisotropy at
>> >>> work in Sagnac experiments,
>> >>
>> >> Dennis: How on Earth does it show that? Do you still not understand
>that the
>> >> same effect occurs for sound--and for sound clocks?
>> >
>> >What I have explained is treated extensively and precisely in the
>references
>> >I have given. So it is now your turn: please show your math or at least
>some
>> >references to them
>>
>> Dennis: You have given no evidence for your claim above. That's just a
>strange
>> Bourhis conclusion.
>
Bourhis>For the moment I have not seen either Tom Roberts or Steve Carlip

>contradicting me. Usually they do not hesitate to point out errors in
>whatever messages they are made. Do you think you know better Special
>and General Relativity than all of us ?

Dennis: Forgive me for not being convinced by
proof-by-non-contradiction-by-authority, when my point seems so clear. I'll
repeat it:

Bourhis wrote: ">Now consider two slow moving clocks C+ and C-, respectively


co- and
>counter-rotating. If they are synchronized with a clock at point A, when
>they come back to A, C+ and C- are respectively late and ahead with
>respect to A and the difference between their reading is exactly the
>above time delay T.
>

>This shows clearly that there is much more than a speed anisotropy at
>work in Sagnac experiments, "

Obviously, your argument involving changing clocks cannot rule out the
possibility that speed anisotropy can the cause of the Sagnac effect. This is
true because open-air sound clocks moving around a rotating rim experience the
same effect as described in your premise--and speed anisotropy is what causes
the sound Sagnac.
A cannot entail B when I show an instance where A is true and B is false.

D:>> Also, you gave references for the interesting argument that the length


of
>a
>> rim is not the same CW as CCW for a rim observer. Very strange to say the
>> least.
>

B: >Strange is not by itself a scientific comment.

Dennis: Well, it certainly contradicts what the rim observer measures--as well
as the procedure for determining lengths in all known experiments, including
the one you agreed to.

Bourhis: . This comes from the fact that


>neither the rim nor the light rays are well defined geometrical objects
>entirely embedded in space at a given instant.

Dennis: Ether theories don't have this problem with hypothesizing an
*unobservable* problem with geometrical objects entirely embedded in space at a
given instant.

Bourhis: >This is a definitive result of SR and/or GR, and of course of LET

Dennis: Please stop confusing your idiosyncratic views of a 1904 paper with
ether theories followed today. I have asked you many times to stop this.

B:>Sun-Tzu said : "When involved in a guerilla, launch first the attacks


>which do not cost you anything".

Dennis: I think we should be on an objective search for the truth, not a
guerilla attack.

B:>> >As many people, you consider that defining the length of co- and


>> >counter-rotating light rays or the length of the rim itself is a trivial
>> >question.
>>
>> Dennis: By "A trivial question," Bourhis means using stationary rulers to
>> measure a length of a stationary object. Bourhis feels you can't do that.
>> By "many people", Bourhis means all engineers, scientists, and
>physicsts--who
>> aren't addressing the Sagnac question--*ever* who measure length in the
>normal
>> way.
>

Bourhis: >That is to say in situations where relativistic effects are


completely
>negligeable. That is to say by working with a Galilean framework which
>was indeed a good approximation.

Dennis: Trusting your stationary rulers to measure length of stationary objects
is a "Galilean framework"? And ignoring what you actually measured and going
by theory instead is the relativistic method. Okay, agreed.

D:>> In fact, this was the method of measuring length for Sagnac which


>> **Bourhis himself agreed to** within the last few weeks--before he
>researched
>> the subject--because it's somewhat embarrassing to deny it. And in fact, I
>defy
>> Bourhis to find any experiment in the history of science where length was
>> determined in the way he advocates.
>

Bourhis: >The definition of length used in [4] and precisely explained in [3]


is
>used routinely in GR for a very long time. Moller [5] described it in a
>book published in the 60's and it is therefore even older. It is a very
>general definition because it is equivalent locally to the kind of
>intuitive method you advocate which uses rulers

Dennis: So your desk having a different length CW than CCW is
"intuitive"--despite what you observe, despite what you measure, and despite
the way the rest of the planet measures length?

Bourhis: >Difficulties do only occur when one considers global properties,

Dennis: "Difficulties do only occur" for the theory you advocate. No
"difficulty" occurs for the ether analysis--or for sound Sagnac. And it is
absolutely no problem to measure the distance around a square or circle by
rulers. Engineers, architects, and scientists do it all the time without the
slightest bit of problem at all. The problem only occurs for your theory. So
you have to throw out this obvious measuring method that everyone uses, that is
completely natural, that relies on commonsense definitions of length, and that
you yourself agreed to within the last two weeks--not because engineers or
architects or scientists have a "difficulty" in making the measurement--but
because relativists have a "difficulty" reconciling that obvious measurement
with their theory.

B: And yes this is a new effect


>predicted by Special Relativity and LET but not by Galilean Relativity.
>But how is it possible to be surprised by this fact ?

Dennis: Actually, I would be very surprised to find out that my desk is longer
CW than CCW--as I think most people would. Apparently, you should have been
too because you agreed that it's the same length in both directions two weeks
ago. However, I'm not surprised people are willing to believe that desk is
longer CW than CCW (despite what they measure) in defense of a theory they
advocate.

D: >> It is a length that can only be


>> determined based on theoretical assumptions, i.e., by putting pen to
>paper--and
>> it is a length that *contradicts* what is actually observed by real
>> experimenters using real stationary rulers.
>

B:You are again completely confused. An observer moving infinitely slowly


>along the rim will measure the perimeter:
> L = 2 pi R/(1-(R Omega)^2)^(1/2)
>in units where c=1. But this is not the length of a light ray
>propagating along the rim in the non-inertial frame I have defined
>precisely above. Now how do you propose to measure with rulers the
>length of a light ray ?

Dennis: When you measure the average speed of something that moves from one
point to another (in your frame)--all scientists and engineers measure the
distance between the points (according to your frame) and divide by the time of
travel. I explained this for a week, showing how you glue rulers around the
rim to measure the path. You accepted this incredibly obvious method of
distance measurement that all scientists and all engineers use.
What happened since then?
Do you think it's reasonable to change such reasonable measurement methods
simply because you have since found it that it poses a problem for a theory you
want to advocate?
Don't you think it's more reasonable to abandon the theory--not the universally
accepted and reasonable measurement method?


Dennis McCarthy


Dennis McCarthy

unread,
Mar 9, 2000, 3:00:00 AM3/9/00
to
>>
>> A clock has nothing to do here - a coincidence indicator does the job. What
>> d'you mean by "the same speed"? D'you postulate it? If so, it's exactly the
>> same unanswered question (2).
>
Berry: >No, our observers measure it. They, for some strange reason, are under

the
>impression that the speeds they measure with their clocks and rulers are
>the actual speeds in their frame.
>
>I suppose _you're_ going to say that their measurements are invalid for
>some reason now, aren't you...

Dennis: I'm not sure what is meant here. The rim observer using stationary
rulers and a single stationary clock (or just a "coincidence indicator")
clearly measures anisotropy.
It is the relativists who want to argue that the distance measured by the
stationary rulers is actually wrong--and hypothesize that there are
unobservable deformations of the path that make it different CW than CCW.
It would seem to be that if your theory has a difficulty with a well-known
standard, intuitive, logically consistent, universal measurement method--it's
not the measurement method that should be abandoned.


Dennis McCarthy