Speed of Light: A universal Constant?
in [Relativity]
|
Prev: Joan-Claude van Dirk Helps to Trivialize Special Relativity
Next: GOD=G_uv Measure your IQ in 30 seconds
From: George Dishman on 13 Apr 2005 16:24 "Henri Wilson" <H@..> wrote in message news:hjmo51pkkip4o5bobs2eb88sh42d98t0gj(a)4ax.com... > On 12 Apr 2005 05:37:59 -0700, "Dishman" <george(a)briar.demon.co.uk> wrote: > >>Henri Wilson wrote: >>> On Mon, 11 Apr 2005 20:33:31 +0100, "George Dishman" >><george(a)briar.demon.co.uk> >>> wrote: >>> >>> > > >>> >> I cannot see any connection with SR. >>> > >>> >I think it is obvious. The the speed of the light >>> >is c in the lab frame, a simple deduction from >>> >the postulates of SR. To emphasise the point, >>> >consider two such experimental setups in the same >>> >lab. One is fixed to the floor while the other >>> >moves through the lab on rails at constant speed. >>> >My analysis holds good for both because the speed >>> >is c in an inertial (non-rotating) frame >>> >referenced to the centre point of _either_ table. >>> >>> But you don't know if the number of fringes moved is the same >>> in both cases for the same rotation! >>> >>> To show that you will have to move one setup at a very high speed. >> >>Hold on Henri, let's finish the point before >>looking at another. We were discussing the >>basis of the explanation I gave, not >>experimental results. That can come later. >> >>You said that you couldn't see the connection >>between the standard explanation as shown in >>my animation and SR. The link is that it uses >>the fact that the speed of light is the same >>in all inertial frames to determine the >>behaviour of the light. Do you now follow that >>part? > > Well, is a rotating frame an inertial one? No, an inertial frame has coordinate axes that are not rotating and the origin is not accelerating. Given those conditions, SR says that the measured speed of light in vacuum, defined as always as the rate of change of the coordinates of the light, will have a fixed value, 'c'. That is what I used to derive the motion of the dots in the right hand diagram of the animation hence it is SR-based. For the Ritzian diagram on the left, the initial velocity of the light is the sum of the velocity of the source and a vector of magnitude 'c' directed tangentially. Thereafter it maintains its speed. Again that analysis is only true if those quantities are expressed in non-rotating coordinates >>You also said: >> >>> >> It is based on an aether concept that >>> >> there is an absolute frame. >> >>That is not true because the speed would be c >>in the aether frame in that case. Do you see >>that? Here are the details: >> >>> >Whether some other aether-based theory might look >>> >similar is irrelevant, the explanation is valid >>> >for an SR analysis. >>> > >>> >That said, you should also note that what I posted >>> >is _not_ a valid analysis for most absolute frame >>> >theories where the speed of light is c in that >>> >frame and you have to consider the speed of the >>> >lab with respect to the preferred-frame to find >>> >the speed of the light at any point. >>> > >>> >The simplest example might be a Galilean aether >>> >with the lab moving through the aether in a >>> >directon that lies in the plane of the turntable. >>> >The speed of the light when moving round the table >>> >in the same direction as the lab is moving wrt. >>> >the aether is c-V while on the other side of the >>> >table is it c+V. Of course this mostly cancels out >>> >as both beams traverse almost one full turn but >>> >the beam going the same way as the table rotation >>> >covers the section between the emission point and >>> >where it hits the detector twice while the other >>> >beam never travels that part. The result is that >>> >the speed in that section varies depending on the >>> >alignment of that part with the direction of >>> >aether flow and as it varies slightly round the >>> >table, there would be a slight modulation of the >>> >signal with orientation. That would be tiny and >>> >unmeasurable in reality but it is there >>> >theoretically. Clearly that is a quite different >>> >approach and a slightly different result. >>> > >>> >With a Lorentz-invariant aether, we know it gives >>> >the same results as SR so it could borrow the >>> >analysis I gave. Still, from first principles, you >>> >should start as I did for the Galilean aether to >>> >find the speed at any point round the path but >>> >taking account of length contraction which turns >>> >the table into an ellipse instead of a circle. >>> >Then note that the frequency of the light emitted >>> >will be altered by the Lorentzian equivalent of >>> >"time dilation" depending on the speed of the >>> >source through the aether at the time of emission. >>> >Then find the phase difference between the delayed >>> >version of this that reaches the detector by the >>> >two paths. >>> > >>> >I'll say good luck if you want to try that >>> >approach, I don't like the look of it at all, but >>> >it must end up giving the same result as SR. >>> > >>> >Anyway, the point is that any preferred-frame >>> >analysis must take account of the motion of the >>> >table centre in the preferred-frame and that >>> >factor does not appear in what we have done so >>> >far so you cannot mistake the analysis for an >>> >absolute-frame theory, and as I said at the top, >>> >even if some aether theory did by chance coincide, >>> >it doesn't prevent this being a valid SR analysis. >>> >>> It might be an analysis but is it correct? >> >>The SR one certainly is. It has been around >>for over 90 years, ever since Sagnac did the >>experiment, and has been examined by hundreds >>of students and experimeners. It has been peer >>reviewed more times than you and I have had >>hot dinners! > > The christian bible has been 'peer reviewed' for 2000 years and still > hasn't an > ounce of truth in it. That isn't comparable, we are only asking whether all the aspects of SR have been accounted for in my derivation. >>The aether analyses above are just outlines to >>make the point that they would have to take the >>speed of the lab wrt the aether into account >>hence my analysis is obviously not aether-based. > > aether is dragged with the apparatus. I was going to snip the aether stuff above but maybe you should remind yourself of what I said, if I had used a dragged aether, not only would the speed of the lab relative to the aether have to appear in the calculations, I would also have had to deal with the drag coefficient by calculating the speed of the aether in some other inertial frame. The more you add, the more obvious it becomes that my analysis could only be SR based. >>If you now understand those points and can see >>that my animation is SR-based, we can move on. > > I think it is based on something completely separate from SR. The bottom line here Henri is that since I wrote the page, I know what I used to work out the values, and it wasn't an aether- based model of any kind, I used SR. > but there is a > purely coincidental connection. > <snip> >>> >BTW, I haven't mentioned refractive index as I >>> >assumed you would realise it cancels out in both >>> >theories. >>> >>> Does it? I'll have to think about that. >> >>The easy way is again in the rotating frame. >>Both beams move at c/n instead of c in the >>Ritzian analysis so acts like a scaling >>factor on the output. The SR version would be >>more complex but it can be treated as a drag >>(ref. Fresnel's experiments with water in >>u-tubes) so would again produce just a scale >>factor. > > an important one. It is only a scaling factor. Zero remains zero. <snip the model> I assumed you would be using VB rather than trying to do it in Java. If you let me know how it differs from mine, I might try doing a Java version. I can't promise, I've got a lot on too, but it would be interesting to compare them. >>For your Ritzian model, as I say, I have been >>waiting several years for someone to resolve >>the problem and I haven't seen any credible >>suggestions of what might be missed. Still, >>perhaps you will be the one to solve the >>problem. > > I have already explained what happens but you cannot see it. No, you have talked about the angles being other than 90 degrees, which I agree, and about the speeds at the mirrors, but you haven't explained why either of those would affect the time taken other than indirectly through their influence on the path length. For example: >>> >> In the four mirror system, the light is reflected >>> >> at an angle that is not 90 (during rotation) >>> > >>> >The angles are illustrated in the simple path- >>> >drawing applet but it doesn't affect the time >>> >taken along the path so there is nothing to >>> >put into the animation. >>> >>> It DOES affect the time. Why do you claim it doesn't? >> >>Because the time is determined fully by the >>length of the path. Unless you are suggesting >>some sort of delay due to absorption and >>subsequent re-emission during the reflection >>process, I don't see how it can add anything >>to the path-length time. > > You will have to wait for my animation. I rest my case ;-) >>> >> There is also a quite complex velocity change to consider >>> >> at each reflection. >>> > <snip> >>> > >>> > http://www.briar.demon.co.uk/Henri/speed.gif >>> >>> I don't understand that diagram. >>> >>> Why are the green lines sloping? >> >>The purple and green lines are the vectors, >>not the path of the light. The actual path is >>the red line and the sloping green line shows >>the direction the light must be emitted such >>that it adds to the motion of the mirror >>(shown by the purple vector arrow) to produce >>the resultant motion necessary for it to hit >>the second mirror. It is doing standard vector >>addition graphically, nothing fancy. > > Why do you think the actual path is the red line. That is the > source of your misunderstanding. Because it is drawn in the lab frame hence the path of the light must a straight line. The red line is the only way to draw a straight line between the points of reflection at A and B. Start with points A and B as the points of reflection. The vertical red lines down from them are the reflected beams and although shown as exactly vertical, that is only a laziness, the angle would be more or less than 90 degrees from the A-B line depending on which beam is being considered. The horizontal red line is one leg of the path and the diagram then shows by symmetry that the if the light leaves one mirror at c, it arrives at the other also at c. It is not stubbornness on my part as you say below but the fact that you have never attempted to tell me what you think is wrong with this diagram. I do make mistakes Henri, I'm human, but you'll have to get specific if you think there is an error because I can't see it. >>> >Can you think of any other possibilities? I have >>> >asked about half a dozen Ritz supporters including >>> >yourself over about six years and so far nobody >>> >has found anything I have missed. >> >>I guess you haven't spotted anything else yet. > > I have already explained. > You are too stubborn to accept the truth. So far you have identified two aspects and your explanations are: > You will have to wait for my animation. and > Why do you think the actual path is the red line. That is the > source of your misunderstanding. I don't consider those to be explanations in any sense. I'm quite willing to wait to see your animation but until then you haven't explained anything. George
From: Henri Wilson on 13 Apr 2005 17:43 On Wed, 13 Apr 2005 21:24:15 +0100, "George Dishman" <george(a)briar.demon.co.uk> wrote: > >"Henri Wilson" <H@..> wrote in message >news:hjmo51pkkip4o5bobs2eb88sh42d98t0gj(a)4ax.com... >> On 12 Apr 2005 05:37:59 -0700, "Dishman" <george(a)briar.demon.co.uk> wrote: >> >>>behaviour of the light. Do you now follow that >>>part? >> >> Well, is a rotating frame an inertial one? > >No, an inertial frame has coordinate axes >that are not rotating and the origin is >not accelerating. Given those conditions, >SR says that the measured speed of light >in vacuum, defined as always as the rate >of change of the coordinates of the light, >will have a fixed value, 'c'. That is what >I used to derive the motion of the dots in >the right hand diagram of the animation >hence it is SR-based. > >For the Ritzian diagram on the left, the >initial velocity of the light is the sum >of the velocity of the source and a vector >of magnitude 'c' directed tangentially. >Thereafter it maintains its speed. Again >that analysis is only true if those >quantities are expressed in non-rotating >coordinates. I'm now wondering how to calculate the velocity component of light as it reflects from a moving 45 deg mirror. I can see that this analysis depends entirely on the answer to that question. >I would also have had to deal with the drag >coefficient by calculating the speed of the >aether in some other inertial frame. The more >you add, the more obvious it becomes that my >analysis could only be SR based. > >>>If you now understand those points and can see >>>that my animation is SR-based, we can move on. >> >> I think it is based on something completely separate from SR. > >The bottom line here Henri is that since I >wrote the page, I know what I used to work >out the values, and it wasn't an aether- >based model of any kind, I used SR. > George, I am now losing quite a lot of sleep thinking about this Sagnac thing. ,,thanks to you. I want to know more about the optical system used in ring gyros. What I have realised is that the standard SR explanation doesn't work!! FOR A CONSTANT RATE OF ROTATION, THERE WOULD BE NO FRINGE MOVEMENT. The pattern would remain fixed. And that is not what happens, surely. Your version of Sagnac would be sensitive to angular acceleration only. >>> >>>The easy way is again in the rotating frame. >>>Both beams move at c/n instead of c in the >>>Ritzian analysis so acts like a scaling >>>factor on the output. The SR version would be >>>more complex but it can be treated as a drag >>>(ref. Fresnel's experiments with water in >>>u-tubes) so would again produce just a scale >>>factor. >> >> an important one. > >It is only a scaling factor. Zero remains zero. Yes. OK. According to your model, there should be a theoretical displacement of fringes for a certain angular change. Refractive index comes into that directly. > ><snip the model> > >I assumed you would be using VB rather than >trying to do it in Java. If you let me know >how it differs from mine, I might try doing >a Java version. I can't promise, I've got a >lot on too, but it would be interesting to >compare them. Like I said, the problem largely boils down to what happens when light reflects from a moving 45 mirror. I don't know if Huygens can answer that one convincingly. There doesn't appear to be any experimental evidence either. > >>>For your Ritzian model, as I say, I have been >>>waiting several years for someone to resolve >>>the problem and I haven't seen any credible >>>suggestions of what might be missed. Still, >>>perhaps you will be the one to solve the >>>problem. >> >> I have already explained what happens but you cannot see it. > >No, you have talked about the angles being >other than 90 degrees, which I agree, and >about the speeds at the mirrors, but you >haven't explained why either of those would >affect the time taken other than indirectly >through their influence on the path length. >For example: > >>>> >> In the four mirror system, the light is reflected >>>> >> at an angle that is not 90 (during rotation) >>>> > >>>> >The angles are illustrated in the simple path- >>>> >drawing applet but it doesn't affect the time >>>> >taken along the path so there is nothing to >>>> >put into the animation. >>>> >>>> It DOES affect the time. Why do you claim it doesn't? >>> >>>Because the time is determined fully by the >>>length of the path. Unless you are suggesting >>>some sort of delay due to absorption and >>>subsequent re-emission during the reflection >>>process, I don't see how it can add anything >>>to the path-length time. >> >> You will have to wait for my animation. > >I rest my case ;-) George, my latest thoughts are that we are looking in entirely the wrong direction for an explanation of Sagnac. Yours (and SR's) plainly doesn't work. I now don't think it is fundamentally related to light speed or path differences. I think light speed is unimportant in this. Rather, I believe there is some kind of 'gyroscopic effect', maybe associated with the planes of the fields. "Emitted light has a built-in rotational reference" as it were. This approach might also reveal a fundamental association between light and the Earth's gravitational field, since that is one reference for 'vertical'. >>>> > http://www.briar.demon.co.uk/Henri/speed.gif >>>> >>>> I don't understand that diagram. >>>> >>>> Why are the green lines sloping? >>> >>>The purple and green lines are the vectors, >>>not the path of the light. The actual path is >>>the red line and the sloping green line shows >>>the direction the light must be emitted such >>>that it adds to the motion of the mirror >>>(shown by the purple vector arrow) to produce >>>the resultant motion necessary for it to hit >>>the second mirror. It is doing standard vector >>>addition graphically, nothing fancy. >> >> Why do you think the actual path is the red line. That is the >> source of your misunderstanding. > >Because it is drawn in the lab frame hence the >path of the light must a straight line. The red >line is the only way to draw a straight line >between the points of reflection at A and B. > >Start with points A and B as the points of >reflection. The vertical red lines down from >them are the reflected beams and although >shown as exactly vertical, that is only a >laziness, the angle would be more or less >than 90 degrees from the A-B line depending >on which beam is being considered. ok, that's why I was confused. The departure from 90 is all important. > The >horizontal red line is one leg of the path >and the diagram then shows by symmetry that >the if the light leaves one mirror at c, it >arrives at the other also at c. It is not >stubbornness on my part as you say below but >the fact that you have never attempted to >tell me what you think is wrong with this >diagram. I do make mistakes Henri, I'm >human, but you'll have to get specific if >you think there is an error because I can't >see it. I'll have another look. My problem now is to work out the reflection angle from a moving 45 mirror. Is the reflected angle identical to the incident one? What is the new speed of the beam in the nonrotating frame? > >>>> >Can you think of any other possibilities? I have >>>> >asked about half a dozen Ritz supporters including >>>> >yourself over about six years and so far nobody >>>> >has found anything I have missed. >>> >>>I guess you haven't spotted anything else yet. >> >> I have already explained. >> You are too stubborn to accept the truth. > >So far you have identified two aspects and your >explanations are: > >> You will have to wait for my animation. > >and > >> Why do you think the actual path is the red line. That is the >> source of your misunderstanding. > >I don't consider those to be explanations >in any sense. I'm quite willing to wait >to see your animation but until then you >haven't explained anything. It's on the way. Like I said George, this thing is keeping me awake at night. it is very complex. I don't really know what to think at present. ...but as far as I can see, your standard explanation doesn't work. > >George > HW. www.users.bigpond.com/hewn/index.htm Sometimes I feel like a complete failure. The most useful thing I have ever done is prove Einstein wrong.
From: Henri Wilson on 13 Apr 2005 18:02 On Wed, 13 Apr 2005 01:53:07 +0000 (UTC), bz <bz+sp(a)ch100-5.chem.lsu.edu> wrote: >H@..(Henri Wilson) wrote in >news:pkno51hjhep1ole2cib56l4m9i9bl70gsd(a)4ax.com: > >>>I am measuring time of flight between two stationary points. >> >> You have designed a typical TWO-WAY light speed experiment. >> Both the positioning of the clocks and their synchronization involves >> light going the opposite way to the beam you are trying to measure. >> > >I don't care if it is one way, or two way. It doesn't matter. It could be >ten way. I am NOT trying to measure the aether drift. > >There is nothing in the 'time of flight'[TOF] portion of the experiment >that is being changed during the experiment [except the speed and or >energy and or wavelenght of the photons traveling between the detectors]. > >The only thing being changed is the speed of the source [and the speed of >the photons {if the photons actually change speed as per BaT}]. > >If Bat is true, the photons will be traveling at different speeds when the >leave the moving source. The time of flight will be different. > >We actually have 3 experiments >1) source moving toward TOF apparatus. >2) source stationary. >3) source moving away from TOF apparatus. > >if BaT works, TOF will vary. TOF will be shortest in 1, intermediate in 2, >and longest in 3. If we vary the speed continuously between 1 and 2, TOF >should vary continuously. The same between 2 and 3. > >If there is no change AND if our sensitivity is sufficient to detect BaT >[and it should be], then BaT is invalidated. I have designed the experiment for you. You don't need any clocks. You only need a very sensitive and fast PM and fast CRO. I have explained why there is not a hope in hell that the experiment will work. ______ You might not be aware of my own very feasible experiment to compare light speed from two differently moving sources. It requires a relay station on the edge of the moon. A fast moving distant transmitter (eg, orbiting Saturn) sends a sharp radio pulse to Earth when the edge of the moon is closely aligned. The pulse moving at c+v is detected by the R/X on the moon and quickly relayed to Earth at around c (since the moon has negligible radial velocity wrt earth). There will be a brief but known apparatus delay is the process. To prove the BaT correct, the original pulse and the relayed pulse will have to arrive at different instants at the Earth receiver. (The known delay must be subtracted). Even at these distances and speeds, the time difference will be only around 100 us at best. HW. www.users.bigpond.com/hewn/index.htm Sometimes I feel like a complete failure. The most useful thing I have ever done is prove Einstein wrong.
From: Henri Wilson on 13 Apr 2005 18:12 On Wed, 13 Apr 2005 03:57:57 GMT, Sam Wormley <swormley1(a)mchsi.com> wrote: >Henri Wilson wrote: > >> >> Have a look at the brightness curves of the following variable stars: >> >> R Aquilae >> R Andromedae >> R Arietis >> R Aur >> X Aur >> R Boo >> S Boo >> U Boo* >> V Boo* >> V CVn** >> R Cam >> V Cam" >> X Cam >> Z Cam >> R Cas* >> S Cas** >> t Cas** >> W Cas >> S Cep* >> T Cep* >> Omicron Ceti >> R Com >> R Crb*** >> S Crb >> V Crb >> W Crb >> R Cyg >> S Cyg >> V Cyg >> W Cyg >> AF Cyg*** >> Chi Cyg >> R Dra >> R Gem >> S Her* >> RU Her** >> SS Her >> AH her >> R Hya >> SU Lac >> X Oph >> U ori >> R Scuti** >> R Ser >> V Tau >> R Uma >> S Uma >> T Uma >> CH Uma*** >> S Umi >> R Vul >> V Vul* >> >> They all match the ballistic predictions. >> > > Here's the light curve for Omicron Ceti > http://www.aavso.org/vstar/vsots/1298.shtml > > What is it you want me to notice? Try http://www.britastro.org/vss/gifl/00098.gif This is a most typical brightness curve as predicted by the BaT for low eccentricity orbits. Notice that the period appears dead constant for over a century. That could only happen if the variation was in synch with something very stable, like an orbit. It certainly could not be the result of a diffusive, random or chaotic process. HW. www.users.bigpond.com/hewn/index.htm Sometimes I feel like a complete failure. The most useful thing I have ever done is prove Einstein wrong.
From: George Dishman on 13 Apr 2005 19:30
"Henri Wilson" <H@..> wrote in message news:s03r51d3gmd6qick1ffuuiebr31gljr9lk(a)4ax.com... > On Wed, 13 Apr 2005 21:24:15 +0100, "George Dishman" > <george(a)briar.demon.co.uk> > wrote: > >> >>"Henri Wilson" <H@..> wrote in message >>news:hjmo51pkkip4o5bobs2eb88sh42d98t0gj(a)4ax.com... >>> On 12 Apr 2005 05:37:59 -0700, "Dishman" <george(a)briar.demon.co.uk> >>> wrote: >>> > >>>>behaviour of the light. Do you now follow that >>>>part? >>> >>> Well, is a rotating frame an inertial one? >> >>No, an inertial frame has coordinate axes >>that are not rotating and the origin is >>not accelerating. Given those conditions, >>SR says that the measured speed of light >>in vacuum, defined as always as the rate >>of change of the coordinates of the light, >>will have a fixed value, 'c'. That is what >>I used to derive the motion of the dots in >>the right hand diagram of the animation >>hence it is SR-based. >> >>For the Ritzian diagram on the left, the >>initial velocity of the light is the sum >>of the velocity of the source and a vector >>of magnitude 'c' directed tangentially. >>Thereafter it maintains its speed. Again >>that analysis is only true if those >>quantities are expressed in non-rotating >>coordinates. > > I'm now wondering how to calculate the velocity component of light as it > reflects from a moving 45 deg mirror. > I can see that this analysis depends entirely on the answer to that > question. For the Ritzian analysis, I agree entirely and that depends to a degree on how the physics works, but both variants I can imagine give the same answer. That is what the diagram below tries to explain, but it is _my_ understanding and really it is an "aunt sally", a target set up for you to correct based on your model. Huygens will tell you the rule that relates the angles but only if you first know how the speed is affected. >>I would also have had to deal with the drag >>coefficient by calculating the speed of the >>aether in some other inertial frame. The more >>you add, the more obvious it becomes that my >>analysis could only be SR based. >> >>>>If you now understand those points and can see >>>>that my animation is SR-based, we can move on. >>> >>> I think it is based on something completely separate from SR. >> >>The bottom line here Henri is that since I >>wrote the page, I know what I used to work >>out the values, and it wasn't an aether- >>based model of any kind, I used SR. >> > > George, I am now losing quite a lot of sleep thinking about this Sagnac > thing. :-) I'm pleased you realise that this is more difficult than I think you had at first assumed. It is a powerful test. > ,,thanks to you. I want to know more about the optical system used in ring > gyros. There is a lot more to the detail in real products, in particular additional modulation is used to improve the performance. This will explain a bit more. I can try to find some other references I saw some weeks ago if it will help. http://www.physik.fu-berlin.de/~bauer/habil_online/node11.html > What I have realised is that the standard SR explanation doesn't work!! > > FOR A CONSTANT RATE OF ROTATION, THERE WOULD BE NO FRINGE MOVEMENT. > > The pattern would remain fixed. And that is not what happens, surely. > Your version of Sagnac would be sensitive to angular acceleration only. Why? The animation shows that there is a clear time difference produced in the SR version. What you say is inconflict with what it shows yet you just make that bald statement without giving any reason. >>>>The easy way is again in the rotating frame. >>>>Both beams move at c/n instead of c in the >>>>Ritzian analysis so acts like a scaling >>>>factor on the output. The SR version would be >>>>more complex but it can be treated as a drag >>>>(ref. Fresnel's experiments with water in >>>>u-tubes) so would again produce just a scale >>>>factor. >>> >>> an important one. >> >>It is only a scaling factor. Zero remains zero. > > Yes. OK. According to your model, there should be a theoretical > displacement of > fringes for a certain angular change. Refractive index comes into that > directly. For a certain angular displacement in the time it takes the light to traverse the path, or in other words for a constant angular speed. >><snip the model> >> >>I assumed you would be using VB rather than >>trying to do it in Java. If you let me know >>how it differs from mine, I might try doing >>a Java version. I can't promise, I've got a >>lot on too, but it would be interesting to >>compare them. > > Like I said, the problem largely boils down to what happens when light > reflects > from a moving 45 mirror. I don't know if Huygens can answer that one > convincingly. There doesn't appear to be any experimental evidence either. Huygens will define the angles but not the speed, for that you need physics. I can imagine two possibilities: if Ritzian particles bounce like balls then the speed of the reflected ray relative to the mirror is the same as the incident ray while if they are absorbed and re-emitted it would be c relative to the mirror regardless. Both those give the same result though. Can't you work it out from the Doppler shift on reflection? >>> You will have to wait for my animation. >> >>I rest my case ;-) > > George, my latest thoughts are that we are looking in entirely the wrong > direction for an explanation of Sagnac. Yours (and SR's) plainly doesn't > work. Plainly it does, the animation makes it quite clear that there is a difference in the arrival times on the right hand diagram. Try setting the slider to maximum and watching the dots as a pair. > I now don't think it is fundamentally related to light speed or path > differences. I think light speed is unimportant in this. > > Rather, I believe there is some kind of 'gyroscopic effect', maybe > associated > with the planes of the fields. "Emitted light has a built-in rotational > reference" as it were. This approach might also reveal a fundamental > association between light and the Earth's gravitational field, since that > is > one reference for 'vertical'. The gyros work at any angle. Remember they are used in flight control for fly-by-wire jet fighters ! The gyroscope effect isn't far wrong though and ring laser gyros are often explained that way, but that simply means the rotational speed is fixed in the non-rotating frame regardless of the rotation of the source, fibre, laser material etc. which is exactly the basis of SR. >>>>> > http://www.briar.demon.co.uk/Henri/speed.gif >>>>> >>>>> I don't understand that diagram. >>>>> >>>>> Why are the green lines sloping? >>>> >>>>The purple and green lines are the vectors, >>>>not the path of the light. The actual path is >>>>the red line and the sloping green line shows >>>>the direction the light must be emitted such >>>>that it adds to the motion of the mirror >>>>(shown by the purple vector arrow) to produce >>>>the resultant motion necessary for it to hit >>>>the second mirror. It is doing standard vector >>>>addition graphically, nothing fancy. >>> >>> Why do you think the actual path is the red line. That is the >>> source of your misunderstanding. >> >>Because it is drawn in the lab frame hence the >>path of the light must a straight line. The red >>line is the only way to draw a straight line >>between the points of reflection at A and B. >> >>Start with points A and B as the points of >>reflection. The vertical red lines down from >>them are the reflected beams and although >>shown as exactly vertical, that is only a >>laziness, the angle would be more or less >>than 90 degrees from the A-B line depending >>on which beam is being considered. > > > ok, that's why I was confused. The departure from 90 is all important. Not to the speed. I'll revise the diagram over the weekend, I'm too busy before then. >> The >>horizontal red line is one leg of the path >>and the diagram then shows by symmetry that >>the if the light leaves one mirror at c, it >>arrives at the other also at c. It is not >>stubbornness on my part as you say below but >>the fact that you have never attempted to >>tell me what you think is wrong with this >>diagram. I do make mistakes Henri, I'm >>human, but you'll have to get specific if >>you think there is an error because I can't >>see it. > > I'll have another look. > My problem now is to work out the reflection angle from a moving 45 > mirror. No, that part is easy. It is whatever is necessary to get the light to the detector. Remember there is dispersion of the beam so the photons that happen to be going in the right direction are the only ones that count, we can disregard any others. A hint might be that if the table turns fast enough, the mirror will be out of the beam by the time the light gets there. > Is the reflected angle identical to the incident one? Yes, if the speed is the same (regardless of its value). That lets you solve for the unique path that meets the above criterion. > What is the new speed of the beam in the nonrotating frame? My understanding is that both variants of the Ritzian model I outlined above give the same result, c relative to the mirror and hence reflected angle equal to the incident, but you might come up with an alternative. >>So far you have identified two aspects and your >>explanations are: >> >>> You will have to wait for my animation. >> >>and >> >>> Why do you think the actual path is the red line. That is the >>> source of your misunderstanding. >> >>I don't consider those to be explanations >>in any sense. I'm quite willing to wait >>to see your animation but until then you >>haven't explained anything. > > It's on the way. > > Like I said George, this thing is keeping me awake at night. > it is very complex. I don't really know what to think at present. I know it is a challenge, but then that is why is is a valuable test. At the end of the day, the standard formula dt = 4Aw/c^2 has been used for decades and thoroughly tested so we can treat it as empirically correct. The SR analysis gives the same formula as you can see here: http://www.mathpages.com/rr/s2-07/2-07.htm so the aim of your thoughts should be to show how you can derive that formula using a ballistic light model. Every attempt I ahve made says there will be no delay or phase change or fringe shift, however you measure it, other than some second-order effects that will be about a billion times too small. > ..but as far as I can see, your standard explanation doesn't work. It works just fine. Remember in the animation the SR model is on the right. The mathpages link above gives the same analysis but just a static drawing. best regards George |