On this holiday weekend I finally have some time to devote to this.
On 10/18/21 8:27 PM, Prokaryotic Capase Homolog wrote:
> Is Thomas J. Roberts (2006) paper a reliable source?
Yes :)
Yes. I submitted it to Rev. Mod. Phys. (where Miller published his
results), but they rejected it as not of current interest. In 2006 I was
just starting a new job in physics, and this paper was far removed from
that, so I did not pursue publishing it.
> and b)
> It's in conflict with the papers from Shankland et al. (1955)
Not really. They did not do the error analysis. What conflict to you see?
> and
> Consoli et al. (2013).
Yes. Essentially any valid paper on SR will be in conflict with that
paper. The authors simply do not understand that Special Relativity and
Lorentz Ether Theory are experimentally indistinguishable, and they
spent 70 pages attempting to avoid that fact.
> References
>
> Shankland et al. (1955):
https://doi.org/10.1103/RevModPhys.27.167 In
> their statistical analysis, they find a signal.
>
> Consoli et al. (2013):
https://arxiv.org/abs/1302.3508 They point out
> an error in a method Roberts used.
>
> --Sebastian Pliet (talk) 15:07, 18 October 2021 (UTC)
>
> For various reasons, I consider Shankland's analysis to be somewhat
> deficient. By modern standards, Joos' eighty-year-old results are far
> from "very accurate" and contradict Consoli et al's assertions which
> are largely based on Joos. Quite frankly, the paper by Consoli et al.
> has very much of a bad smell to me in their assertion that modern
> measurements of anisotropy are intrinsically incapable of detecting
> aether drift precisely because they are performed in high vacuum to
> eliminate artifacts -- this is a very common crackpot argument. I
> note also that Eur. J. Physics has frequently published articles on
> special relativity that I consider fringe. Publication in a reliable
> journal does not guarantee reliability of any individual paper. I'll
> bring your questions up with Tom. Prokaryotic Caspase Homolog (talk)
> 01:21, 19 October 2021 (UTC)
In their first paragraph Consoli et al say "it is not obvious how to
distinguish experimentally between the two formulations [Einstein's and
Lorentz's]". They just don't understand that it is IMPOSSIBLE to do so.
No mumbling about "condensation of elementary quanta", or anything else,
can change the basic fact that the two theories are mathematically
equivalent for any experimental prediction. This would only be possible
if there is significant "physics beyond the standard model" that involve
violations of Lorentz invariance; at present there is no evidence of
either..
On their p33, Consoli et al claim my Fig. 3 does not include
"first-harmonic effects". While I have not studied their paper to
determine the truth or significance of that, IT IS IRRELEVANT, because
a) That is clearly an instrumentation effect -- any real signal
cannot possibly have a non-zero first harmonic (their term).
b) I am criticizing Miller's analysis, not that of Consoli et al
(published 5 years after my paper).
c) If a real instrument has a significant first harmonic, it
must be because SOMETHING CHANGES during a 180-degree
rotation. Such a VARYING interferometer is a major problem,
as the experiment demands rigidity and stability. As can be
seen from my Figs. 2 & 3, Miller's interferometer is NOT
stable, and its drift is clearly NOT linear (not even within
a turn). But as section IV of my paper shows, it is possible to
separate the temporal drift from any orientation dependence.
MILLER assumed linearity, which means he ignored the differences between
dots and lines of my Fig. 3, which is as large as 0.5 fringe -- huge
compared to his claimed variation (amplitude ~ 0.06 fringe in my Fig. 1,
remembering his data are 0.1 fringe).
On their p33, Consoli et al present a different analysis of Miller's
data, in which they claim significance. That is IRRELEVANT to my paper,
which is discussing MILLER'S analysis. But it is worse than that: In
their Table 6, Consoli et al have hidden the 6-fringe drift of the
interferometer between turn 1 and turn 20, which dwarfs their values.
They did this by dividing by 2𝛌 -- they never justified that, and it is
clearly unwarranted because the fringe position 𝛌 could be zero, or
100, or any other value depending on where one started counting.
Moreover, like Miller they are ASSUMING a linear drift during each turn,
for an interferometer that is manifestly not drifting linearly (see my
Fig. 2, a plot of his raw data).
In later posts you said the following:
> I made some corrections to my Wikipedia Talk Page response:
How do I see that, and other editors' talk about this content page? When
I click on the "Talk" link of this Wikipedia page, I see my own Talk
page (empty), not editors talking about that specific page. I thought
there would be a single page summarizing various editors' comments on
this specific content page.
> Book
>
https://iopscience.iop.org/book/978-1-64327-738-7/chapter/bk978-1-64327-738-7ch8
I have requested it via inter-library loan. May take a week or more.
BTW in 2006 I gave a colloquium on Miller's results at Case Western
Reserve University, where both Michelson & Morley and Miller performed
their experiments. Their Physics Department was VERY interested to learn
of an explanation for Miller's anomalous results. Prof. Fickinger and I
spent a few hours in their Archives, which have Miller's original data
sheets. We searched for runs with no adjustments [#], and found several
dozen, including some runs with less than 1/2 fringe variation. I was
never interested enough to analyze them.
[#] Miller knew his instrument was drifting. When it
got large enough he would place a coin on one arm to
re-center the image. This happened in >90% of his
runs, so the majority of his runs are not really
using a single interferometer.
Tom Roberts