Requiem for Relativity

In Barbarossa's Cavern There Are No Stars (Part VII)

Above, I very roughly estimated a Visual extinction of 1.0 mag for passage through the full thickness of Barbarossa's cloud; a more accurate version of this estimate (see below) is 0.3 mag. This would of course be due to the dust component. What about the gas?

The column of hydrogen atoms amounts to a layer of dissociated neutral hydrogen at standard temperature and pressure, 12cm thick; add another 1cm for the helium. This is 1/60,000 the number of gas molecules in a column of Earth's atmosphere, so, Visual extinction, due to the gas, would be negligible. These molecules are illuminated only 1/197.7^2 = 1/40,000 as intensely by the sun. The regional sky brightening seen from Earth's surface, due to sunlight scattered by the gas atoms of Barbarossa's cloud, would be somewhat less than the sky brightening, due to starlight scattered by Earth's atmosphere.

Let Barbarossa's cloud be a homogeneous oblate spheroid with a=26AU, b/a=0.5. Let the path to Theta Crt transect a chord 0.6 * 2 * a. Then the mass of the cloud, assuming 75% H by weight and that 90% of Fruscione's N(H) is within the cloud, would be 0.028 Earth masses (1/120,000 the mass of the Barbarossa system). Because estimates from metal atom abundances can overestimate H abundance by 0.5 log unit (Fruscione, 1994, p. 129), a lower bound for the mass would be 0.01 Earth masses. (In interstellar matter, dust averages only 0.7% the mass of gas.) There would be 740,000 gas atoms (helium or dissociated hydrogen) per cubic cm; this is 100x the density of the Martian exosphere as measured by Mars Express (Wurz et al, Geophysical Research Abstracts, Vol. 8, 01954, 2006). The equivalent altitude on Earth is below that at which aurorae begin to manifest, so, if Barbarossa has a magnetic field, the sun should cause aurorae in Barbarossa's cloud.

What about sky brightening due to sunlight scattered by the dust of Barbarossa's cloud? If the densest part of Barbarossa's cloud intercepts (i.e., extinguishes, either by absorption or scattering) ~30% of starlight (0.3 mag extinction)(Bohlin's 1978 empirical formula relating N(H) to extinction, cited by Draine, AnnuRevAstronAstrophys 2003, 41:241+, p. 244, sec. 2.1.2, eq. (2), gives 0.25 mag), then it also intercepts 30% of sunlight.

I follow Endrik Kruegel's text, "The Physics of Interstellar Dust" (ch. 4, "Case Studies of Mie Calculus"). For a spherical particle which is a "strong absorber" (sec. 4.1.3)(optically not very different from amorphous carbon, which is discussed in detail in sec. 4.1.5), define x = circumference/wavelength. For the "typical" interstellar dust grain, diam = 0.1mu (Kruegel, sec. 4.1.5.1), wavelength = 0.6mu for visible light, so x = 1. For x < 0.5, the ratio of scattering to extinction is proportional to x^3, and for x=0.5, is (Fig. 4.3) about 1/3 * 0.5^3 = 1/24. For x > 2, the ratio of scattering to extinction is about 1/2 overall, but "g", the average cosine of the scattering angle, is about 0.8, which implies that backscattering can't possibly be more than 1/5 of the amount found when there is isotropic scattering; so, the ratio of backscattering to extinction is < 1/10. For x = 1, the situation is intermediate. So, a fair estimate of backscattering to extinction, is 1/24.

More precisely, I could for x < 0.5, neglect scattering, and for extinction (which in this domain is proportional to x times the geometric cross section), integrate the often assumed a^(-3.5) power law for particle size distribution (JE Dyson, "The Physics of the Interstellar Medium", p. 53), assuming (as Draine suggests) the law holds at least halfway down to molecular sizes; and assuming it achieves the geometric cross section, when x = 1. The ratio of scattering to extinction is maximum at the "typical" (maybe this is why it's considered typical) diameter 0.1mu; again I could use the power law to integrate scattering between x = 0.5 & x = 2 (using a constant scattering efficiency ~1/6 seen near x = 1, from Kruegel's Fig. 4.3); Simpson's rule (based again on Fig. 4.3) to get from the scattering, the extinction due to this domain; using "g" = ave(cos(theta)) = 0.2 at x = 1, and a first-order spherical harmonic, I can estimate the relative backscattering efficiency as 40% in this domain. This more precise calculation merely changed the estimate to 1/25 from 1/24.

The effective albedo of Barbarossa's cloud then would be at least 70% (a lower bound for the fraction of light which escapes a second interception) * 30% * 1/24 / 2 (because of isotropy) = 0.44%, ~1/30 that of Luna. The cloud's surface would be 1/(30*40,000) as bright as Luna's; it would be Vmag -12.7+15.2 = +2.5 on a 0.52deg diam disk, i.e. +0.8 on a sq. deg. The brightest part of the northern Milky Way (Zavarzin, Astrophysics 23:647+, Table 1) has brightness +4.05 on a sq. deg.

So, if the ~0.3 mag extinction (much less than that, hardly would be consistent with the observed relative dimness of R2 & B2 in the region) is due to "typical" dust grains, the Barbarossa cloud would be too bright. However, the extinction can be achieved with arbitrarily little scattering, if these "strong absorber" grains are small enough. Diameter 0.01mu, gives scattering/extinction = ~ 1/3 * (1/10)^3 = 1/3000, 3000/24 = 5.2mag less, i.e., sky brightness a plausible +6.0 on a sq. deg. Alternatively, the physical properties of the "dust" might be different than presently believed.

cc

To: organizers of 2nd (Sept. 2008) Crisis in Cosmology Conference

Dear sirs:

The link didn't work for me, so here is my application.

Sincerely,
Joseph C. Keller

Presentation Type: Oral Presentation

Paper Title:
Brown Dwarf, Barbarossa, Causes "Cosmic" Microwave Background Dipole

Author & Affiliation: Joseph C. Keller (B. A., Harvard)

Primary contact: Joseph C. Keller

Contact email: *******

Short Abstract:

A cold brown dwarf, Barbarossa, causes the "Cosmic" Microwave Background dipole. Barbarossa+Frey+Freya, with a dark nebula, lie at the (+) CMB dipole.

Full Abstract (slightly < 2500 character count):

A cold brown dwarf, Barbarossa, causes the "Cosmic" Microwave Background dipole. Barbarossa+Frey+Freya, with a dark nebula, lie at the (+) CMB dipole.

Exclusively near there, red & blue USNO-B magnitudes dim between c.1954 and c.1985; likewise, Johnson's bright star photometry c.1964 vs. Harvard magnitude as published mostly in 1908. Barbarossa's nebular size, is given by extrapolating Hubble's nebular size relation. Interstellar absorption lines before the two studied nearby stars in this direction, are exceedingly strong.

My original finding was that automated USNO-B R1 & R2 magnitudes, differing enough to be misidentifications of wanderers, outlined an orbital path there. Dots of magnitude ~ +18, though not of typical starlike appearance, on all relevant red and infrared survey plates, and on prospective photos by Joan Genebriera, Steve Riley, and Robert Turner, lie within arcseconds of an e < 0.1, 198 AU orbit slightly leading the (+) CMB dipole. Frey has a 3-yr, e = 0.65 orbit around Barbarossa with retrograde apsis precession in 24 yr. Freya (not yet identified) is inferred to orbit in 6 yr. perpendicular to the ecliptic, causing Frey's precession, and lateral deviation of the Barbarossa-Frey c.o.m. The projection of Barbarossa's orbit onto Jupiter's, follows the mean position of a Jupiter/Saturn conjunction resonance.

Claimed COBE & WMAP error bars rule out such a near orbit. However, only a cause within the solar system, explains the correlation, of the Maxwellian moments of the CMB anisotropy, with the plane of the ecliptic.

My theory of CMB production by gravitational fields (electrons boiling from the surface of the sun's, 52.6 AU radius, ether island, detected by anomalies in Pioneer10's transmission there), and, a Newtonian theory of nodal regression resonances in the outer solar system, give equal estimates of Barbarossa's mass. Subtraction of Barbarossa's gravity makes the Pioneer Anomaly consistent with gravitation by a smoothly decreasing mass density.

I have ~100 references in major refereed journals and other authoritative sources. See my posts to the messageboard of Dr. Van Flandern's www.metaresearch.org website.

Earth's atmospheric twinkling is said to be a phenomenon of timescale ~ 1/30 sec. A lightning flash on Barbarossa might make a pointlike image almost as though Earth's atmosphere did not exist. This would amount to "active optics". Previously on this messageboard I estimated the likely brightness of lightning on such a giant planet (or cold brown dwarf), and found it competitive with reflected sunlight.

In Barbarossa's Cavern There Are No Stars (Part VIII)

Three Articles from the Annual Review of Astronomy & Astrophysics


1. Salpeter, "Formation & Destruction of Dust Grains", 1977. In Sec. 5.1, Salpeter cites observations by Zappala, of a circumstellar dust shell which is basically a solid sphere (inner diam. << outer diam.), though this is around a presumed mass-losing giant star. In Sec. 4.1, Salpeter says that the almost constant ratio of blue to red extinction, suggests that interstellar dust grain size distribution is about the same almost everywhere in the galaxy. On the contrary, the formulas and graphs of Kruegel's text say that for realistic "strong absorber" grains in the domain x < 0.5 (i.e., grain circumference < 0.5 * wavelength) extinction is proportional to grain cross-section times x; therefore the ratio of blue to red extinction = 0.55/0.44 = 1.25 is the same for any grain size distribution whatever, if only the extinction is mainly by small grains ( < 0.05 micron).

2. Aannestad & Purcell, "Interstellar Grains", 1973. In Sec. 3, p. 325, eqs. (6) & (7) (the "rho-g" factor of eq. (7) is a misprint), the authors say that (basically because of the "x" proportionality discussed above) extinction is proportional to the volume, hence mass, of dust whatever its grain size. Their lower bound of 0.5% dust::gas, agrees with the refined estimate of 0.8% in recent texts, and with the estimate 0.7% based on element abundances. They say, "This result applies equally well to grain mixtures. It can be applied with only slight modifications to 'grains' that are nothing but large molecules...".

3. McCray & Snow, "The Violent Interstellar Medium", 1979. In Sec. III.B, p. 228, the authors cite Shull's calculation that sputtering from high-velocity interstellar shocks, can reduce grain size to ~ 0.02 micron.

So, grains mainly moderately smaller than usually supposed, are quite possible. In (VII) above, I used formulas and graphs in Kruegel's text to show that such grains would reduce the apparent brightness of Barbaross'a nebula, to plausible levels.

Why GPS does not disprove the ether theory, and why this is relevant to Barbarossa.


(To the best of my knowledge and belief, this ether drift calculation essentially first was made by James Clerk Maxwell. It also paraphrases an earlier post of mine to this messageboard.)

Let all velocities be measured in a frame of reference centered on Earth. Suppose Earth sits in an ether drift of speed v << 1 (choose units so c = 1). Let an orbiting satellite have speed u << 1 parallel to the drift, and speed w << 1 perpendicular to it. The satellite's clock is slowed by a fraction:

0.5*((u-v)^2 + w^2) = 0.5*(u^2+w^2+v^2) - u*v

The clock on Earth is slowed by a fraction:

0.5*v^2

The difference between the clock rates is a fraction:

0.5*(u^2 + w^2) - u*v

Suppose Jim is a physicist who believes in textbook relativity. Jim will correct for what Jim thinks the difference between the clock rates is:

0.5*(u^2 + w^2) (i.e., half the square of the velocity vector relative to Earth)

Now let the satellite move along any path whatever, so that it is now more downstream in the ether, by a distance d. Jim doesn't know about the u*v term. That is, the clock is faster than Jim thinks it is, by a fraction u*v. The satellite clock now is ahead of what Jim thinks it says, by

integral(u*v*deltat) = v*d (because u*deltat is the distance, d, that the satellite moved downstream)

Now for simplicity suppose that the satellite's path was, to move from Earth, to a distance d straight downstream from Earth in the ether. The time really required for light to move from the satellite, to Earth, is d/(1 - v) = d * (1 + v) to first order in v. Jim thinks the satellite clock is v*d behind what it really says, so Jim thinks the time interval was d * (1 + v) - v*d = d. Jim might think this disproves the ether theory. Really, all it proves is that, even for a one-way test, the ether theory differs from textbook relativity, only to second order in v.

What GPS can do, is essentially repeat the Dayton Miller round-trip experiment (Miller found an ether drift, though somewhat smaller than expected; anyone who looks at the graph in Michelson & Morley's original paper can see that Michelson & Morley got roughly the same result as Miller; yes, a significant, positive result, though somewhat smaller than expected. I dare anyone to look at the original Michelson & Morley paper and say he doesn't see that. Recently Yuri Galaev in Kharkov, using two different, novel interferometric schemes considerably different from Michelson's, quantitatively confirmed Miller's result, at least to within an order of magnitude.) The discrepancy a 10 km/s ether drift would cause in a round-trip geostationary GPS time is equivalent to only 30,000km * 2 * 0.5 * (10/300,000)^2 = 3cm.

If the ether exists, it's likely to be related to the sun's gravitational field. Then, it's also likely to be related to the gravitational field of a fundamental particle. The distance from the sun, at which the sun's field becomes weaker, than that produced inside the most compressed possible proton (Gaussian distribution of deBroglie waves, following Merzbacher or other quantum mechanics texts) is 52.6 AU. Pioneer10 showed transient abnormal signals at this distance (JD Anderson's best explanation was a gravitational encounter with Edgeworth-Kuiper belt objects there, but quantitatively the frequency shifts were too big, for too long, to be due to acceleration from any plausible such encounter).

It so happens that the gravitational escape energy of electrons at this distance from the sun, is roughly their mean thermal kinetic energy at the "Cosmic" Background temperature. Depending on the details of the mechanism, various factors of order unity might be involved. The "Cosmic" Background temperature, is basically the gravitational potential of an electron at the surface at which the sun's gravitational primacy ends.

Primordial pinheads aside, the only known object big enough and symmetrical enough to explain the "Cosmic" Microwave Background, is the sun with its fields. The first (dipole) and some higher Maxwellian moments of the CMB distribution, are significantly correlated with the plane of the ecliptic. Only solar system influence, by three or more bodies, can explain this. Planets distort that mathematical surface at which the macroscopic gravitational field becomes equal to some small critical value; thus the gravitational potential of an electron at this surface also varies, due both to the gravitational potential of the planet, and to the distortion of the mathematical surface, due to the gravitational field of the planet. A brown dwarf outside the surface is best for causing a dipole; planetesimals near the surface best for local variations.

Please don't ignore the data merely because they lack an explanation consistent with textbook relativity theory. Here's another synopsis I've been providing to inquiring scientists:

"The 'ether' is like an approximately spherical pot of water. The sun is in the center of this 'water'. The boundary of this 'water' is at 52.6 AU. There's some activity on this boundary, on this distant 'movie screen' all around the sun. We call this activity the 'Cosmic' Microwave Background. (Someday we might be able to detect a 'CMB'-radiating surface around Sirius. The radius of the 'movie screen' sphere goes as sqrt(M), the 'CMB' temperature as sqrt(M), the absolute magnitude of the 'CMB' thus as M^3, but the absolute magnitude of the star as M^(3 or 4), so the easiest case is a bright nearby star, of mass enough greater than the sun's, that its 'CMB' curve is distinguishable.) The position or 'locus' of the 'screen' is determined by some critical value of the macroscopic gravitational field strength vis-a-vis the internal gravitational fields of fundamental particles. This critical equation allows unknown processes of some kind to occur. The screen isn't quite a perfect sphere, nor is the 'activity' on it everywhere exactly equal. But it's close to perfect, because the sun has almost all the mass in the solar system, so the gravitational field is almost perfectly symmetrical. Maybe the sun provides the CMB energy. Maybe something else does. But the sun does define the locus of the activity, except for small asymmetries caused by smaller gravitating solar system bodies."

In Barbarossa's Cavern There Are No Stars (Part IX)

The Johnson (c.1964) vs. Harvard (c.1908) magnitudes near Barbarossa's 1964 position


As detailed above, in regions near Barbarossa, USNO-B blue magnitudes were dimmer c.1983 than c.1954, and USNO-B red magnitudes dimmer c.1987 than c.1954 (these years can't be exact, because several overlapping plates were available to compute magnitudes). Interpolating between tested regions, the greatest blue dimming occurred about 3 deg ahead of Barbarossa and the greatest red dimming about 3 deg behind, along Barbarossa's track. During the time interval, Barbarossa moved 4 deg. So, the fastest increase in extinction, seems to be about 3 + 4/2 = 5 deg ahead of Barbarossa for blue, and about -3 + 4/2 = a degree behind Barbarossa, for red.

One explanation, would be that Barbarossa's cloud is nonhomogeneous: the most red-extinguishing material lags behind the most blue-extinguishing. The real explanation may be even stranger. Historical photometry shows that Barbarossa's cloud has no effect on the magnitude of stars of approx. Type G0. However, exclusively in the region of Barbarossa, bluer stars are dimmed, and redder stars are brightened, by amounts an order of magnitude greater than can be explained by error in zenith angle correction. The red dimming following Barbarossa, might be due to replacement of Barbarossa's cloud by some complementary medium which extinguishes red more than blue.

Methods. VizieR's online documentation of Harvard's Henry Draper catalog, gives its vintage as 1918-1924, but to secure data even older and of more certain vintage, I used the Harvard magnitudes listed, to 0.1 mag precision, for the stars on pp. A413-A415 (RA 10:26:15 to RA 11:40:43, B1900) of the USNO "Catalogue 4526 Stars" (Publications of the USNO, 2nd ser., vol. 9, pt. 1, published 1920 but dated 1917). The listed Harvard magnitudes mostly were from "Annals of the Astronomical Observatory of Harvard Coll." vol. 50, 1908, or vol. 54; many magnitudes were personal communications, from Prof. Pickering of Harvard, to the USNO. So, no magnitudes were later than 1917; many were earlier than 1908. An even older USNO publication (House Misc. Docs., 2nd ser., v. 2672, 1884) contained the USNO "Catalog of Stars 1845-1877"; I eschewed these because they took their magnitudes from various European catalogs, and often differed > 1.0 magnitude from later measurements. Only one difference between the Harvard magnitudes, and the Johnson (1966) online photometry catalog magnitudes, was truly a non-normal statistical outlier; this printed Harvard magnitude also differed from the online Draper catalog, so I used the Draper value, which was more plausible. I checked several other magnitudes whose differences from Johnson were greatest; only one more, which happened to be the magnitude with the second biggest difference from Johnson, differed from Draper (listed to 0.01 mag) by more than rounding error; likewise I adopted the Draper value here. I noticed several Boss (1928) "San Luis Catalogue" magnitudes, which Boss said were taken from Harvard, did agree exactly with the corresponding Harvard magnitudes printed.

Those stars on the three pages, which also were in HL Johnson's online "UBV Photometry of Bright Stars" (1966) catalog, numbered 61, and comprised my study. Johnson's observations (Comm. of the Lunar & Planetary Lab. vol. 4, pt. 3, #63) were dated approx. 1964.0 +/- ~ 1 yr and were taken at 32deg N lat in Arizona. He also used Cape, South Africa, observations by others; the text indicates these were taken in 1963 or 1964, and that no use was made of any observations from prior to 1955.

Spectral types were taken from the online Jaschek catalog when found (most); otherwise from the recent online Kharchenko catalog (many). All the Kharchenko spectral types were checked against Draper; all agreed within the 1/2 color rounding of Draper.

I found five of the 61, listed as variable in the "Bright Star Catalog 2000.0", vol. 2. Information given on these five, suggested amplitudes not big enough to require expulsion from my study.

Results. The magnitude changes, c.1908 to c.1964, within each color type, B through M, were fitted by exhaustive computer search, to a function of the star's Declination difference from Barbarossa's orbit. Barbarossa's orbit was approximated by a Mercator projection line through the (B1900) 1954 and 1986 center-of-mass positions. The function used, consisted of a constant term, plus a normal curve centered on Barbarossa's Declination for that star's RA. The three adjustable parameters giving the least-squares fit are:

magnitude change, Type B: -0.20 + 0.47 * exp(-deltaDecl^2/12^2) n=3
magnitude change, Type A: -0.10 + 0.10 * exp(-deltaDecl^2/11^2) n=12
magnitude change, Type F: -0.09 + 0.04 * exp(-deltaDecl^2/16^2) n=6
magnitude change, Type G: -0.12 - 0.04 * exp(-deltaDecl^2/23^2) n=14
magnitude change, Type K: -0.14 - 0.12 * exp(-deltaDecl^2/6^2) n=21
magnitude change, Type M: -0.11 - 0.19 * exp(-deltaDecl^2/14^2) n=4
(for Type M, one outlier was discarded)

Note that the denominators, 12^2, etc., imply that the standard deviation of the normal curve, in degrees of Decl, is 12/sqrt(2), etc. The weighted mean standard deviation is 9.1deg.

A Type M star near the N pole, with a moderately big magnitude change, tended to force a very large standard deviation (100deg) for best fit. So I decided the exclusion of this outlier would give a more representative result.

The linear variation of the coefficient of the bell curve term, proves the statistical significance of this result. The magnitude change, Johnson vs. Harvard, was different in the neighborhood of Barbarossa. Along this section of its track about 20 deg long, the "track effect" averages about 18 deg wide. By interpolation (using average number types for each color), there is no "track effect" for type G0; that is, no effect for starlight of the composition of sunlight. Blue undergoes extinction here relative to other parts of the sky, but red undergoes what amounts to "negative extinction": red stars brighten more here, than in other parts of the sky.

This is not due to Barbarossa's track acting as a proxy for the equator and revealing errors in zenith angle correction. Boss, in his 1928 catalog, gives 0.25 * (sec(z) - 1) for this, noting that this value also applies to the "northern Harvard measures". (Boss also gives the useful figure of +/- 0.08 mag error for one visual magnitude observation as done by his team 1909-1911; he says at least two, sometimes three, observations were made even for his hurried program.) With access to Cape observations, surely no Harvard magnitudes are based on observations lower than the traditional limit of z = 45. Even such observations would be extinguished only 0.10 mag for yellow light.

The difference between B and V band extinction would be only 0.024 mag, using the standard 4.16:1 ratio. The difference between Type A Visual and Type G Visual extinction can be estimated from the lambda^(-1) law, by replacing the Visual sensitivity window with the Type G spectrum as a proxy, considering the Type A0 spectrum to be the same as Type G2 but transformed to 9500K from 5800K (temperatures used by Johnson, Comm. Lunar & Planetary Lab vol. 3, #53, p. 73+, 1965), and expanding both factors of the convolution's integrand as Maclaurin series. For small changes in star temperature, the extinction is proportional to sqrt(temperature). This approximation says there's 28% more extinction in V for Type A0, than for Type G2. Again using the average type numbers in my sample's colors, this implies only 0.027mag more extinction for my Type A than my Type G, even at 45deg zenith angle. Likely, Johnson corrected for color differences in extinction, and Harvard (like Boss in 1928) didn't. This would brighten Johnson's Type A, vs. Type G, in the Barbarossa region, by 0.027 mag, not dim it by 0.14 mag as observed. The difference observed is five times too big and of the wrong sign. So, zenith angle isn't the explanation.

Another possible explanation is that Harvard magnitudes might consistently have relied on Cape observations for stars south of the celestial equator. Few stars in my sample of 61, are more than 10 deg south of Barbarossa's track; so, the sample might be basically Boston-measured stars north of about Decl +5, and Cape-measured stars from Decl +5 to -15, i.e., along the seemingly affected region. Systematic differences between Johnson and Harvard are explainable by the historical evolution of photometric calibration; but it was not necessary for Harvard and contemporary Cape calibrations to differ. To quantitate this possibility, I also tabulated, as above, stars between Decl 0 and -10 on pp. A383-386, A419-420, & A425-432 of the abovementioned 1917 (pub. 1920) USNO catalog. These regions were chosen randomly to get three roughly equal-size samples all > 20 deg from Barbarossa and > 20 deg from the galactic equator.

This control study comprised 31 stars. For convenience, I used only Kharchenko spectral types. No magnitude changes were big enough to be statistical outliers; no Harvard magnitudes were checked against Draper; the variable star catalog check was omitted. The standard error of the mean, for the magnitude changes in the most common colors, A & K, was 0.018 & 0.025 mag, resp. Subtracting the first, all-sky, difference term (see above), gave the local effect:

Type B: +0.22 mag
Type A: +0.05
Type F: +0.10
Type G: 0.00
Type K: -0.016
Type M: -0.08

Large statistical uncertainties for some of the Types, suggest replacing these numbers with a linear interpolation between Types A & K:

Type B: +0.07 mag
Type A: +0.05
Type F: +0.03
Type G: +0.006
Type K: -0.016
Type M: -0.04

(These quantities would practically equal the coefficients of the bell curve terms, because the borders of these control regions are at only 0.5 standard deviation from Decl -5.)

The difference in local effect, between Types A & K is 0.066/0.22 < 1/3 as big as for the Barbarossa region sample. Even at the extreme choice of error bars (A up and K down), the difference is only (0.066 + 0.018 + 0.025)/0.22 = 1/2 as big. So, no more than half the regional effect near Barbarossa, on the change in magnitude between c.1908 and c.1964, can be explained by Cape calibration and other Declination effects.