Spectra of quasars
A) mportant observations are neglected by Weinberg:
i) Burbidge  and Karlsson studied redshifts (relative frequency shifts of star spectra) of quasars selected as isolated in space, and having low redshifts. They observed that redshifts of most of them may be written Z(n)=nK, where K is empirical Karlsson's constant K=0.061, computed from thousands of observations, and n an integer 3, 4, 6, ... These authors tried to use this formula to study spectra of quasars, without good results.
Karlsson's formula may be written:
Z(p,q)= p(3K)+q(4K), where p and q are nonnegative integers.
ii) It was remarked that 3K (resp. 4K) are redshifts which transform Lyman beta (resp. gamma) frequency f(b) (resp. f(g)) of H atom into Lyman alpha frequency f(a). This shows that H atom seems play a role in redshift of these quasars. Thus, assume that low pressure hydrogen is at a temperature between 5000 and 40000 K, so that it is in ground, 1S, atomic state.
How can these "quantized redshifts" be produced? After an absorption by atomic hydrogen of beta (resp. gamma) line by a continuous spectrum, absorbed line is redshifted until it reaches alpha frequency. Thus, it appears that it is the lack of alpha absorption which stops redshift, that is that redshift results from an invisible alpha absorption: by redshift, region of spectrum being absorbed at Lyman alpha frequency changes permanently so that absorption is low and width of line written into spectrum corresponds to redshift: very broad, weak absorption line is invisible.
Remark that with old astrophysical theory, absorption of observed sharp, saturated lines results from H atoms in filaments which must be: - thin, so that cosmological redshift in filament is negligible; - low pressure to avoid collisional broadening, thus thick to obtain observed saturation of absorption.
During the stop of redshift, gas spectrum is written into light; assuming a very low pressure, lines are sharp. They are saturated if stop is long.
iii) Spectra show many sharp saturated lines named "redshifted Lyman alpha lines" . It is strange that no lines are named "redshifted Lyman beta or gamma line". By Rydberg's formula, assuming that, as in a Doppler redshift, achromatic redshift multiplies all frequencies by a constant, compute frequencies of beta and gamma lines having same redshift than a redshifted alpha line. One out of both frequencies is frequency of a line named (redshifted) Lyman alpha line; observed line is not a doublet, it is sharp, coincidence is perfect. Applying again the same process, we finally obtain Lyman beta or gamma frequency. Unshifted Lyman beta line is observed, Lyman gamma not, because its frequency is out shifted emission band of star.
As previous process is a bijection, opposite process generates by steps the spectrum of "Lyman alpha lines":
-1- Assume a "white spectrum" up to far UV or X.
-2- Absorb Lyman alpha, beta and gamma lines by H atoms (possibly not pure: also metal lines,..).
-3- Redshift all frequencies until an absorbed line of start frequency f reaches Lyman alpha frequency f(a) of H: assuming no chromatic dispersion, frequencies are multiplied by f(a)/f.
-4- No more redshift, all gas frequencies are absorbed.
-5- If it remains energy at beta frequency, a slow redshift puts absorbed line off alpha frequency, go to -3- .
-6- multiply frequencies by a dispersion function which modifies mainly low frequencies.
B) Physics of redshift.
Many authors proposed redshifts by INCOHERENT interactions of light with gas. This cannot work because these interactions, similar to interactions of sunlight with colliding molecules which provide blue sky, blur images.
Light-hydrogen interaction must be COHERENT, and exchange energy, so that it is named coherent Raman interaction. But, exciting beam and Raman scattered beam have different energy, thus (except in optically anisotropic crystals), different frequencies correspond to different wavelengths, so that light scattered along exciting beam has variable phases, it self-absorbs.
Happily, there is a trick: Fourier computation shows that chopping a monochromatic light beam into pulses broadens its spectrum, so that exciting and scattered beam may have a region of equal frequency in which light beams interfere to an intermediate frequency: Frequency of exciting beam is shifted. This effect, named "Impulsive Stimulated Raman Scattering" (ISRS), may be computed precisely, giving conditions set by G. L. Lamb: "length of light pulses must be shorter than all implied time constants". In chemistry, ISRS uses femtosecond laser pulses, to study dynamics of reactions. Ordinary (incoherent) light is made of ~1 nanosecond pulses, computation shows that observation of shift (order of magnitude inversely proportional to cube of length of pulses) requires astronomical paths.
This result takes Lamb's conditions into account:
- Period of quadrupolar (Raman) resonance must be longer than 1ns, that is frequency must be lower than 1 GHz. Hyperfine resonance frequency in ground state (1S) of H at 1.420 GHz is too high. Resonances in excited states work, in particular in 2P state reached by Lyman alpha absorption. Higher states work also, but weakly because frequencies are too low.
- Collisional time must be longer than 1 nanosecond, so that pressure must be very low. Collisions are too rare to de-excite efficiently hyperfine levels excited by Raman absorption. But, there exist always a possible interaction with low temperature background radiation. Thus efficient effect is parametric, made up of several ISRS. This parametric effect named "Coherent Raman Effect between Incoherent Lights" (CREIL) increases entropy of sets of light beams by coherent frequency shifts.
C) Structuring space.
Space is split into spherical regions in which, either there is no energy at Lyman alpha frequency, or atoms are pumped to 2P state.
In this last case, gas becomes an amplifier at Lyman alpha frequency. A flash bursts when amplification coefficient is large enough, a relaxation process starts. The flash is seen as flare close to star. By competition of modes, flash absorbs also energy from observed beam in which it writes an absorption line.
D) Conclusion. Explanation of quasar spectra requires only ordinary physics well verified in labs.
Menzel's point of view , that coherent interactions are negligible in nebulae is rejected.
-- Hubble's law does not evaluate distances, but column densities. Thus distances are exaggerated by hot regions in which density of excited H atoms may be large. Consequently:
Distances of spiral galxies, thus their size are exaggerated: no need of dark matter for their stability.
This exaggeration of distances inflates bubbles in maps of galaxies which become spongious.
and so on ...
-- As CREIL is an interaction with matter, as refraction, it has a dispersion observed and attributed to a variation of fine structure constant!
-- CREIL increases entropy of a set of light beams. In particular, it transfers energy from sunlight to weak microwave waves exchanged with probes. This blue-shifts signals used to measure distance of Pioneer 10 and 11 probes. This happens where solar wind cools into excited H atoms, between 5 and 15 AU.
Negligence knocks not only ISRS but also an other important coherent interaction of light with matter: superradiance.
For instance, Strömgren spheres, made of protons and electrons around a star, are surrounded by a spherical shell of excited atomic hydrogen able to amplify light beams, mainly those, tangent to sphere, which have the longest path inside shell: We see limbs of Strömgren's spheres:
-- by amplification of brightness of background stars (H. Arp);
-- by observation of bubbles;
-- by observation of pearl collars.
In an equatorial plane of SN 1987A, planets absorb or scatter radiation of star, Strömgren's sphere becomes an hourglass whose we see three limbs. Light radiated by star is absorbed by hourglass which reemits it after various interactions; hourglass transforms light as ruby of a flash pumped laser. This fast transformation explains that SN1987A disappeared when its collars appeared.
No need of heavy stars well aligned with Earth: Strömgren's spheres are more abundant!