wrote:
We consider the sun and its atmosphere as a spherical lens causing
deviation of stars light and also decrease of its speed in the lens.

There is no doubt of this -- it is measured all the time.

But do not deceive yourself into thinking this explains the observed
gravitational bending of light, and the Shapiro time delay. The solar
atmosphere cannot possibly do that:

a. Bending occurs at angles 90 degrees from the line-of-sight to
the sun, which means the observation ray never gets closer to
the sun than the earth, and near earth the solar atmosphere is
observed to be FAR too small to account for the measurements.

b. These effects occur EQUALLY for light rays of many different
wavelengths; the index of refraction of the solar atmosphere
varies significantly for the different wavelengths used.

All of the measurements are consistent with a model consisting of GR
plus a model of the optical effects of the solar atmosphere. The latter
alone is woefully inadequate.


Your attempt to "explain" lack of dispersion is wrong:
But how can the solar atmosphere, as the observations show, be
non-dispersive? Surely if we can consider the solar lens as a small
lens or prism in an optical laboratory and allow a narrow beam of
some non-monochromatic light to pass through it (not towards its
center), then we must expect dispersion of the light passed through
the (solar) lens due to its refraction in the lens. But that such a
dispersion is not observable when observing the stars light passing
beside the sun is because of this fact that it is not only a single
narrow beam of the light of a star that reaches the sun but numerous
beams of its light reach the sun parallel to one another. The reason
of their parallelism is that the star is distant from the sun very
much.
In this manner instead of a single beam which may pass through a lens
in an optical laboratory, we are here dealing with numerous parallel
beams. According to the justification related to Fig. 1, these beams
when passing through the sun's atmosphere (or in other words when
passing beside the sun) are deflected into different directions in
proportion to their distances from the sun's center. It is natural
that each beam is also dispersed in the solar lens simultaneous with
its deflection. Then, we shall have numerous differently oriented
deflected beams each of which simultaneously dispersed, after passing
of the beams through the sun's atmosphere. It is clear that different
dispersions of different deflected beams (related to the primary
parallel beams adjacent to each other) will be intermingled with each
other (eg the rays a and b in Fig. 6 that have different wavelengths
are mixed with each other due to their parallelism), and consequently
an earthy observer won't observe any dispersion but only the deflection
of the beam will be observable for him or her; indeed this is just
the same reason that why in an optical laboratory the phenomenon of
separation of different wavelengths of a sufficiently thick beam of
some non-monochromatic light is not observed in the middle part of the
beam after its refraction in a prism (or a spherical lens).

Yes indeed, in a simple laboratory looking at a thick beam bent by a
dispersive lens one will only observe the separation of light into
individual colors at the edges of the thick beam, and not at the center.
That's because near the center of the observed beam different colors
come from differet parts of the incident beam. But that is NOT how the
VLBI measurements of bending by the sun are made -- those measurements
accurately measure not only the intensity of each wavelength, but also
its direction. For that simple laboratory, if it were able to limit the
angle of the observed light well enough, it would indeed separate light
into different colors throughout the beam. The VLBI measurements do so,
but don't see any dispersion, so solar atmosphere cannot account for
this. And also, reasonable models for the optical effects of the solar
atmosphere near earth are FAR too small to account for the obsevations.



Tom Roberts