Ever heard of the Lyman Alpha Forest?
Lyman Alpha Forest
There are clouds of gas between distant quasars and the Earth that absorb ultraviolet light at the wavelength of the Lyman alpha line of hydrogen at a wavelength of 122 nm. Quasars also emit a strong Lyman alpha emission line. But the absorbing clouds all have smaller redshifts than the quasar since they have smaller distances. As a result the absorption lines are all on the blue or shorter wavelength side of the quasar emission line.
The cartoon above shows a quasar with its Lyman alpha emission line redshifted from the ultraviolet into the red, and the Lyman alpha absorption lines from four intervening clouds appearing as orange, yellow and green-blue.
The figure below shows two actual quasar spectra. One is the nearby quasar 3C273 while the other is a large redshift object. This figure was adapted from Bill Keel's web site.
The clouds that produce the absorption lines are much more numerous at high redshifts so a very dense array of absorption lines results: the "forest".
We know that there are a small number of very big clumps of hydrogen in the distant Universe: the galaxies. We also know that smaller galaxies, the dwarf galaxies, are very much more common. Most of the clouds in the Lyman alpha forest are much less massive than dwarf galaxies and these small clouds are much more numerous. We can only see these very low mass clouds by the absorption they produce in the strongest line of the most abundant element: Lyman alpha. Thus by studying the Lyman alpha forest we can learn about the density fluctuations in the Universe on the smallest observable scales.
Note that if Arp were correct and quasars had a redshift much larger than the redshift due to their distance, then there should be a gap on the blue side of the Lyman alpha emission line before the absorption lines began. Such gaps are not seen. So if Arp were correct the Lyman alpha forest would have to be an intrinsic property of the quasar, which would be a very unlikely situation. Distant galaxies are seen which also show the Lyman alpha forest, so we know that the intervening clouds do exist. For Arp to be correct the intrinsic absorption lines would have to act exactly like the intervening clouds would act under the standard hypothesis that the quasar redshift is entirely cosmological.