Leo Vuyk was intrigued by a new interesting paper by Mike Hawkins:
rule 8 of Baez's index. ;-)
Hawkins has observed power variations of 900 quasars with frequencies between inverse months and inverse decades. Because quasars at apparent distances 6 or 10 billion years differ by the redshift as seen through the frequency of the electromagnetic waves, the corresponding light variations should be slowed down or sped up by the same factors.
But they're not.
Hawkins offers two explanations. One of them tries to cancel the time dilation by the black hole growth. The size of the black holes inside the quasars depends on the redshift (age) in such a way that the spectrum ends up to be nearly identical.
The other explanation is that all these variations originate as the photons arrive here - by "microlensing". Well, I don't think it necessarily has to be "microlensing" but anything that modifies the light on its path. However, quasars may be driven by the accretion disks around galactic black holes - and their radius could actually be comparable to a light year so that the variations could be caused by the complex dynamics inside the accreation disk.
Of course, one needs to mention a more dramatic possible explanation, the heretical theory by Halton Arp claiming that the quasars are actually not super-far at all but they're connected to the nearby galaxies by ropes so all the previous conclusions derived from the observed quasar redshifts have been wrong: the redshifts originate as their "intrinsic redshift". :-)
While I still view this Arp's possibility as a remote one, it would seem foolish to me to reject it a priori, especially given the tantalizing observations such as Hawkins' ones. Hawkins' observations are not quite new: he published a nearly identical paper in 1993 and 2001.