[response to Rowan Hamilton... and isn't your first name supposed to be "William"?] Ultra high energy cosmic rays (say >50 GeV) cannot come from "dense bodies like neutron stars or black holes" at least not directly. There simply is not any high enough energy production process available. The measured ray energy is far exceeding the rest mass of whatever particle the ray is made of. With a black hole, you are never going to get above order 1 nuclear rest mass worth of energy, say 10 GeV should be an upper bound. So it is clear that they either (a) must be accelerated by something that acts like a particle accelerator, or (b) come from new physics. The highest energy cosmic rays we've seen are up to 10^20 eV and far exceed what human-built accelerators can produce, which is about 10^13 eV. It has been conjectured that MHD effects in multiple plasma shockwaves, acting over length scales of many light years, and perhaps radio signals from pulsars (or quasars or active galactic nuclei) might somehow effectively build particle accelerators. Your claim that "positrons should annihilate at a much higher rate as they propagate across the universe" is true, but fails to explain why the positron fraction in rays should depend on the ray-energy in the observed decreasing-then-increasing manner. The characteristic energy scale for reactions involving e+ and e- should be of order 511 KeV, which is far below the energy scale of the "valley" in the curve, so that selective annihilation would seem to be ludicrous as an attempted explanation. Further, one could consider energy dependence of cross sections for electron or positron collisions with protons, but again that energy scale (0.5-2 GeV) seems wrong to explain the observation (although less wrong). Collisions with photons, or EM accelerator effects, should act the same on both e+ and e-, so they cannot explain it. Here is how one could try to explain the observation without invoking new physics: (a) postulate there are two cosmic ray creation processes, one (I) creates e+ e- pairs, other (II) creates just hot electrons, but not pairs. (b) The second process is hooked up to a custom accelerator which nearly always outputs electrons with energies in the "valley" range, 7 GeV. (c) the first process is hooked up to all accelerators with all output energies. As a result of these hypotheses, positrons would be a constant (or decreasing with energy) fraction of the cosmic rays output by process I, but process II gives you a lot of extra electrons near 7 GeV, causing the positron fraction to exhibit a valley there. I would think that explanation would probably yield visible signatures, which, if they are not present, should rule it out. It looks to me like this experimental result is very interesting and may indicate the presence of new physics. And if it is old physics, then it's old physics that needs a lot more understanding. -- Warren D. Smith http://RangeVoting.org <-- add your endorsement (by clicking "endorse" as 1st step)