Quasars, companion galaxies and Poisson statistics

Arp has presented a sample of quasars lying close to the companion galaxies of bright spirals, from which he estimates a value of 10-17 for the probability that the galaxies and quasars are sited independently on the celestial sphere; Browne, however, has found a simple fallacy in the statistics which accounts for about 10 of the 17 orders of magnitude. Here we draw attention to an obscure part of Arp's calculation which we have been unable to repeat; if it is carried out in what seems to be the most straightforward way, about five more orders may be accounted for. In consequence, it is not clear that the sample contains any evidence damaging to the popular notion that the redshifts of quasars indicate distance through the Hubble Law. (author)