The statistics of bright galaxies

Bhavsar and Barrow (1985. Mon. No. R. astr. Soc. 213, 857) demonstrated the use of extreme value statistics for the problem of the distribution of first-ranked galaxy magnitudes, M1. A model in which these galaxies are drawn from two distinct populations -'special' and 'statistical' - describes the distribution of M1 very well. Here we further explore this assertion and calculate the values and probabilities of the related physical parameters. In particular we find that models in which the luminosities, rather than magnitudes, of the 'special' galaxies have a Gaussian distribution do marginally better. In this model special galaxies are present in 73 per cent of the richness 0 and 1 clusters. However, with the available data, statistical methods would not rule out the possibility that all clusters had a special galaxy. (author)