Key points are not available for this paper at this time.
The ratio of angular diameters of the first three ranked galaxies in E and 50 aggregates define contrast parameters that correlate well with Bautz-Morgan (BM) cluster types. Interpreted as an apparent magnitude difference, 5 log 01/02 varies from 1.3 mag for Bautz-Morgan class I clusters to 0.4 mag for class III. The absolute magnitudes also change with BM class. Magnitude residuals from the Hubble diagram show that the first-ranked galaxy is absolutely brighter in class I clusters than in class III by AMv = 0.6 mag. However, the second and third ranked are fainter by 0.5 mag in class I compared to class III clusters. This startling, but well-determined, inverse effect suggests that the dominance of first-ranked galaxies in clusters occurs at the expense of the fainter members. Because both the Bautz-Morgan class and the first-ranked absolute magnitudes are independent of cluster richness, we argue that the Bautz-Morgan effect is more likely to be related to an initial condition of cluster formation than to later evolution by processes that depend on the rate of interaction of cluster members. New data, obtained by extensive counting, are given for the galaxy population (N) of all groups and clusters in our sample. There is no correlation of first-ranked absolute magnitudes with N at a level more significant than 1 , but a significant correlation does exist for second- and third- ranked galaxies. The Hubble diagram using the V magnitudes of first-ranked galaxies in 98 E and 50 groups, corrected for aperture effect, K-dimming, galactic absorption, Bautz-Morgan effect, and cluster richness shows good agreement with a linear redshift-distance relation, and has the small dispersion of a(Mv) = 0.28 mag for the distribution of horizontal residuals. The effect of the BM and the richness corrections on the value of q0 in our sample is negligible compared with the large errors of the current determinations. Only the grossest alternatives to q0 (apparent) = 1 t 1 (such as q0 = - 1 or q0 > 3) can be discarded from the data now available, no matter how the material is analyzed. New data for many clusters with large redshift (z > 0.4) are needed for a finer solution. However, the prediction of steady-state cosmology (q0 = - 1) is clearly at variance even with the present data. Subject headings: galaxies, clusters of - galaxies, photometry of
Sandage et al. (Wed,) studied this question.