3. Galaxy velocity distribution
We use the extensive spectroscopic catalogues of galaxies in this region to analyse several fields around each cluster with the goal of determining the three dimensional structure of the ABCG 85/87/89 complex. The catalogue by Durret et al. (1998) includes 551 galaxy velocities in the direction of ABCG 85; among these, 305 galaxies lie in the velocity range 13000-20000 km s-1, which are assumed to belong to ABCG 85. We also used five galaxy velocities for ABCG 87 from the ENACS catalogue (Katgert et al. 1997).
To analyse the distribution of velocities, we use a wavelet reconstruction as described in Fadda et al. (1997). This method provides velocity density profiles. The calculation of significance for features derived in the wavelet reconstruction is described in their paper. Mathematically, it is possible to perform analyses to very small velocity scales; however, the features found may not be significant. We will therefore use only the largest scales. Notice that even if we are not able to be very confident in each peak for each sample, the fact that the same velocity feature is observed in adjacent zones, using independent velocity samples, argues in favour of its true existence.
3.1. Velocity subsamples
We defined seven subsamples of galaxies chosen in circular regions on the sky (see Table 2 and Fig. 5). One is located at the center of the cluster ABCG 85 (named Sc) and two (named A89 and A87) are located at the positions of ABCG 89 and ABCG 87 as indicated by the NED database. Two control subsamples are chosen at the symmetrical positions of A 87 and A 89 relative to the ABCG 85 center: C87 is symmetric to A87, and C89 to A89. To these five samples, we add two comparison samples: C1 between A87 and C89 and C2 symmetric to C1 relative to the ABCG 85 center (see Fig. 5). The radii of all these circles have been chosen in each case to avoid superposition of possible structures, while obtaining a significant number of velocities in each sample.
Table 2. Characteristics of the seven samples.
Velocity clustering is observed between 13000 km s-1 and 32000 km s-1, as illustrated in Fig. 6. The velocity distributions for the control samples C1 and C2 are displayed in Fig. 7. The most prominent feature centered at about 17000 km s-1 corresponds to the ABCG 85 cluster. In Fig. 6, a second feature is seen at about 23000 km s-1.
3.2. The 20000 - 32000 km s-1 velocity range
3.2.1. A galaxy sheet between 20000 km s-1 and 26000 km s-1
The peak observed between 20000 km s-1 and 26000 km s-1 (Fig. 6) is particularly strong in the region of A89 (Figs. 6 and 8); this peak (hereafter ABCG 89b) has a velocity dispersion km s-1. Such a value is too large for a single relaxed group, thus suggesting either substructure or contamination by other structures. As Fig. 6 shows, the maximum between 20000 and 26000 km s-1 is not at the same velocity for all the samples. In particular, the peaks for the control samples C87 and C89 are higher: 24000 km s-1 for C87 and 25000 km s-1 for C89.
The clustering of velocities, together with the variation of the peak velocity, suggest the presence of a sheet of galaxies inclined with respect to the plane of sky. This sheet is not visible in the two control samples C1 and C2.
3.2.2. A structure at km s-1: ABCG 89c
A second peak is observed in the A89 velocity distribution at about 28500 km s-1 (see Figs. 6 and 8). An enhancement is also found at this velocity in the C1 control sample (Fig. 7), but no peak is present in the five other velocity samples in Fig. 6. This high redshift component of A89 has a velocity dispersion of = km s-1, typical of a galaxy group or small cluster. Obviously ABCG 89 is not a simple cluster, but rather a complicated superposition of structures.
3.3. The 13000 - 20000 km s-1 velocity range
We have analysed the distribution at various scales. However, due to the inhomogeneous number of galaxy velocities in the various samples distributed in various velocity intervals, the scales differ, depending on the number of galaxies in each sample.
3.3.1. A structure at km s-1: the "Foreground Group"
A peak in the velocity distribution at 14000 km s-1 is observed for Sc (Fig. 9, top and bottom panels), A87 (Fig. 10, bottom panel) and the comparison sample C89 (Fig. 12). This peak corresponds to the "Foreground Group" detected by using the Serna & Gerbal (1996) hierarchical method and identified on the X-ray image with the excess to the west of ABCG 85 (Durret et al. 1996).
3.3.2. Structures in the [15000-19000 km s-1 ] range
In velocity space for the Sc and C89 samples, a dip at km s-1 follows the "Foreground Group" (see Figs. 9 and 12). Notice that the galaxy samples A89, C87 and C1 contain no galaxy velocities smaller than km s-1, while this is not the case for the A87 sample (Fig. 10); in all these samples, most of the galaxy velocities lie between 15000 and 19000 km s-1.
In the case of Sc (top panel in Fig. 9), A89, C89, C87 and for the two control samples C1 and C2, a maximum is observed at km s-1. These velocity distributions are clearly not Gaussian for any of the samples.
At a smaller scale, in the case of Sc, the velocity density breaks into a two-maxima density distribution (Fig. 9, bottom panel) with a first maximum at km s-1 followed by a dip (the "Dip" in the following) at km s-1, then a second maximum at km s-1 followed by a shoulder at km s-1. It is interesting to notice that the velocity of this second maximum is very close to that of the cD galaxy.
Even if the relative levels of peaks and dips are not the same for both samples, the succession of features in the Sc velocity density distribution is very similar to that in the density profile of A87 (Fig. 10, top panel): a first maximum followed by the "Dip" then a second maximum.
In several samples we observe a similar succession of maxima and minima, such as the 14000 km s-1 maximum, a small dip at 15000 km s-1, the "Dip" at 16100 km s-1, and a maximum at km s-1 (Figs. 9, 10 and 12). A peak (at km s-1) corresponds in Fig. 10 (bottom panel) to the shoulder noted in the Sc sample (Fig. 9, bottom panel).
© European Southern Observatory (ESO) 1998
Online publication: June 12, 1998