4 Discussion

The three-dimensional distribution of the galaxies in the redshift sample is shown in Fig. 8 in the form of wedge diagrams, where the angular position of each object on the sky has been converted to proper distance from the line of sight, appropriate for the given redshift in a $\Omega_{\rm M} = 1$, ${\Omega_\Lambda} = 0$ world model. The cluster Cl0024+1654 shows up clearly as a sheet at $z\simeq0.4$. The expanded views show that Cl0024+1654 is not a simple isolated cluster but that there is a foreground "clump'' at z=0.38, superimposed onto the main cluster and connected to the latter via a narrow bridge. In a companion paper (Czoske et al. 2001) we discuss an interpretation of this structure as being due to the foreground cluster having passed through the main cluster.

Figure 5: V-I colour-magnitude diagrams. The top-left diagram shows the full photometric catalogue, the top-right diagram the full spectroscopic catalogue. The next three diagrams split the spectroscopic catalogue according to redshift, showing foreground and background galaxies as well as galaxies around the cluster redshift $z\sim 0.39$. Note the clearly visible cluster sequence at $V-I \simeq 2$ in the latter diagram. The bottom-right diagram shows the members of the newly discovered group of galaxies at $z\sim 0.495$ (see Sect. 4). The parallelogram marks the subsample used in the completeness map (Fig. 7), 20 < V < 23, 0.6 < V-I < 2.4.

Figure 6: Completeness of the spectroscopic survey in V magnitude. For each galaxy from the catalogue this is given as the ratio of the numbers of galaxies in the spectroscopic and photometric catalogues in a given bin width centered on the magnitude of the galaxy. Pluses mark galaxies taken from the whole survey area (as outlined in Fig. 7), crosses galaxies within $3\hbox {$^\prime $ }$ of the cluster centre. In the former case, a bin width of 0.5 mag was used, in the latter a bin width of 1 mag.

Figure 7: Map of the completeness variation of the spectroscopic catalogue as gray-scale with overlaid contours. The completeness at any point is determined in a circular top-hat encompassing the 10 nearest neighbours in the spectroscopic survey; the map is smoothed with a Gaussian of width $30\hbox {$^{\prime \prime }$ }$. Contour lines are spaced in 10% steps. The 50% contour is marked by a bold line, contours at less than 50% are drawn in black, higher contours in white.

Figure 8: Three-dimensional distribution of the objects in our redshift catalogue. In the two upper panels the objects are projected onto the right ascension axis, in the lower two onto the declination axis. The upper panel of each pair shows the large-scale distribution from z=0 to z=1, the lower panel an expanded view of the environment of the cluster Cl0024 itself. The dashed line marks the direction towards the potential perturbation detected by Bonnet et al. (1994). Two groups at $z\sim 0.495$ are marked by rectangles. The conversion from angular position on the sky to proper transverse distance was done assuming an Einstein-de Sitter Universe with $H_0=100\,{\rm km}\,{\rm s}^{-1}\,{\rm Mpc}^{-1}$.

We find a pair of compact groups of galaxies at z=0.495 about $10\hbox{$^\prime$ }$ to the north-east of the centre of Cl0024+1654 (see Fig. 8 at $x\simeq1$ Mpc, $y\simeq2$ Mpc). The northern group includes 8 galaxies, centered at $574\hbox{$^{\prime\prime}$ }$ north and $203\hbox{$^{\prime\prime}$ }$ east of the cluster centre, the southern group includes 6 galaxies centered at $369\hbox{$^{\prime\prime}$ }$ north and $250\hbox{$^{\prime\prime}$ }$ (median positions); the projected distance between the groups is thus $\sim$ $740\,h^{-1}\,{\rm kpc}$. The mean redshifts are $\overline{z_{\rm N}}=0.4921$ and $\overline{z_{\rm S}}=0.4970$, the formal velocity dispersions $\sigma_{\rm N}=657\,{\rm km}\,{\rm s}^{-1}$ and $\sigma_{\rm S}=647\,{\rm km}\,{\rm s}^{-1}$. Student's t-test rejects the hypothesis that the two groups have the same mean redshift at 99% confidence, so we assume that we are really seeing two separate groups. The velocity dispersions are presumably enhanced by tidal interaction between the groups.

Figure 9: "True'' colour images of the apparent centres of the northern (top) and southern (bottom) groups of galaxies at $z\simeq 0.49$. The images were created from the V and I band images, the green channel is an average of these two images. Supposed member galaxies of the group are conspicuous by their yellow colour, corresponding to $V-I\sim 2.4$. Note the blue arc-like structure around the galaxy at z=0.4907.

Figure 9 shows colour images of the northern and southern groups created from the I- and V-band CCD images. The galaxy at z=0.4907 is surrounded by three objects of similar, blue colour. It is tempting to interpret this group as multiple images of the same background object. In this case, using the curvature radius ($\sim$ $5\hbox{$^{\prime\prime}$ }$) as an estimate for the Einstein radius and $z_{\rm s}=1$ as a rough guess for the redshift of the background source, we obtain $5.8\times 10^{12}\,h^{-1}\,M_{\odot}$ for the mass within this radius.

Another overdensity in Fig. 8 occurs at $z\sim0.18$. These galaxies are however distributed fairly uniformly across the field with no obvious spatial concentration and are therefore just part of the general large-scale structure in the Universe.

In the first detection of a coherent shear field around a cluster of galaxies, Bonnet et al. (1994) found a signal to the north-east of the centre of Cl0024+1654, indicating a concentration of mass at a point where no overdensity of galaxies is apparent in the two-dimensional images. The direction to this dark "clump'' is indicated by a circle in Fig. 4b and by the dashed line in Fig. 8. There is no significant over-density along this line which could explain the spatial tightness of the signal observed by Bonnet et al.