3. Discussion and conclusions
The X-ray luminosity of Cl 0500-24 is very moderate. Actually, it is quite small for a "rich" cluster of galaxies. The Abell Cluster A370 (at a comparable redshift), which optically appears about equally rich as Cl 0500-24, is roughly ten times more luminous in X-rays (Lea & Henry 1988; Fabricant et al. 1991; Bautz et al. 1994). Does this show that there is a wide spread of X-ray luminosities for clusters that appear similarly rich optically?
We argue that the X-ray emission of Cl 0500-24 originates mainly from the northern subclump. But even when comparing the bolometric luminosity of Cl 0500-24 only with the velocity dispersion of the northern subcluster, it is unexpectedly low (Edge & Stewart 1991b). This indicates at least a moderate spread in the optical richness versus X-ray luminosity relation.
There are other examples that illustrate this, e.g. the two distant clusters Cl 0939+4713 and Cl 0016+16, which are both optically very rich (Dressler 1994; Koo 1981) but have quite different bolometric X-ray luminosities (Schindler & Wambsganss 1996; Neumann & Böhringer 1996). Various properties of these two clusters, along with those of Cl 0500-24, are compiled in Table 1 for comparison.
The subcluster around galaxy C is even less X-ray luminous. Judging from optical data alone, it is well justified to assign the centre of the subcluster with redshifts peaking at (cf. introduction) to the position of galaxy C. Galaxy C is the most luminous galaxy of all galaxies with redshifts around , and its velocity is very close to the maximum of their velocity distribution. To be more quantitative about the X-ray luminosity of subcluster C, we subtracted a spherically symmetric profile (see Fig. 4) centred on the X-ray maximum from the total X-ray emission. The remaining X-ray flux at the position of C has a count rate of counts/s/arcmin2, above a background of counts/s/arcmin2. If subcluster C has any X-ray emission at all, it is extremely small. Scaling from the difference in measured velocity dispersions, the mass difference of the subclusters is by far not sufficient to account for the different X-ray luminosities. Subcluster C must have a much lower gas mass fraction than subcluster N.
Given the bimodal velocity distribution of the galaxies in Cl 0500-24 (cf. Introduction), there are two possibilities for the spatial distance of the subclusters along the line of sight. One possible scenario is that the two subclusters have a physical distance which is smaller than that corresponding to the redshift difference, and are in the process of colliding. The other possibility is that the two subclusters are relatively unrelated and the different velocities are caused mainly by the Hubble flow. In this case they have a distance of more than 60 Mpc.
Unfortunately, the X-ray morphology does not give a final answer to the question whether the two subclusters are interacting or not, for two reasons: Firstly, the HRI image is affected by statistical fluctuations as it contains only 440 source counts. Secondly, a collision along a direction close to the line of sight cannot be easily distinguished from a virialized cluster by looking at the X-ray image only (Schindler & Müller 1993).
As the relative velocity of the subclusters of almost 3200 km/s is very high, in a merging scenario the subclusters would have to pass right through each other along the line of sight. In fact, the projected distributions of the galaxies with redshift preferentially north, and with preferentially south indicates, that such a collision could not be exactly along the line of sight, but with some angle relative to it. So the velocity difference is only the contribution projected onto the line of sight, and the real 3dim relative velocity would be even larger than the measured difference. N-body models show that relative velocities around 3000 km/s or more are reached only during a very short period of time as the subclusters pass through each other (Schindler & Böhringer 1993; Huss et al. 1996). If the subclusters are just passing through each other, an X-ray emission enhanced by about a factor of two relative to a quiescent state is expected during a collision (Schindler & Müller 1993). However, given the low observed X-ray luminosity, and the fact that it is very unlikely that we see Cl 0500-24 exactly in this extreme short phase corresponding to such an encounter with high relative velocity we conclude that it is extremely unlikely that the two subclumps are currently undergoing such a collision. Due to these reasons we conclude that the scenario of two non-interacting subclusters at a large distance is much more likely.
This view of two unrelated subclusters that appear as one richer cluster due to the projection is also supported both by the morphology and the luminosity of the X-ray emission of the cluster. The X-rays originate from regions that are related to the subclump around N with galaxy velocities peaking at . There is no significant detection from the supposed centre of the cluster near galaxy C (cf. Giraud 1990) or other galaxies at redshift .
New insight into the dynamical state of Cl 0500-24 may come from a temperature measurement with ASCA. During a collision the intra-cluster gas is heated (Schindler & Müller 1993) so that in this case a higher temperature than the 4 keV from the relation is expected.
The mass assigned by Infante et al. (1994) to the more massive subclump around N at is even higher than the X-ray mass determined here. That can be another indication that the X-rays trace the potential of only one subcluster while the southern subcluster around galaxy C apparently does not affect the X-ray mass determination. The mass found by Wambsganss et al. (1989) for the cluster from a simple gravitational lens model for the straight arc is higher than the X-ray mass determined here. This also fits into this picture of two subclusters qualitatively, because the gravitational lens effect integrates all the mass along the line of sight, and so unavoidably considers all the subclumps discussed here. For a more detailed discussion of the reason of the discrepancy between X-ray and lensing mass one has to wait for better X-ray data, e.g. a temperature determination from ASCA data, as well as deeper optical data, with the possibility to trace the cluster potential via weak lensing.
The total mass of ( within 1 Mpc which is about the virial radius according to the spherical collapse approximation (Gunn & Gott 1972; White et al. 1993a) is a relatively low value compared to the typical mass range of clusters of (Böhringer 1995). It is an order of magnitude lower than the mass of the nearby clusters Coma and Perseus (Böhringer 1994). The gas mass fraction of % inferred from this number is at the upper limit of the typical range for clusters of 10-30% (Böhringer 1995). This high fraction makes Cl 0050-24 another example for the so-called baryonic catastrophe (Briel et al. 1992; White et al. 1993b).
Combined, all the arguments mentioned above point into the direction that Cl 0500-24 consists of two concentrations with little direct interaction. Only one of the subclusters has X-ray emission. But even considering the fact that only part of the optically visible cluster is associated with the X-ray properties, the values for the X-luminosity and the derived total mass are comparably small.
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998