3 Analysis of the specific sources

In this section we analyze in details the more probable associations of galaxy clusters with the unidentified EGRET sources which are listed in Table 1: all of these galaxy clusters are found within the $95 \%$ confidence level position error contours of the associated EGRET source. As for the flux of each source, we report only the first entry given in the Third EGRET catalogue (Hartman et al. 1999) providing also the viewing period (VP) of the source detection and the significance level, (TS)^1/2, of the detection. The reader may refer to Hartman et al. (1999) for the full list of information about the EGRET source under consideration.

Figure 4: We show here the maps of the galaxy clusters which are more probable candidates for the associations with unidentified EGRET sources: 3EGJ0038-0949 associated with Abell 85 (upper left), 3EGJ0253-0345 associated with Abell 388 (upper right), 3EGJ1424+3734 associated with Abell 1914 and Abell 1902 (mid left), 3EGJ2218-7941 associated with Abell 1024S and Abell 1014S (mid right) and 3EGJ1310-0517 associated with Abell 1688 (bottom left). The cluster positions and X-ray brightness contours (when available) are superposed to the maps of the EGRET sources. The intensity scale of the EGRET maps goes from black (minimum) to white (maximum).

3.1 3EG J2219-7941

This EGRET source has been detected with a (TS)^1/2 = 4.4 in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV}) =(13.5 \pm 3.6) \times 10^{-8}~{\rm cm}^{-2}~{\rm s}^{-1}$ with a power-law spectral index $\gamma=2.50 \pm 0.29$. Its flux does not change significantly over all the VPs with the exception of the VP = 10.0 in which a flux increase by a factor $\sim$2 has been recorded; however, this flux is consistent with the flux detected in the other periods at the $2 \sigma$ confidence level, so that it can be considered a stationary gamma-ray source (see Fig. 1). There is no identified gamma-ray source counterpart for this EGRET source and, thus, it is a good candidate for the galaxy cluster association. Two Abell clusters fall within the $95 \%$ confidence level position error contours of this source: Abell 1014S and Abell 1024S. The EGRET source map is very broad with an effective radius of the $95 \%$ confidence level position error circle of $\theta_{95} = 0.63$ $\deg$. The elongation of the EGRET source probability map is aligned with the position of the two clusters, suggesting a possible contribution to the gamma-ray emission from both clusters (see Fig. 4).

There is no definite X-ray information on these two galaxy clusters available in the literature. However, we found that both these clusters have some radio information. Given the low declination of the sources, they are not covered by the NVSS. So, we looked for radio sources associated with the cluster in the SUMMS survey (Bock et al. 1999). We found that Abell 1014S at z=0.048 (Abell et al. 1989) has several bright SUMSS radio sources located within its Abell radius, $\frac{1.7}{Z}$ arcmin. A preliminary analysis of the RASS field also yielded a flux $F_{0.5-2.5 ~{\rm keV}} \la4.1 \times 10^{-12}~{\rm erg}~ {\rm cm}^{-2} ~{\rm s}^{-1}$.

The cluster Abell 1024S has an estimated redshift of z=0.1053 (taken from a 2002 version of Andernach & Tago 1998 compilation of Abell clusters redshifts, H. Andernach 2002, priv. comm.) and we found two bright SUMSS radio sources within its Abell radius.

A preliminary analysis of all these radio sources yields a total flux evaluated at 1.4 GHZ of $S_{1.4} \approx 270$ mJy. A more refined analysis of the radio and X-ray sources in the field of this EGRET source will be presented in a future paper Colafrancesco & Andernach (2002, in preparation).

Based on the previous evidence, we consider now that this is a probable association between galaxy clusters and an EGRET gamma-ray source.

3.2 3EG J1825-7926

This EGRET source has been detected with a (TS)^1/2 = 4.9 in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV}) =(18.4~ \pm~ 4.5) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma=2.47 \pm 0.31$. Its flux does not change significantly over all the VPs with the exception of the VP = 38.0 in which a flux increase by a factor $\sim$2 has been recorded; however, this flux is consistent with the flux detected in the other periods at less than the $2 \sigma$ confidence level (see Fig. 1). The quite low upper limit of < $13.5 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ found in the VP = 402.+ has $({\it TS})^{1/2} = 0$ and thus is not statistically significant. The Abell cluster Abell 3631 falls at the border of the $95 \%$ confidence level position error contours of the EGRET source. The position error map of this EGRET source is quite broad with $\theta_{95} = 0.78$ $\deg$ and it is elongated in the south-west north-east direction. There is no other known gamma-ray source counterpart for this EGRET source. However, there is poor information available on the cluster Abell 3631. In particular, there is no NVSS radio source found in the field of this cluster and no X-ray information. There is one bright radio source in the field of 3EGJ1825-7926 found in the SUMSS survey, but it is $\sim$20 arcmin away from the optical center of Abell 3631.

Due to the previous evidence, there is no strong hint indicating the possible association of this cluster with the EGRET source 3EG J1825-7926 and so we consider this case as likely due to projection effects.

3.3 3EG J0348-5708

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.1$ in the VP = P2. It has a flux $F_{\rm P2}({>}100~{\rm MeV})=(22.1 \pm 7.6)\times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ but the power-law spectral index remains unconstrained. Its flux does not change over all the VPs in which it has been detected and it is a stationary gamma-ray source (see Fig. 1). The quite low upper limits < $10 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ found in other independent VPs have all (TS)^1/2 = 0 and can not be considered as statistically significant. The other reported upper limits on this source are consistent with the detection fluxes.

The Abell cluster Abell 3164 falls within the $95 \%$ confidence level position error contours of the EGRET source. Also, there is no other identified gamma-ray source counterpart for this EGRET source. The EGRET source map is relatively broad with $\theta_{95} = 0.42$ $\deg$ and has a comet-like tail in the east direction.

Abell 3164 is an irregular cluster with a redshift of z = 0.057 (Struble et al. 1999) for which there is poor information available. Ebeling et al. (1996) estimated its X-ray luminosity to be $L_{0.5-2.4~{\rm keV}} \approx 1.48 \times 10^{44}$ erg s^-1 and its temperature as $kT \approx 3.6$ keV from the ROSAT data. There are no evidence for NVSS radio sources found in correlation with this cluster, given the low cluster declination. A few SUMSS radio sources are found at more than $\sim$20 arcmin from the cluster center.

Based on the previous evidence, we do not find any strong hint for the probable association of this cluster with the EGRET source 3EG J0348-5708 and we consider also this case as likely due to projection effects.

3.4 3EG J0159-3603

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.3$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100 {\rm ~MeV})=(9.8 \pm 2.8)\times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma=2.89 \pm 0.51$. Its flux does not change over all the VPs and it is a stationary gamma-ray source (see Fig. 1). Two Abell clusters fall within the $95 \%$ confidence level position error contours of this source: Abell 219S and Abell 2963. Other galaxy clusters are found in the vicinities of the EGRET source (see Fig. 5). The EGRET source map is quite broad with $\theta_{95} = 0.79$ $\deg$ and the elongation of the EGRET probability map is aligned with the position of the two clusters, suggesting a possible contribution to the gamma-ray emission from both clusters. There is no other known gamma-ray source counterpart for this EGRET source.

There is little optical and X-ray information on both the clusters Abell 219S and Abell 2963. There are nonetheless three NVSS radio sources associated with the cluster Abell 2963: they have radio flux at 1.4 GHz of $S_{1.4}= (23.44\pm1.57), (4.46 \pm 0.45)$ and $(2.43\pm0.45)$ mJy, respectively.

In view of these evidence, we consider that this is a candidate for a probable association between galaxy clusters and an EGRET gamma-ray source.

3.5 3EG J0616-3310

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.7$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100 {\rm ~MeV}) =(12.6 \pm 3.2) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma=2.11 \pm 0.24$. Its flux changes significantly over several VPs and in the VP = 419.5 it increase by a factor $\ga$4 with respect to the VP = P1234. Due to such strong flux variations in comparison with other cases shown in Fig. 1, it is hard to consider it as a stationary gamma-ray source.

Two clusters (Abell 577S and Abell 575S) fall close to the $95 \%$ confidence level error contour for the position of this EGRET source. Another cluster (Abell 573S) falls within 1 $\deg$ radius from the center of the EGRET source. However, the shape of the EGRET map of this source is quite compact and round with $\theta_{95}=0.13$ $\deg$.

The two clusters Abell 577S and Abell 575S have very few morphological and physical information (see Table 1). Nonetheless, the cluster Abell 577S is associated with three NVSS radio sources with flux $S_{1.4} = (4.32 \pm 0.48),$ $(14.92 \pm 0.61)$ and $(12.48 \pm 1.57)$ mJy, respectively. Also Abell 575S is correlated with other three NVSS radio sources with flux $S_{1.4} = (5.86 \pm 0.48), (14.17 \pm 0.59)$ and $(3.03 \pm 0.47)$ mJy, respectively. Note that also the cluster Abell 573S is associated with 7 NVSS radio sources.

Due to the previous evidence, and in particular the flux variation over the various VPs, we consider this association as suspect and probably due to projection effects.

3.6 3EG J2034-3110

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.0$ in the VP = P1. It has a flux $F_{\rm P1}({>}100~{\rm MeV})=(17.4 \pm 5.2) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma=3.43 \pm 0.78$. Even though the flux variations over the different VPs and the upper limit of < $6.2 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with $({\it TS})^{1/2} = 0$ found in the VP = 209.0 are not statistically significant, the behaviour of this EGRET source is quite different from the other sources here selected as possible association with galaxy clusters, which are expected to be quite stationary over different VPs. Due to such flux variations in comparison with other cases shown in Fig. 1, we do not consider it as a stationary gamma-ray source.

Nonetheless, this EGRET source is quite broad and irregular with a quite large value of $\theta_{95}= 0.73$ $\deg$. The cluster Abell 886S falls within the $95 \%$ confidence level position error contour of the source. There is no other gamma-ray source counterpart in the field of 3EG J2034-3110.

The cluster Abell 886S has no detailed information available, it is not associated with any NVSS radio sources and there are no other hints for the presence of galaxy activity in its environment.

Due to the previous evidence, and in particular the flux variation over the various VPs, we consider also this association as suspect and probably due to projection effects.

3.7 3EG J1234-1318

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.8$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV})=(7.3 \pm 1.7) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a quite low power-law spectral index $\gamma=2.09 \pm 0.24$. The flux variations over the different VPs are not statistically significant, and for this reason it can be considered as a stationary gamma-ray source. The upper limit of < $8.9 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ has $({\it TS})^{1/2} = 0$ and its very poor statistical significance does not affect strongly the previous conclusion.

Two galaxy clusters (Abell 1558 and Abell 1555) fall within the $95 \%$ confidence level position error contour of the EGRET source. This EGRET source is quite regular with $\theta_{95}= 0.76$ $\deg$ even though source confusion may affect its flux and/or its position (see Hartman et al. 1999). No other known gamma-ray source counterpart has been found in the field of this EGRET source.

Abell 1558 has an estimated redshift of z=0.116 (Andernach 2002, priv. comm.) and it is associated with two NVSS radio sources with flux $S_{1.4} = 2.92 \pm 0.47$ and $7.73 \pm 1.24$ mJy, respectively. Also the cluster Abell 1555 at z =0.127 is associated with two NVSS radio sources with flux $S_{1.4} = 7.14 \pm 0.48$ and $4.24 \pm 0.48$ mJy, respectively. No other information is available on these clusters at both optical and X-ray frequencies.

Due to the previous evidence, we find that the association of this EGRET source with the two Abell clusters here mentioned is still questionable.

Figure 5: We show here the positions of the galaxy clusters which are probable candidates for the associations with unidentified EGRET sources: 3EGJ0215+1123 associated with Abell 331 (upper left), 3EGJ0439+1105 associated with Abell 497 (upper right), 3EGJ0159-3603 associated with Abell 2963 and Abell 219S (lower left) and 3EGJ1337+5029 associated with Abell 1758 (lower right). The cluster positions and the X-ray brightness contours (when available) are superposed to the maps of the EGRET sources. The intensity scale of the EGRET maps goes from black (minimum) to white (maximum).

3.8 3EG J0038-0949

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.1$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV}) =(12.0 \pm 3.7) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma=2.70 \pm 0.44$. The flux variations over the different VPs are at less than the $2 \sigma$ level and so are not strongly statistically significant. Also the low upper limit of < $11.8 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ obtained in the VP = 327.0 has (TS)^1/2 = 0 and it is not statistically significant. However, the behaviour of the flux changes in the different viewing periods over which this EGRET source has been detected is somewhat different from a purely stationary source as shown in Fig. 1. This EGRET source is elongated in the east-west direction and has a value $\theta_{95}= 0.59$ $\deg$.

The optical and X-ray center of the cluster Abell 85 is found slightly beyond the $95 \%$ confidence level position error contour of the EGRET source (see Fig. 4). However, due to its large extension ($\ga$30 arcmin radius) a large part of this nearby (z=0.056) cluster falls within the $95 \%$ confidence level position error contour of the EGRET source and hence can be considered to be spatially correlated with it. No other known gamma-ray source counterpart is found in the field of this EGRET source.

Abell 85 is a bright X-ray cluster with a luminosity $L_{2-10~{\rm keV}}=(7.65 \pm 0.52)\times 10^{44}~{\rm erg} ~{\rm s}^{-1}$ and a temperature of $kT = 6.2 \pm 0.4$ keV (Wu et al. 1999) and shows a strong activity in its ICM.

In fact, there are several bright radio galaxies within the cluster Abell 85 and also several bright NVSS radio sources correlated with the cluster as well as in the field of the relative EGRET source. Abell 85 has been observed with the VLA in the B and C configurations and a flux S_1.4 = 55 mJy has been reported by Owen & Ledlow (1997). Abell 85 contains also a diffuse, relic radio source (see, e.g., Slee et al. 2001) found off-center with respect to the X-ray center of the cluster (see Fig. 6). Giovannini & Feretti (2000) estimated that the diffuse radio halo flux at 1.4 GHz is S_1.4 = 46 mJy, consistently with the result of Owen & Ledlow (1997), with a power $P_{1.4} = 6.1 \times 10^{23}$ W Hz^-1.

There is also evidence of a hard X-ray emission excess which is spatially correlated with the radio relic source and is due probably to Inverse Compton Scattering (ICS) of the CMB photons with the relativistic electrons of the radio relic (Bagchi et al. 1998; Lima-Neto et al. 2001). Such a non-thermal X-ray emission is spatially correlated with the Very Steep Spectrum radio source MRC 0038-096 (see Bagchi et al. 1998), without any detected optical counterpart, which is $\sim$7 arcmin south-west of the X-ray center of the cluster.

The positive detection of both synchrotron radio and ICS X-ray emission from a common ensemble of relativistic electrons leads to an estimate of the average magnetic field, $B \approx 0.95 \pm 0.10~\mu$G, on the cluster scale. Further, the radiative flux and the estimated value of B imply the presence of relativistic electrons (with radiative lifetime $\ga$10⁹ yr) with Lorentz factor $\gamma_{\rm L} \approx 700 {-} 1700$ (Bagchi et al. 1998). Electrons with these energies can easily emit gamma-rays at $E_{\gamma} > 100~{\rm MeV}$ by bremsstrahlung in addition to the ICS emission tail which is present in the gamma-ray region probed by EGRET.

Even though the cluster Abell 85 is offset with respect to the center of the EGRET source map, there are good reasons to believe that it may contribute substantially to the gamma-ray flux of the EGRET source 3EG J0038-0949 in addition to the possible gamma-ray flux possibly produced by the active radio-galaxies which are living in the cluster environment.

Figure 6: One of the most probable associations between galaxy clusters and EGRET unidentified gamma-ray sources: Abell 85. This cluster has a radio halo/relic inhabiting the cluster and a number of identified radio galaxies. Shown are the EGRET image with the cluster X-ray brightness contours (right), the ROSAT-HRI X-ray image of the cluster (center) with the NVSS radio sources in the field (red circles) and the radio halo/relic image obtained with the VLA at 327 MHz (left).

3.9 3EG J1310-0517

This EGRET source has been detected with a $({\it TS})^{1/2} = 5.0$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV})=(7.9 \pm 1.8) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma=2.34 \pm 0.22$. The flux variations over the different VPs are at less than the $2 \sigma$ level and are not statistically significative (see Fig. 1). Also the lowest upper limit < $10.9 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$, obtained for this source in the VP = Virgo4 with $({\it TS})^{1/2} = 0.5$ has a very poor statistical significance. The EGRET source is elongated in the south-north direction and has a value $\theta_{95} = 0.78$ $\deg$.

The center of the cluster Abell 1688 is found within the $95 \%$ confidence level position error contour of the source, even though quite off-center with respect to the EGRET map center (see Fig. 4). There is, however, no other known gamma-ray source counterpart in the field of this EGRET source.

Abell 1688 is one of the most distant clusters listed in Table 1 and has little information available at both optical and X-ray wavelengths. Kowalski et al. (1984) gave an estimate of its redshift $z\sim 0.19$ and of its X-ray luminosity, $L_{2-10~{\rm keV}} \la4.79 \times 10^{44}$ erg s^-1 as obtained from the HEAO-A1 all-sky survey. There are, nonetheless, 4 NVSS radio sources correlated with the position of Abell 1688 and they have flux $S_{1.4} = 8.63 \pm 0.50$; $59.83 \pm 2.25$; $10.29 \pm 0.53$; $3.66 \pm 0.51$ mJy, respectively.

Due to these evidence, this could be considered as a probable - but still questionable - association between galaxy clusters and EGRET gamma-ray source.

3.10 3EG J0253-0345

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.0$ in the VP = 317.0. It has a flux $F_{317.0}({>}100~{\rm MeV}) =(17.3 \pm 5.7) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index which is unconstrained. There are no flux variations over the two VPs in which the source has been detected (see Fig. 1). The low upper limit < $4.2 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ derived in the VP = 21.0 has $({\it TS})^{1/2} = 0$ and is not statistically significant. The EGRET source map is quite extended and round with a high value of $\theta_{95}= 1.13$ $\deg$.

The cluster Abell 388 falls within the $95 \%$ confidence level position error contour of the source (see Fig. 4) and other Abell clusters are found in the field of this EGRET source. No other possible counterpart of the gamma-ray source is found in the field of this EGRET source.

Abell 388 has a redshift of z =0.134 and an estimated X-ray luminosity of $L_{2-10~{\rm keV}} \la3.47 \times 10^{44}$ erg s^-1, as reported in the HEAO-A1 all-sky survey (Kowalski et al. 1984). No other relevant X-ray information is available for this cluster. There are 8 NVSS radio sources correlated with this cluster with flux $S_{1.4} = 8.42 \pm 0.50, 4.69\pm0.47, 10.08\pm1.13, 4.18\pm0.48, 7.58\pm0.49, 15.32\pm1.13, 3.91\pm0.48$ and $4.03 \pm 0.48$ mJy, respectively.

Due to these evidence, this could be considered as a probable - but not yet definite - association between galaxy clusters and EGRET gamma-ray source.

3.11 3EG J0439+1105

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.2$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV})=(9.4 \pm 2.4) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 2.44 \pm 0.29$. There are no statistically significative flux variations over the different VPs in which the source has been detected (see Fig. 1) and the statistically significant upper limits are consistent with this conclusion. The EGRET source error area is quite extended with a high value of $\theta_{95}= 0.92$ $\deg$.

The cluster Abell 497 falls within the $95 \%$ confidence level position error contour of the source. No other gamma-ray source counterpart is found in the field of this EGRET source (see Fig. 5).

Abell 497 has an estimated redshift of $z \sim 0.14$ and an estimated X-ray luminosity of $L_{2-10~{\rm keV}} \sim 3.89 \times 10^{44}$ erg s^-1 (Ulmer et al. 1980). There is a NVSS radio source correlated with the cluster within its Abell radius with a flux $S_{1.4} = 3.27 \pm 0.47$ mJy.

Due to these evidence, this could be considered as a probable - but not yet definite - association between galaxy clusters and EGRET gamma-ray source.

3.12 3EG J0215+1123

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.4$ in the VP = 21.0. It has a flux $F_{21.0}({>}100~{\rm MeV})=(18.0 \pm 5.0) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a quite low power-law spectral index $\gamma = 2.03 \pm 0.62$. There are no other definite detection of this EGRET source over other viewing periods (see Fig. 1). The upper limit of $6.0 \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ obtained in the VP = 317.0 is not statistically significant since it has $({\it TS})^{1/2} = 0$. The EGRET source is quite extended with a high value of $\theta_{95}= 1.06$ $\deg$ and is elongated in the south-north direction.

The cluster Abell 331 falls near the center of the EGRET source (see Fig. 5) and there is no other gamma-ray source counterpart in the field of 3EG J0215+1123. There is another Abell cluster (Abell 330) which is found at $\sim$1.1 $\deg$ south of Abell 331. The cluster Abell 331 has an estimated redshift of z = 0.186 and an X-ray luminosity of $L_{2-10~{\rm keV}} \la5.01 \times 10^{44}$ erg s^-1 as estimated in the HEAO-A1 all-sky survey (Kowalski et al. 1984). There are four NVSS radio sources which are correlated with this cluster within its Abell radius and their flux is $S_{1.4}= 3.68 \pm 0.48$; $3.38 \pm 0.48$; $4.73 \pm 0.49$; $133.93 \pm 4.42$ mJy, respectively.

Due to these evidence, this could be considered as a probable - but still questionable - association between galaxy clusters and EGRET gamma-ray source.

3.13 3EG J2248+1745

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.1$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV}) =(12.9~ \pm~ 3.5) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 2.11 \pm 0.39$. The flux variations over the different VPs are at less than the $2 \sigma$ level, but the behaviour of this source is quite different from the stationary ones in Fig. 1. The EGRET source is quite extended with a value $\theta_{95}= 0.94$ $\deg$ and is elongated in the south-north direction.

The cluster Abell 2486 falls within the $95 \%$ confidence level position error contour of the source. No other gamma-ray source counterpart is found in the field of this EGRET source.

Abell 2486 has an estimated redshift of z = 0.143 and an X-ray luminosity of $L_{2-10~{\rm keV}} \la1.48 \times 10^{44}$ erg s^-1 as estimated in the HEAO-A1 all-sky survey (Kowalski et al. 1984). There are two NVSS radio sources which are correlated with this cluster within its Abell radius and their flux is $S_{1.4} = 4.60 \pm 0.51$ and $4.84 \pm 0.49$ mJy, respectively.

Due to these evidence, and in particular to the flux variations over the various VPs, this case should not be considered as a possible association between galaxy clusters and an EGRET gamma-ray source.

3.14 3EG J1212+2304

This EGRET source has been detected with a $({\it TS})^{1/2} = 3.3$ in the VP = Virgo2. It has a flux $F_{\rm Virgo2}({>}100~{\rm MeV}) =(19.7 \pm 7.7) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 2.76 \pm 0.60$. Even though the flux of this source may be affected by confusion, there are quite strong flux variations over the three different VPs in which this source has been detected (see Fig. 1). The EGRET source is quite extended with a value $\theta_{95}= 0.88$ $\deg$ and is elongated in the south-north direction.

The cluster Abell 1494 falls within the $95 \%$ confidence level position error contour of the source. No other gamma-ray source counterpart is found in the field of this EGRET source.

Abell 1494 has an estimated redshift of z=0.159 and an X-ray luminosity of $L_{2-10~{\rm keV}} \la1.95 \times 10^{45}$ erg s^-1 as estimated in the HEAO-A1 all-sky survey (Kowalski et al. 1984). There are five NVSS radio sources which are correlated with this cluster and their flux is $S_{1.4} = 4.91 \pm 0.48$; $8.05 \pm 0.49$; $23.68 \pm 1.74$; $17.53 \pm 0.66$; $5.15 \pm 0.47$ mJy, respectively.

Due to these evidence, and in particular to the flux variations over the various VPs, this case should not be considered as a possible association between galaxy clusters and EGRET gamma-ray source.

3.15 3EG J1347+2932

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.0$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV})=(9.6 \pm 2.9) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 2.51 \pm 0.61$. There are no strong flux variations over the different VPs in which this source has been detected (see Fig. 1), even though the flux of this source may be affected by confusion. However, we noticed that there is only one independent detection of this source in the VP = 4.0 which does not allow to draw any definite conclusion on its possible variability. The EGRET source is quite extended with a value of $\theta_{95}= 0.95$ $\deg$ and is irregular.

The cluster Abell 1781 falls within the $95 \%$ confidence level position error contour of the source. No other gamma-ray source counterpart is found in the field of this EGRET source.

Abell 1781 has a redshift of z = 0.0618 and an X-ray luminosity of $L_{2-10~{\rm keV}} \sim 1.15 \times 10^{44}$ erg s^-1 as estimated in the HEAO-A1 all-sky survey (Kowalski et al. 1984). There is one radio galaxy (FIRST J134159.7+294653) at a redshift of z = 0.0457 which is apparently in the foreground of the cluster. There is evidence for 36 NVSS radio sources correlated with Abell 1781 within its Abell radius.

Due to these evidence, and in particular to the uncertainties in the flux variations over the various VPs, this case should not be considered as a possible association between galaxy clusters and EGRET gamma-ray source.

3.16 3EG J1424+3734

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.4$ in the VP = P1. It has a flux $F_{\rm P1}({>}100~{\rm MeV})=(16.3 \pm 4.9)\times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 3.25 \pm 0.46$. There are no strong flux variations over the different VPs in which this source has been detected (see Fig. 1) and the lowest upper limit of < $16.1\times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ found in the VP = 201.+ is not statistically significant because it has $({\it TS})^{1/2} = 0$. The EGRET source is quite regular with a value $\theta_{95}= 0.88$ $\deg$ and with an emission tail in the north-west side of the field.

Two rich clusters (Abell 1914 and Abell 1902) fall within the $95 \%$ confidence level position error contour of the source (see Fig. 4). No other gamma-ray source counterpart is found in the field of this EGRET source.

Abell 1902 has a redshift of z = 0.16 and is associated with the X-ray source RXJ1421.6+3717 (Boehringer et al. 2000) with an X-ray flux $F_{0.1-2.4~{\rm keV}} = (4.3 \pm 10.8)\times 10^{-12}~{\rm erg}~{\rm cm}^{-2}~{\rm s}^{-1}$ and a luminosity $L_{0.1-2.4~{\rm keV}} \sim 6\times 10^{44}~{\rm erg}~{\rm s}^{-1}$. Its X-ray luminosity has been also estimated to be $L_{2-10~{\rm keV}} \sim 1.26 \times 10^{44}$ erg s^-1 in the HEAO-A1 all-sky survey (Kowalski et al. 1984). There is one radio source (FIRST J142140.4+371731) associated to the cluster galaxy MAPS-NGP 0-272-0323568 found at a redshift of z = 0.16. There is also a NVSS radio source with flux $S_{1.4} = 3.51 \pm 0.46$ mJy which is associated with the cluster.

Abell 1914 has a redshift of z = 0.1712 and is associated to the X-ray source RXJ1426.0+3749 with an X-ray flux $F_{0.1-2.4 ~{\rm keV}} = (12.90 \pm 5.2)\times 10^{-12}~{\rm erg} ~{\rm cm}^{-2} ~{\rm s}^{-1}$. Its X-ray luminosity has been estimated to be $L_{0.1-2.4~{\rm keV}} \sim 15.91 \times 10^{44}$ erg s^-1 (Boehringer et al. 2000). Three NVSS radio sources are correlated with Abell 1914 and they have flux $S_{1.4} = 9.91 \pm 1.50$, $30.86 \pm 1.77$, $20.75 \pm 4.08$ mJy, respectively. The cluster Abell 1914 also hosts the Very Steep Spectrum radio galaxy 1474+380 (4C 38.39) (Komissarov & Gubanov 1994). This cluster has also a bright radio halo (see Fig. 7) detected in the VLA with a flux of S_1.4 = 50 mJy and a power $P_{1.4} = 6.31\times 10^{24}$ W Hz^-1 (Giovannini & Feretti 2000).

Due to the previous evidence we consider this case as a probable candidate for the correlation of galaxy clusters and EGRET unidentified gamma-ray sources.

Figure 7: One of the most probable associations between galaxy clusters and EGRET unidentified gamma-ray sources: Abell 1914. This cluster has a radio halo/relic inhabiting the cluster and a number of identified radio galaxies in the ICM. Shown are the EGRET source image with the cluster X-ray brightness contours (right), the ROSAT-HRI X-ray image of the cluster (center) with the NVSS radio sources in the field (red circles) and the radio halo/relic image observed with the VLA at 1.4 GHz (left).

3.17 3EG J1337+5029

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.4$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV})=(9.2 \pm 2.6) \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 1.83 \pm 0.29$. There are no strong flux variations over the different VPs in which this source has been detected (see Fig. 1). The EGRET source is quite regular with a value $\theta_{95}= 0.72$ $\deg$.

The rich cluster Abell 1758 falls within the $95 \%$ confidence level position error contour of the source, very close to the center of the EGRET source map (see Fig. 5). No other gamma-ray source counterpart is found in the field of this EGRET source.

Abell 1758 is the most distant cluster listed in Table 1. It has a redshift of z = 0.279 and is associated with the X-ray source RXJ1332.7+5032 with an X-ray flux $F_{0.1-2.4~{\rm keV}} = (5.6 \pm 9.5) \times 10^{-12}~{\rm erg} ~{\rm cm}^{-2} ~{\rm s}^{-1}$ and a luminosity $L_{0.1-2.4~{\rm keV}} \sim 1.8 \times 10^{45} ~{\rm erg} ~{\rm s}^{-1}$ (Boehringer et al. 2000). The ROSAT-PSPC observation yielded a temperature $kT \approx 4.1$ keV which is found to be much lower than the ASCA (SIS+GIS) temperature of $T \approx 9.33$ keV (Rizza et al. 1995). An X-ray luminosity of $L_{2-10~{\rm keV}} \approx (1.43 \pm 0.06) \times 10^{45}$ erg s^-1 has been estimated independently by Wu et al. (1999).

There are four NVSS radio sources correlated with this cluster with a flux $S_{1.4} = 109.83~ \pm~ 3.86$, $8.69~ \pm~ 1.42$, $14.83~ \pm~ 1.51$, $5.18 \pm 0.43$ mJy, respectively. Abell 1758 also hosts the narrow tailed radio galaxy 87GB 133050.3+504752 (Feretti et al. 1992). This cluster also shows a diffuse radio emission (see Fig. 8) which could be possibly identified with an extended radio halo (Giovannini & Feretti 2000).

Due to the previous evidence we consider this case as a probable candidate for the correlation of galaxy clusters and EGRET unidentified gamma-ray sources.

Figure 8: One of the most probable associations between galaxy clusters and EGRET unidentified gamma-ray sources: Abell 1758. This cluster has a radio halo/relic inhabiting the cluster and a number of identified radio galaxies in the ICM. Shown are the EGRET image with the cluster X-ray brightness contours superposed (right), the ROSAT-HRI X-ray image of the cluster (center) with the NVSS radio sources in the field (red circles) and the radio halo/relic image observed with the VLA at 1.4 GHz (left).

3.18 3EG J1447-3936

This EGRET source has been detected with a $({\it TS})^{1/2} = 4.5$ in the VP = P1234. It has a flux $F_{\rm P1234}({>}100~{\rm MeV}) =(11.0~ \pm~ 2.7)~ \times 10^{-8}~{\rm cm}^{-2} ~{\rm s}^{-1}$ with a power-law spectral index $\gamma = 2.45 \pm 0.34$. There are no strong flux variations over the different VPs in which this source has been detected (see Fig. 1). However, the only two independent detections of this source do not allow to draw any definite conclusion on its variability. The EGRET source is quite regular with a value $\theta_{95}= 0.87$ $\deg$. The cluster Abell 774S falls within the $95 \%$ confidence level position error contour of the source. No other gamma-ray source is found in the field of this EGRET source.

Abell 774S is a poor cluster with very limited information at other wavelengths. There are 9 NVSS radio sources correlated with this cluster within its Abell radius. There are also 2 PMN sources within the Abell radius of Abell 774S with flux $S_{4.85} =126 \pm 11$ mJy and $67\pm10$ mJy, respectively. The latter one coincides with NVSS J144952.4-395732 with S_1.4=55 mJy, an inverted-spectrum radio source.

We thus believe that this case of spatial correlation is likely due to projection effects.