next previous
Up: Identification of selected sources I.


Subsections

3 Identification of sources from the ROSAT Galactic Plane Survey

3.1 The RGPS source list

The 93 sources discussed in this paper were extracted from the RGPS source list using selection criteria described in section 3.5 below. The RGPS source list results from the survey reduction performed by SASS-I in October 1991. This RASS reduction has well known documented flaws, namely a PSPC count rate overestimated by a factor of $\approx$ 1.2, the appearance of some ghost sources caused by problems in the satellite attitude reconstruction in spinning mode and uneven sensitivity due to the strip accumulation.

Spacecraft problems prevented the all-sky detection of sources between ecliptic longitudes 41$^\circ$ and 49$^\circ$ and ecliptic longitudes 221$^\circ$ and 229$^\circ$ leaving about 5% of the galactic sky without coverage.

The 1991 SASS-I version stacked photons in 2$^\circ$ wide strips along the great scan circles resulting in frequent multiple detections of the same source in adjacent strips (if overlapping sufficiently). The lists of sources derived from each strip were merged into a single master list totaling about 15000 sources at |b| $\leq$ 20$^\circ$. This list constitutes the database for the RGPS. Details on the merging procedure can be found in Motch et al. (1997b).

Errors on ROSAT X-ray positions are quadratic sums of the statistical uncertainty with which the centroid of the X-ray image is positioned on the pixel grid by the Maximum Likelihood source detection algorithm and of a systematic attitude error, estimated to be of the order of 8$^{\prime\prime}$ (Motch et al. 1997b).

Among the 93 sources, a total of 87 are bright enough to be also found in the RASS bright source catalogue (1 RXS, Voges et al. 1996) discussed in the previous section. The SASS process used for the production of the bright source list (SASS-II) detects sources on square sky areas, eliminating thus the uneven sensitivity resulting from the strip approach used in 1991. Updated attitude reconstruction removed ghost sources sometimes present in the 1991 version and human data screening ensured high quality. Most RGPS and BRASS positions are fully consistent with a median difference in position of 7.6$^{\prime\prime}$. The 90% confidence radii are similar. In a few cases mentioned below in the notes on individual objects, the offset between the RGPS and BRASS positions is significantly larger than r90 probably because of the use of an improved attitude solution in the SASS-II reduction. The 1991 SASS-I reduction also uses slightly different energy ranges than the SASS-II reduction for computing hardness ratios 1 and 2:

\begin{displaymath}
{ HR1}\ = \frac{(0.40-2.40)-(0.07-0.40)}{(0.07-2.40)} \ {\rm (SASS-I)} \end{displaymath}

\begin{displaymath}
{ HR2}\ = \frac{(1.00-2.40)-(0.40-1.00)}{(0.40-2.40)} \ {\rm (SASS-I)}\end{displaymath}

where (A-B) is the raw background corrected source count rate in the A-B energy range expressed in keV. There is no one to one relation between SASS-I and SASS-II hardness ratios as their values depend on the details of the observed count distribution in energy. However, a meaningful average graphical relation can be estimated using HR1 and HR2 plots for sources with accurate hardness ratio determinations. This relation was used to propagate the ranges from one reduction to the other.

3.2 Optical observations

Optical material was collected at the Observatoire de Haute-Provence, CNRS, France for northern sources and at ESO, La Silla for the southern sky. The present observational material was acquired during several runs dedicated to the identification of ROSAT galactic plane survey sources in general which were carried out from 1991 till 1995 by various observers.

At OHP, multicolour CCD imagery was usually obtained with the 1.2m telescope few days before the spectroscopic run with the 1.9m. In most cases we used B and I band filters with additional U band exposures in a few instances. Pixel size was 0.85 or 0.77 arcsec on the sky depending on the CCD chip used. At the 1.9m telescope, we operated the CARELEC spectrograph (Lemaitre et al. 1990). Most of the time we used two dispersions, a low resolution mode 260 Å/mm ($\lambda \lambda$ 3500 - 7500 Å; FWHM resolution $\approx$14 Å) and a medium resolution mode in the blue 33 Å/mm ($\lambda \lambda$ 3800 - 4300 Å; FWHM resolution $\approx$ 1.8 Å).

ESO data were acquired at the occasion of four runs in May 1991, April 1992, February 1994 and February 1997. In 1991, 1994 and 1997, we used the Boller & Chivens spectrograph at the ESO 1.5 m telescope. Medium dispersion gratings were used in all cases yielding a FWHM resolution of 4-5 Å and a wavelength range $\lambda \lambda 3900 - 7200$ Å. In 1992 we used EFOSC 2 at the ESO-MPI 2.2 m telescope with the same instrumental setting as described in Motch et al. (1994).

All spectral and photometric data reductions were performed using standard MIDAS procedures (Banse et al. 1983). Spectra were corrected for bias and flat-field and later calibrated in wavelength using arc lamps. In most cases we could acquire flux standard stars. However, uncertain weather conditions and the narrow slit entrance width sometimes used may introduce large errors in the derived flux. These errors may be of the order of a factor 2 or more.

3.3 Optical data analysis

For active coronae, spectral classification was carried out as outlined in Motch et al. (1997a) using Turnsheck et al. (1985), Jacoby et al. (1984) and Jaschek & Jaschek (1987) stellar atlases.

Visual magnitudes were in most cases not derived from our CCD imagery as they usually lacked photometric calibration. Instead we used magnitudes extracted from the SIMBAD database, the Guide Star Catalogue (Lasker et al. 1990) or for the faintest counterparts, from the USNO-A1 catalogue (Monet 1997). The GSC magnitudes of stars having a spectral type were corrected for colour effects according to relation (1) of Russell et al. (1990) assuming a main sequence unreddened object. After colour correction, the remaining 1$\sigma$ error on magnitudes is $\approx$0.2 mag.

The coordinates of the optical counterparts were in first priority extracted from the SIMBAD database which usually gives entries from astrometric catalogues (e.g. PPM). When no accurate SIMBAD positions were available we used the GSC coordinates and for the remaining identifications positions computed interactively using the Aladin sky atlas (Bonnarel et al. 1997) at the Centre de Données de Strasbourg (CDS).

The Aladin project aims to provide multi wavelength cross-identification. This tool is designed as an interactive X-window client accessing images from the CDS image server, Simbad database, CDS catalogue server and on-site catalogues. The Aladin collection contains a high resolution image archive of Schmidt plates digitized by the Paris MAMA facility and covering a significant portion of the sky, mainly in the Magellanic Clouds and southern Galactic Plane. The integration of the STScI Digital Sky Survey -1 in the system provides full-sky coverage albeit with a lower spatial resolution and astrometric accuracy than that of the MAMA archive. For the plates digitized by the MAMA, astrometric calibration is based on PPM standards and reaches an accuracy better than 0.3$^{\prime\prime}$ rms.

3.4 Optical identifications

The strategy used for optical identification of ROSAT sources in the galactic plane has been extensively discussed in Motch et al. (1997a). For stars, we used two criteria based on the Ca II H&K or H$\alpha$ flux to X-ray flux ratio and a priori probability of positional coincidence in the relatively small X-ray r90 (<r90> = 25$^{\prime\prime}$). A star was identified as the counterpart of the X-ray source when its line emission level was compatible with the X-ray flux, or when no spectral line could be measured, on the basis of positional coincidence. The surface density of optically bright active galactic nuclei, cataclysmic variables, hot white dwarfs and Be stars is small enough that the discovery of one such object in the ROSAT error circle is highly significant.

For all identified sources but one (RXJ0254.6+3931) we show an optical spectrum of the counterpart, either low or medium resolution but in general, do not provide finding charts as the identifier and positions are in principle sufficient to localize the object. However in few cases where the counterpart does not appear in the USNO catalogue we show finding charts. We also show finding charts for all accreting sources.

For the 17 cases where we failed to find the counterpart we show on a finding chart the observed candidates in order to ease further follow-up studies but do not plot the spectra.

By default, the finding charts are extracted from the STScI DSS-1 and the RGPS/SASS-I 90% confidence error circle is shown. In some instances where the DSS-1 data are not able to show the candidates because of crowding or extreme colours, we use instead our CCD images.

Comments on individual sources, spectra and finding charts are given in Sect. 5.

3.5 Source selection

The ROSAT sources discussed in this paper were extracted from the entire RGPS source list covering the whole galactic plane (apart from the small interruptions mentioned above). We mainly used four different selection criteria, hard, soft, absorbed soft and bright candidates, all based on SASS-I hardness ratios and count rates. Hardness ratio boundaries were chosen following the observed repartition of identified sources in the preliminary cross-correlation made in 1991 (Motch 1991a). These installments of X-ray selected sources were then distributed among several optical groups for identification at the telescope. In this paper, we present the optical work done by one of these observing groups. Some optical identifications which have been already published in dedicated papers are not repeated here whereas work at the telescope is pursued for a number of other sources.

To these purely X-ray selected samples, we add a couple of AGN extracted from not yet fully published lists of identifications in sample areas (e.g., GPS1-4; Motch et al. 1991b, Taurus region; Guillout et al. 1996a). Three other sources are from a so far barely investigated region located at l $\sim$130$^\circ$. In the following, we consider all sources together independently of their original selection. A small fraction of these sources was re-observed by the HRI in order to get more accurate positions. These cases are mentioned in Sect. 5.

Hard sources were defined as having $HR1\geq$0.7 and $HR2\geq$-0.1 ($HR1_{\rm BRASS}$$\geq$0.65 and $HR2_{\rm BRASS}$$\geq$-0.15). We furthermore imposed the additional condition that integrated galactic absorption had to be larger than 4 1021 H atom cm-2 and rejected extended sources. The requirement of large galactic absorption was aimed at screening the extragalactic component and searching preferentially for galactic accreting binaries. Applied to the entire RGPS database, this selection yielded 157 entries with count rates larger than 0.1 cts s-1 among which 84 had no identification in SIMBAD. We report here on 26 of these unidentified sources. We added to this sample an additional set of 13 sources sharing constraints on HRs but not on $N_{\rm H}$ and flux. These sources arise from early selections or from sample areas.

Soft sources were defined as having $HR1\leq$-0.4 ($HR1_{\rm BRASS}$$\leq$ -0.75) with a maximal error of 0.5 and a PSPC count rate $\geq$0.1cts s-1. Such soft sources do not emit much at energies higher than 0.4 keV and HR2 is therefore essentially undefined. A total of 112 sources fulfilled this criterion among which 43 had no identification in SIMBAD. We expected to preferentially find white dwarfs and in general low luminosity soft sources in this sample. Several of these sources were also detected by the Wide Field Camera and subsequently identified.

Absorbed soft sources were defined as having $HR1\geq$-0.4 and $HR2\leq$-0.4 ($HR1_{\rm BRASS}$$\geq$-0.6 and $HR2_{\rm BRASS}$$\leq$-0.2). This sample was designed to discover intrinsically soft luminous sources undergoing relatively large interstellar absorption as a result of their remote location. Above 0.1 cts s-1, 667 RGPS sources fulfil these constraints among which 199 had no identification in SIMBAD.

Sources which do not fall in any of the other hardness ratio ranges are split into the "Bright" group (PSPC count rates larger than 0.25cts s-1) and the "Other" group.

All samples were cleaned by discarding those sources having an obvious identification in SIMBAD (catalogued X-ray source, bright active corona). However, in the course of this identification programme, some of these sources were recovered by other instrumentations (Sky Lab SLX, WFC RE, EUVE, etc.) or identified by other groups and now appear as such in SIMBAD.

We show in Fig. 6 the position of the various X-ray selected samples in the HR1/HR2 diagram and give number repartition by selection criteria in Table 2.


 
Table 3: Repartition of the number of sources by selection criteria

  
\begin{figure}
\hspace{-3mm}
\psfig {figure=selection.ps,width=8.8cm,bbllx=0.0cm,bburx=22cm,bblly=0.0cm,bbury=22cm}\end{figure} Figure 6: Position of the various X-ray selected samples in the HR1/HR2 diagram. X = hard X-ray binary like, A = absorbed supersoft candidates, S = soft white dwarf candidates, B = X-ray bright sources, O = other remaining sources

In Table 4 we list the RGPS source positions, count rates, hardness ratios and 1 RXS identification when available. Table 5 gives optical identification, position, V and B when available. Spectra are shown in Fig. 8 and finding charts in Figs. 9 and 10.

3.5.1 Hard sources

Among the already published identifications of hard sources are the four CVs RXJ0028.8+5917, RXJ0744.9-5257, RXJ1141.3-6410 and RXJ2123.7+ 4217 (Motch et al. 1996). Some of the Be/X-ray candidates reported in Motch et al. (1997b) or the low-mass X-ray binary GS1826-24 identified in Barret et al. (1995) were also found in this sample. Here we report on the identification of RX J1739.5-2942 with a new Be/X-ray binary (see below). Our hard sample also contains the ultrasoft transient SLX 1746-331 (Skinner et al. 1990) which was apparently in outburst at the time of the ROSAT survey observation in September 1990 and not detected in follow-up ROSAT HRI observations on 1994 October 2.

Because of observational constraints, SIMBAD contains hardly any AGN seen through large galactic foreground absorption whereas all luminous galactic X-ray binaries, mostly discovered at higher energies are listed. This strong bias indicates that in hard X-ray selected samples, the number ratio of AGN to X-ray binaries or CVs should be much larger than suggested by the BRASS-SIMBAD correlation (see Figs. 2, 4 and 5). A total of 14 AGN are indeed found in the hard sample (see Table 3), confirming that in the galactic plane, there is no easy way to disentangle absorbed AGN from genuine accreting binaries since both populations exhibit hard spectra and faint bluish counterparts. Not surprisingly, the new identified hard AGN are much more absorbed than the SIMBAD sample and their HR2 is correlated with galactic $N_{\rm H}$ (see Fig. 7). In 10 instances we identify the X-ray source with an optically bright active corona and in 13 cases, we fail to find a likely counterpart. Not unexpectedly, the vast majority of unidentified sources (13 among 17) are found in the hard sample and based on the properties of the identified population we can predict that most of these sources are likely to be absorbed extragalactic objects. In particular, from the lack of bright objects in I band images we can exclude relatively close Be/X-ray binaries as possible identification for these hard sources.

  
\begin{figure}
\psfig {figure=agn_id_h2.ps,width=8.8cm,bbllx=0.0cm,bburx=22cm,bblly=0.0cm,bbury=22cm}\end{figure} Figure 7: Relation between hardness ratio HR2 and galactic absorption for the newly identified AGN


  
Table 3: Magnitude, redshift and selection of newly identified AGN

\begin{tabular}
{cclcc}
Source Name & $V/B$\space & Type & $z$\space & Sel\\ \hl...
 ...1 & 0.080 & X\\  RX J2044.0+2833 & 14.70 &Sy 1 & 0.050 & B\\ \hline\end{tabular}

3.5.2 Soft sources

The majority of the very soft sources discovered in the RASS turned out to be white dwarfs often also detected as bright UV sources in the Wide Field Camera Survey (Pounds et al. 1993) or by the Extreme Ultraviolet Explorer (Malina et al. 1994). Many of these bright sources were readily optically identified from their UV positions (Mason et al. 1995). Again, some active coronae emit low temperature X-ray spectra which are indistinguishable from white dwarf emission at the PSPC survey sensitivity.

Among the already published discoveries in this soft sample are the extremely hot PG 1159 star RX J2117.1+3412 (Motch et al. 1993) and the hot WD RX J2052.7+4639 (Motch et al. 1997a).

The other class appearing in this X-ray selected group is that of soft polars and intermediate polars in which the blackbody like X-ray emission from the polar cap dominates over the hard bremsstrahlung radiation from the shock in the accretion column (Beuermann & Schwope 1994; Haberl & Motch 1995). The polar RX J1802.1+1804 (Greiner et al. 1995) and RE 0751+14 (Mason et al. 1992; Motch & Haberl 1995) were also found in this sample. The exotic population of isolated neutron stars accreting from the interstellar medium should also show up as soft sources. In fact, two of the best candidates known so far were found in the soft sample (RXJ1856-37 Walter et al. 1996, RX J0720-3125 Haberl et al. 1997). RX J2102.0+3359 (see Tables 4 and 5) was also considered for some time as a good lone neutron star candidate. However, UB CCD imagery revealed the presence of a faint UV excess object (star X, B= 21.6, U-B = -1.1) close to the center of the error circle. This picture hints towards identification with an extreme $F_{\rm X}$/$F_{\rm opt}$ ratio AM Her CV such as RXJ0453.4-4213 = RS Cae (Burwitz et al. 1996). However, its firm identification awaits optical spectroscopic confirmation.

These identifications are in agreement with the results of the BRASS-SIMBAD analysis discussed above.

In the soft sample presented here, we report the identification of a new DA white dwarf (RX J1936.3+2632 = 2EUVE J1936+26.5), 2 new polars (RX J0649.8-0737 and RX J0749.1-0549) of rather long orbital period and of a couple of active coronae.

3.5.3 Absorbed soft sources

Based on the obvious observation that no extremely X-ray bright soft source was found in the ROSAT survey nor in previous EUV surveys, we concluded that if a luminous supersoft source analogue to Cal 83 in the LMC were existing in the Galaxy, its observed X-ray spectrum would be dimmed and distorted by interstellar absorption. In order to delimit the range of hardness values to select, we simulated soft black body spectra affected by large interstellar absorption. With the ROSAT PSPC, large photoelectric absorption of a very soft source produces a very peaked distribution centered at an energy which depends on the balance between kT and $N_{\rm H}$. These spectra are characterized by a hard HR1 and a soft HR2.

Unfortunately, at the PSPC resolution, optically thin thermal spectra with many emission lines in the range of 0.3 - 0.5 keV can also produce peaked spectra, albeit with a much lower photoelectric absorption. Some active coronae may emit this kind of spectra in a narrow range of temperature and photoelectric absorption. Such spectra are also often seen in supernova remnants and in the hot diffuse galactic emission in general. Care was therefore taken to exclude extended X-ray sources by examining visually the survey images of all absorbed soft detections.

The first sample of absorbed soft sources in which we discovered the galactic supersoft binary RX J0925.7-4758 (Motch et al. 1994) and the peculiar intermediate polar candidate RX J1914.4+2456 (Motch et al. 1996) had more stringent constraints than set here ($HR1_{\rm SASS-I}$ $\geq$ -0.4 and $HR2_{\rm SASS-I}$ $\leq$ -0.6 with errors on HR1 and HR2 of less than 0.5 and 1.0 respectively and a count rate larger than 0.1cts s-1). This restricted sample is now mostly identified and the remaining sources are associated with active stars.

Relaxing the constraints on HR2 increased the number of candidates but also the fraction of active coronae. Therefore, apart from one AGN (RX J1929.8+4622) and two CVs (RX J1951.7+3716 and RX J1946.2-0444), most sources are identified with stars. In three cases, however, we fail to identify the source. A fraction of these unidentified sources could be AGN with absorbed steep spectra.

3.5.4 Bright sources

A number of unidentified bright RGPS sources of high $F_{\rm X}$/$F_{\rm opt}$ turned out to be previously unknown AGN or CVs. As shown in Fig. 3, selecting high $F_{\rm X}$/$F_{\rm opt}$ sources at intermediate hardness ratios does improve the fraction of extragalactic and cataclysmic variable identifications, however, with slightly less efficiency than other HR based selections. Some very active coronae with magnitudes below that of the limit of HD or SAO catalogues were also discovered.

  \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_2543.ps,width=8.5cm,heig...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_2677.ps,width=8.5cm,heig...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure}
Figure 8: Optical spectra of the proposed identifications sorted by increasing right ascension


\begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_6762.ps,width=8.5cm,heig...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_9904.ps,width=8.5cm,heig...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure}
Figure 8: continued


\begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_13605.ps,width=8.5cm,hei...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_14232.ps,width=8.5cm,hei...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure}
Figure 8: continued


\begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_7391.ps,width=8.5cm,heig...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_12382.ps,width=8.5cm,hei...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure}
Figure 8: continued


\begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_13218.ps,width=8.5cm,hei...
 ...cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=spe_14853.ps,width=8.5cm,hei...
 ....5cm,bburx=18.5cm,bblly=16cm,bbury=27.5cm,clip=true}

&\end{tabular}\end{figure}
Figure 8: continued

  \begin{figure}

\begin{tabular}
{ll}

\psfig {figure=fc_1732.ps,height=8.0cm
,bb...
 ...x=1cm,bburx=17.5cm,bblly=9cm,bbury=27cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=fc_10706.ps,height=8.0cm
,bb...
 ...x=1cm,bburx=17.5cm,bblly=9cm,bbury=27cm,clip=true}

\\ \end{tabular}\end{figure}
Figure 9: Finding charts for unidentified sources. North is to the top and East is to the left. Image width varies with origin (STScI DSS-1 or proper CCD observations). For STCcI images, recognizable by their large pixel size (1.7$^{\prime\prime}$), the field of view is 2.89$\times$2.89 arcmin. The size of finding charts extracted from CCD images slightly changes with origin. We show the ROSAT 90% confidence radius error circle and the position of the optical candidates investigated spectroscopically. Section 5 contains information relevant to each of the finding charts


\begin{figure}
\begin{tabular}
{ll}

\psfig {figure=fc_6861.ps,height=8.0cm
,bbl...
 ...llx=1cm,bburx=17.5cm,bblly=9cm,bbury=27cm,clip=true}

&\end{tabular}\end{figure}
Figure 9: continued

  \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=fc_9571.ps,height=8.0cm
,bbl...
 ...x=1cm,bburx=17.5cm,bblly=9cm,bbury=27cm,clip=true}

\\ \end{tabular}\end{figure} \begin{figure}
\begin{tabular}
{ll}

\psfig {figure=fc_11062.ps,height=8.0cm
,bb...
 ...llx=1cm,bburx=17.5cm,bblly=9cm,bbury=27cm,clip=true}

&\end{tabular}\end{figure}
Figure 10: Finding charts for identified sources. Same as for previous figure. The arrow points at the proposed optical counterpart


 
Table 4:   X-ray characteristics of ROSAT survey sources derived from the SASS as in October 1991. Coordinates are equinox 2000.0. Column (2) indicates the source selection origin as shown in Fig. 6, S, soft, A, absorbed soft, X, X-ray binary like, i.e. hard. Remaining sources are split in B, bright (count rates $\geq$ 0.25 cts s-1) and O, others. Whenever the SASS-I RGPS source is bright enough to appear in the BRASS catalogue we give the 1RXS name


 
Table 4: continued


 
Table 5: Optical identifications of ROSAT survey sources. Coordinates are equinox 2000.0. Column (2) indicates the source selection origin as shown in Fig. 6, and given in Table 4. Column (3) gives X-ray to optical distance in units of the 90% confidence radius. In column Class, "AC" stands for active corona and "Gal" for normal galaxy. The last column indicates whether a finding chart is provided


 
Table 5: continued


next previous
Up: Identification of selected sources I.

Copyright The European Southern Observatory (ESO)