2. Observations and results
The observations were all performed with the 1.52 m "Cassini" telescope equipped with the Bologna Faint Objects Spectrometer and Camera BFOSC (Bregoli et al. 1987, Merighi et al. 1994). During the first run (May 1998) the camera was equipped with the Thomson CCD with 0.56" pixel size and field of view (FOV). On July and September 1998 a Loral CCD detector with 0.41" pixel size and a FOV of was used instead. We performed V, R and I photometry, and low-resolution spectroscopy. The data were reduced using standard ESO-MIDAS and IRAF procedures for bias subtraction, flat-field correction, and one dimension stellar and sky spectra extraction. Cosmic rays were removed from each image and spectrum and the sky-subtracted stellar spectra were obtained, corrected for atmospheric extinction and flux calibrated (when possible).
2.1. Photometry and spectroscopy
V, R, and I images of the whole wide 1WGA J1958.2+3232 field were first obtained on 1998 May 30. In Fig. 1 the ROSAT PSPC error circle is shown together with the ASCA error circle (Israel et al. 1998, 1999). Note that at the time of the first optical observation only the ROSAT PSPC position uncertainty was known. Photometry for each stellar object in the image was derived with the DAOPHOT II package (Stetson 1987), and Color-Magnitude Diagrams (CMDs) were then computed for all color combinations. By the analysis of the V-R vs. V CMD a well defined main sequence is clearly visible (Fig. 2), as expected in an area projected along the galactic plane like the one of 1WGA J1958.2+3232 (; ). However, a number of objects with peculiar colors (redder or bluer) with respect the bulk of the stars were selected; three of them (stars A, B and C) were found to lie within the ROSAT position uncertainty region (see Table 1). Comparison between photometry obtained at the beginning and the end of the run showed no signs of variability to a limit of mag for any of the selected objects. A comparison between May and September 1999 gave similar results.
Table 1. Position and V magnitude of the studied stars within the uncertainty region.
On 1998 May 30 we obtained a low-resolution (20 Å) slitless spectroscopic image of the field with an R filter and a large band grism covering the spectral region (4000-8000 Å). We note that, in general, the choice of the grism, filter and time exposure depends on the grism dispersion and CCD size (which set the maximum spectral range allowed), and the crowding level of the field (which sets the maximum number of non-overlapping spectra to be analysed). The selected filter and grism combination gives a bandpass of 1000 Å, centered on H. A detailed analysis of the spectra obtained in this way allowed us to single out a relatively strong H emission line associated to one (star B) of the three stars previously selected.
Slit spectroscopy was performed over selected stars on 1998 May 30-June 2, July 27-31, and September 14-15 (see Table 2).
Table 2. Journal of the slit spectroscopic observations.
The spectra of star A are undoubtfully those of a classical OB star. However neither emission-lines or other peculiarities are present that would associate this star with the X-rays source. Similar results were obtained for stars D and E for which a slitless spectrum was obtained on May 1998.
Due to its faintness (R=17.0) star C is more difficult to study. Even after a 1.5 h exposure spectrum, the S/N ratio of the spectrum barely reached the unity with our instrumental set-up. The steep rise of its spectrum in the UV argues for a very hot object. However, it is unlikely that it is to be associated with the X-ray source since no obvious emission features are present. Moreover, after the ASCA observation, its position lies outside the intersection region of the two X-ray error circles.
We note that at the border of the ROSAT error circle lies a bright (V12) star. Its 4000-8000 Å spectrum (=5.5 Å) was taken in an earlier observation (June 1996) also from the Loiano Observatory using the BFOSC ; the star resulted to be a strongly reddened B8V spectral-type star, without any spectral peculiarity.
The spectrum of star B clearly shows a very high ionization state (see Fig. 3). The Balmer series lines are all in emission, up to the blue edge of our spectrogram as well as many He I ( 4471, 4922, 5875, 6678, 7065, 8361 Å) and He II ( 4686, 5412 Å). An emission feature centered at 4634 Å is present, that could be attributed to the N III 4634-40 Å doublet, but we cannot exclude the possibility that these feature is mainly due to Fe II lines. The presence of He II emission rules out a classical Be star, whereas the presence of emissions up to H7 exclude that of an Of star (see e.g. Jaschek & Jaschek, 1987).
Table 3 gives the parameters of the strongest emission lines. These lines (H, H, H, He I 5875 Å, He II 4686 Å) show extended and asymmetric profiles, with a red wing more pronounced than the blue wing. At our spectral resolution, this might indicate a line splitting, possibly due to the presence of a disk. The absence of forbidden lines rules out the possibility of a pre-main sequence object or a cataclysmic variable. All this evidence clearly point to the association of this object with the X-ray source.
Table 3. List of the strongest spectral lines identified in the likely optical counterpart of the 1WGA J1958.2+3232 star.
© European Southern Observatory (ESO) 1999
Online publication: April 12, 1999