Astron. Astrophys. 342, L9-L12 (1999)

3. Results and Discussion

Fig. 1 shows the V frame of the And VI dwarf galaxy. It is clearly resolved into individual, faint stars. Similarly, the B and I band frames show faint, resolved stars. The images resemble those of the other M31 dSph companions (see e.g. And V in Armandroff et al., 1998). The stars are very regularly distributed and show no structure or knots which might indicate recent star formation as in dwarf irregular galaxies. From the smoothed frame, we measured a central surface brightness in V of mag/ (24.13 when corrected for galactic reddening using the value of Burstein & Heiles (1978) and the extinction law of Savage & Mathis (1979)). The light distribution follows nicely an exponential law of scale length arc sec and has an overall ellipticity of (Fig. 2). From this smoothed image, we constructed a mask which allowed to measure the total magnitude inside the 25.5 mag/ V isophote (which essentially corresponds to the Holmberg radius). We found V and (B-V)₀ = +0.65 0.1 (already corrected for a galactic reddening). These data which are summarized in Table 1 indicate a classification as dSph (compare e.g. to the tables in Mateo (1998) for the other dSph in the local group).

Fig. 1. Calar Alto 2.2m telescope CCD image in V of the Pegasus dwarf galaxy = And VI. North top, East left, field of view shown is 399.5 arc sec each side.

Fig. 2. Results of the surface photometry on the V frame. Top : Surface brightness versus major axis a, the line shows the fit. Middle : Ellipticity of the isophotes versus a. Bottom : Position angle versus a.

Fig. 3 shows the V, B-V color magnitude diagram (CMD) and Fig. 4 the V, V-I CMD. They are already corrected for a foreground reddening of 0.15 mag in B which was derived from Burstein & Heiles (1978).

Fig. 3. V, B-V color magnitude diagram. Left: Stars in the field surrounding And VI, illustrating the MWG contamination. Right: Stars either belonging to the And VI dwarf galaxy or MWG stars being projected on it. The left panel contains stars from the rest of the CCD frame, an area 3.5 times larger than the area of the dwarf galaxy. Both CMDs are corrected for galactic foreground reddening of 0.15 mag. in B. The right panel also shows as solid lines the red giant branches for M15 from Durrel & Harris (1993), a very low metal abundance globular cluster, and for 47 Tuc from Hesser et al. (1987), a high metal abundance globular. The RGB's of the two MWG globulars have been shifted to a distance modulus of 24.5.

Fig. 4. Same as Fig. 3 for the I, V-I color magnitude diagram of the And VI dwarf galaxy. The right panel also shows the mean RGB lines of five MWG globular clusters from da Costa et al. (1990) ranging in metallicity from very low values (M15) to high ones (47 Tuc) as indicated. The RGB's of the MWG globulars have been shifted to a true distance modulus of 24.5.

The right panels of Fig. 3 and 4 show only those stars with distances from the center of the And VI dwarf galaxy smaller than 175 arc sec. The left panels represent the stars from the rest of the CCD frames, a field about 3.5 times larger. A careful comparison indicates that essentially all stars brighter than I20.5 (Fig. 4) and V21.5 (Fig. 3) are MWG stars. Also, any faint and blue stars with V-I0.6 (Fig. 4) either belong to the MWG, or they are unresolved background galaxies.

Stars which are fainter than I = 20.5 and redder than V-I = 0.6 (V = 22.0, B-V = 0.3) are much more abundant within the area encompassed by the And VI dwarf galaxy than they are in the field. In particular, a strongly populated clump of stars is visible around V-I1.0 in the right panel of Fig. 4. An analogous feature is present in the right panel of Fig. 3. These are the brightest stars of the And VI dwarf galaxy.

These brightest stars could be - according to their colors - either red supergiants and intermediate aged AGB stars or old RGB stars. As there is no evidence for recent star formation, neither in our images (no knotty distribution of the stars, unresolved clumps) nor in the CMDs (no blue and yellow supergiants which are usually brighter and easier to detect, no blue blume), most of these stars are probably RGB stars of an old population. Indeed, a comparison with the RGB tracks of galactic globular clusters strongly supports this identification (Figs. 3 and 4). This is further strengthen by the fact that Armandroff et al. (1999) did not detect H. This comparison also indicates that the RGB of the And VI dwarf galaxy seems to be a mixed bag of metallicities, with most of the stars occupying the region of the low metallicity tracks. Our CMDs also resemble those found for other dSph companions of M31 (see Armandroff et al., 1998, and references therein).

A clear discontinuity in population density is visible in the V,V-I CMD at I = 20.5. We did I-band star counts inside and outside the area of the And VI dwarf galaxy in a color range of 0.6 V-I 2.5. The field number counts were used to statistically correct the counts inside the dwarf galaxy for the foreground contamination. The resulting luminosity function is shown in Fig. 5. The I-band number counts show a strong step at I = 20.50.2. We identify this feature with the TRGB. As we are dealing with a low-metallicity system (see above), we can apply the calibration of Lee et al. (1993), thus, . This yields a true distance modules of 24.50.25. The error gives the measurement error as derived from the location of the TRBG above, and includes the systematic error which was estimated following Schulte-Ladbeck et al. (1998). The leading error in our case is the accuraccy with which we can derive the TRGB magnitude in the I counts.

Fig. 5. Foreground-corrected I-band number counts of And VI RGB and AGB stars (color range 0.6 V-I 2.5). The horizontal bar indicates the value and errors of the TRGB.

The I-band number counts shown in Fig. 5 are still to be corrected for incompleteness which normally set in about 2 magnitudes above the observational limits. Thus, while the counts shown should not be used to construct luminosity functions, the TRGB brightness is still well above the regime of severe incompleteness.

© European Southern Observatory (ESO) 1999

Online publication: December 22, 1998
helpdesk.link@springer.de