5 Discussion

The above analysis shows that AS314 is a distant and a high-luminosity star. Its emission in the H$\alpha$ line is only comparable to those of the most luminous galactic stars of similar spectral types, HD 160529 (Stahl et al. [1995]) and HD 168607 (Chentsov & Luud [1989]), which are considered LBV candidates. AS314 is certainly not a Be star, as was suggested by Venn et al. ([1998]), because the line emission in Be stars arises in a quasi-Keplerian disk (e.g., Hanuschik et al. [1995]), which gives double- or single-peaked, but not PCyg-type, profiles. The narrow wings and cores of the H$\gamma$ and H$\delta$ lines in AS314 points to a low gravity and a low rotational velocity of the star which is common in A-type supergiants (Verdugo et al. [1999]), while Be stars have gravities at least 2 order of magnitude larger and are rapid rotators.

Furthermore, the IR-excess of AS314 is certainly due to thermal emission of circumstellar dust. Our analysis of the high-resolution IRAS maps of its environments constructed by means of the Maximum Correlation Method (Aumann et al. [1990]) and obtained from the Infrared Processing and Analysis Center in Pasadena (California, U.S.A.) shows that in all four IRAS bands the source is point-like and its position coincides with the optical position of AS314 (see Fig. 8). The star's $T_{\rm eff}$ is small enough to effectively heat interstellar dust. Therefore, the dusty particles responsible for the IR excess in AS314 are certainly connected to the star. One possible explanation of their existence is an LBV-type outburst occured in the past. Otherwise the dust may be a remnant of the red supergiant evolutionary phase. In both cases AS314 is a more evolved object than HD160529 and HD168607 quoted above, as the latter do not display any noticeable far-IR excess.

If we assume a lower limit of the dusty grains velocity of 5 kms^-1, then the kinematic age of the object's dusty envelope is less than $3\ 10^4$years. The evolutionary time passed since the end of the main-sequence phase is an order of magnitude larger for a $20\;M_{\hbox{$\odot$ }}$ star (Schaller et al. [1992]). Therefore, the envelope may be formed on the object's way from the main-sequence. On the other hand, Stothers & Chin ([1994]) noted that low-luminosity LBVs (log $L_{\rm bol}/L_{\hbox{$\odot$ }} \sim 5$) might have reached their present positions in the HRD after a violent mass loss during the red supergiant phase. In this case their initial masses are much larger than it is follows from the current theoretical evolutionary tracks. However, since a far-IR excess similar to that of AS314 is not observed in HD160529 and HD168607, which have almost the same fundamental parameters of the underlying stars and stellar winds, the red supergiant scenario seems not to be justified enough.

Figure 7: Theoretical fit (dashed line) to the H$\beta$ line profile of AS314 (solid line) obtained in 1998. The fit parameters are presented in the text. The velocity and intensity are given in the same units as in Fig. 6

The discrepancy between the measured radial velocity of AS314 and that expected from the galactic rotation curve might suggest that it is a runaway star and/or a member of a binary system. The proper motion of the star measured by HIPPARCOS, $\mu = 7.4 \,\pm\,2$ milliarcseconds, at D = 10 kpc gives the tangential velocity $350 \,\pm\,100$ kms^-1. Thus, taking into account the radial velocity of nearly 100 kms^-1, one can estimate the total linear velocity of about 365 kms^-1 for AS314. The distance from the galactic plane turns out to be about 600 pc at D=10 kpc. These values are consistent with those expected for a runaway star ejected from a cluster (Leonard [1993]).

In the Hertzsprung-Russell diagram (Fig. 9) AS314 is located close to the positions of the LBVs in quiescence as well as to the instability line suggested by Stothers & Chin ([1994]). This instability is due to the decrease of helium in the star's core during steady helium burning phase. Stothers & Chin also suggested that such a type of instability cannot be encountered by stars with initial masses $\le 30\; M_{\hbox{$\odot$ }}$. Less massive stars may only become dynamically unstable through mass exchange in a close binary system. The latter possibility may well be the case for AS314 because of its radial velocity variations. However, the data obtained so far are still insufficient to derive certain conclusions about binarity of the object. Nevertheless, since the usual uncertainty in determination of the LBVs fundamental parameters is of the order of 20-30 per cent, the position of AS314 is very close to those of HD160529 and HD168607, taking into account these uncertainties. This suggests that binarity may not be crucial for AS314 in order to encounter the instability phase.

Figure 8: IRAS high-resolution maps of AS314 (top) and an artificial point-like source of the same integrated intensity (bottom) at 60 $\mu$m

Figure 9: Hertzsprung-Russell diagram. The zero-age main-sequence and evolutionary tracks from Schaller et al. ([1992]) for stars of different initial masses (denoted by the numbers in $M_{\hbox {$\odot $ }}$ near corresponding tracks) are shown by solid lines, the Humphreys-Davidson limit by a dashed line, while the Stothers & Chin instability line by a dotted line. Positions of LBVs and candidate LBVs collected from literature are shown by filled circles, those of some high-luminosity B[e] supergiants by triangles. HD numbers of the LBVs, which have fundamental parameters and spectral properties closest to those of AS314, mark the corresponding positions. The position of AS314 is shown by the open circle