5 Conclusions and future work

The new high-resolution spectra of MWC 314 obtained in a range from 4200 to 8936 Å allow us to constrain main physical parameters of the star giving additional evidence that it is an early B-type hypergiant close to the Eddington limit. An unusual chemical composition deduced from the identified photospheric lines suggesting a N/O overabundance, which is similar to those of LBV stars AG and HR Car, also supports its status as an LBV candidate. The lack of a surrounding nebula points out that the star has not experienced any eruption yet or it happened a long time ago. Coronographic observations would help to distinguish between the two possibilities (e.g., Nota et al. 1995).

The double-peaked shapes and the absence of the blue-shifted absorption components in most of the emission lines imply a non-spherical geometry of the stellar wind. The emission-line spectrum of MWC 314 is similar to those of B[e] supergiants, such as Hen S22, Hen S134, R126, having almost the same stellar parameters and non-spherical winds (Zickgraf et al. 1986). In contrast to these objects, at present MWC 314 does not show any infrared excess due to radiation of circumstellar dust, however, one can expect a dust formation event around MWC 314 in the future. In this case the observed lack of oxygen would imply the absence of the silicate features at 10 and 18 mum as in the infrared spectra of the B[e] supergiants (Roche et al. 1993).

Thus, our present study confirms that MWC 314 is one of the most luminous stars in the Milky Way and shows importance of its high-resolution spectroscopy. Besides addition of a new information about the star and its wind it allows us to outline directions of the object's future investigation. High-resolution spectroscopic observations are needed to:

1.

look for the photospheric features in the blue region (e.g., absorption lines of C III, Si III, Si IV) which can be used to refine the stellar parameters,

2.

look for the spectral line variability, which is expected in such a luminous object,

3.

check whether He II lines are present or not.

Finally, we would like to emphasize that other types of observations suggested by Miroshnichenko (1996) are also important. For example, photometric monitoring is capable of revealing outbursts which can be expected from this LBV candidate at any time. Radio observations in continuum can help to ascertain our distance estimate.

Acknowledgements

We are grateful to D.J. Lennon for making his high-resolution atlas of supergiants available to us in numerical form and to O. Stahl for supplying us with the spectrum of MWC 300. We wish also to thank the referee for his comments and suggestions which improved our paper.