4. Spectral types
4.1. Sk-66o41 cluster
It was not possible to obtain the individual spectra of the components of Sk-66o41. Therefore the spectrum (Fig. 5) represents the whole cluster dominated by the brightest central stars (#33, #34, #36). If one were to force a single spectral type, it would be O6.5 V ((f)). However, that classification is not satisfactory, because there are opposite discrepancies. The N iv 4058 emission (stronger than N iii 4634-40) and N v 4604 absorption show the presence of a (non-dominant) O3 component, while the weaker He i lines such as 4009, 4144, 4387 require a component at least as late as O8. Hence, we derive O3 V((f*)) + OB. This result revises the previously published classification O5 V, based on lower S/N spectra (Paper I).
4.2. HNT cluster
Fig. 6 displays the spectra of some of the brightest stars in the HNT cluster. To limit the contamination by nearby fainter stars, the spectra were extracted over a rather narrow range of four pixels (") centered on the location of the relevant star along the slit projected on the sky. We caution however that the resulting spectra are slightly contaminated by the nearby fainter stars, especially near the crowded cluster center.
Our spectra of the brighter members of the HNT cluster correspond to stars of spectral type A-F. We have classified these spectra using the criteria described by Gray & Garrison (1987, 1989a,b). While these criteria are well established for Galactic stars, their application to stars in the LMC is more ambiguous because of the well known metal deficiency of the latter galaxy. Therefore, there is a dispersion among the classifications based upon different criteria for the same star. We find that the ratio of the strength of the CaII K line with respect to the CaII H + H line yields an earlier spectral type than the other criteria based for instance on the strength of the Balmer lines. Since the CaII lines are weakened compared to Galactic standard stars, we favor the classification based on the other criteria. Table 3 lists the results for five bright stars in the HNT cluster.
Table 3. Spectral classification and radial velocities of the stars in the HNT cluster. The radial velocities are derived from the Balmer lines excluding the H line which is heavily blended with the CaII H line. The quoted absolute magnitudes are derived from our photometry assuming an average mag. The last column lists the identification of the fainter nearby stars that are likely to contribute to the spectra.
The luminosity classification of A-F stars largely relies on the intensities of metal lines (e.g. Gray & Garrison 1987, 1989a, b) and is therefore rather unreliable for the LMC stars discussed here. We emphasize however that the shape of the wings of the Balmer lines points towards a giant or even supergiant luminosity classification. Adopting a distance modulus of 18.5 and assuming an average mag towards N 11C (Paper I), we derive the absolute visual magnitudes listed in Table 3. These values roughly correspond to luminosity class III - II. We caution however that these results might be affected by multiplicity of the stars considered here.
We note that our spectroscopic data reveal no trace of a star earlier than A0 in the HNT cluster.
4.3. Spectral types of the field stars
A number of the field stars were studied also spectroscopically. The spectrograms of these stars are displayed in Fig. 5 and Fig. 7. Table 4 summarizes the spectral types and the measured radial velocities. Almost all are O or B types. In the following we discuss the details of some of the individual spectra.
Table 4. Spectral types and radial velocities of the field OB stars
Wo599. This star exhibits an interesting composite spectrum. Although the N iv 4058 emission is prominent, it is weaker than N iii 4634-40. The presence of C iii 4650 emission is noteworthy. The only feature that prevents a pure O4 classification is He i 4387, which is somewhat strange (no other anomalously strong He i lines), and it is broader than the other lines, which is consistent with an origin in other star(s). The required components are most probably not very late and the earlier type must dominate, otherwise He i 4471 would be stronger. We classify Wo599 O4 V ((f+)) + O. The "+" tag means that the Si iv 4089, 4116 lines are in emission; when only the latter is visible, as in this case, it is presumed that the former is canceled by a relatively stronger absorption line.
Wo622. The classification of this star is ambiguous between O9.7 III and B0.2 V (both have He II 4541 Si iii 4552). The main luminosity criterion is He II 4686/He i 4713 and if this is a late O star, it has to be a giant from that ratio (Walborn & Fitzpatrick 1990). Should the bump at the blue edge of H be Si iv 4089, the higher luminosity class is supported, but the presence of this line is uncertain. On the other hand, if the type is as late as B0.2 then the He II 4686/He i 4713 ratio would be consistent with class V. Given the resolution and S/N, we cannot decide. Note however that the derived absolute magnitude (-4.1) is consistent with the later type (Vacca et al. 1996).
Star #211. The classification is B1 V, provided that a feature near He II 4686 is noise. If it were real, the type would have to be earlier.
Star #475. This star lies 45" east of Wo622 on a line joining Wo622 to the middle of HNT, outside the field of our images. Some features of the spectrum look like those of a mid- or late-B supergiant, but they are not consistent. For example, there may be strong C ii 4267, but there is no trace of Mg II 4481, which should be very strong.
Star #476. This star lies 65" west of #204 on a line joining it to #211. It is an O6 V, assuming that a feature at 4387 is noise, since there are three other comparable ones near H. Otherwise He i 4387 would be too strong, but everything else is consistent.
Star #477. This star lies 46" east of Sk-66o41. Its spectral type is B0-1 V depending on the feature near He II 4686. If real, the earlier type applies; if not, the later.
Wo 647. We classify this star as F7-F8. Contrary to what happens for the members of the HNT cluster, the line intensities in the spectrum of Wo 647 match those of the Galactic main sequence or giant standard stars. This star has a radial velocity of km s-1. Our results confirm that Wo 647 is most probably a Galactic foreground star as already suspected in Paper I.
© European Southern Observatory (ESO) 2000
Online publication: January 29, 2001