Astron. Astrophys. 347, 137-150 (1999)

4. Results and discussion

Our photometric observations confirmed the data from the General Catalogue of Variable Stars (Kholopov et al. 1985-1990) that MQ Cas and V1012 Ori show strong brightness variations in the optical region. The other stars have not displayed any variability exceeding in all the passbands used. However, the small number of observations does not allow us to draw more definite conclusions on this subject. The spectroscopic observations revealed an emission component in the H line of BD+11^o829, which suggests that the star should have a significant emission in the H line. The spectra of the three southern objects (AS 116, AS 117, and Hen 938) were obtained in a larger spectral region than in the previous study by Gregorio-Hetem et al. (1992) and with a sufficient resolution to study them in detail.

No reliable estimates of the spectral types of any of our objects have been published before (Thé et al. 1994). We can use the results of our and previously published UBV observations to predict the spectral types of these stars. The averaged and color-indices can be used for the objects with small brightness variations, while for those with Algol-type minima (Grinin et al. 1991) one should use the data in the brightest state. This method works rather well for HAEBEs without a strong emission-line spectrum. The objects with strong emission in the Balmer lines may display additional free-bound radiation shortward of the Balmer jump, which makes their color-indices more negative suggesting an earlier spectral type (e.g., Doazan 1982). Furthermore, circumstellar matter, which is present around the objects, may affect the color-indices making the predicted spectral types uncertain. Below we will compare our photometric estimates with those derived from the spectroscopic observations.

The diagram containing our objects is presented in Fig. 1. It is seen that all of the objects except Hen 938 and possibly AS 116 are located not far from the line of the intrinsic colors for dwarfs (Strajzhys 1977). To predict the spectral types we used the mean color excess ratio = 0.75. The resulting spectral type predictions for our objects along with their averaged color-indices are given in Table 6. The color-indices for MQ Cas and V1012 Ori were averaged in their brightest states. Only the position of V1012 Ori in the color-color diagram is ambiguous and could correspond to any spectral type between B8 and F0. The colors of Hen 938 point to an early-B spectral type. However its high reddening, strong Balmer line emission, and the fact that we used only one UBV observation make this prediction uncertain. Below we discuss the results we obtain for each object separately.

Fig. 1. Color-color diagram containing the studied objects labeled as follows: 1 - MQ Cas, 2 - GSC 1811-0767, 3 - V1012 Ori, 4 - HDE 290380, 5 - BD, 6 - HDE 244604, 7 - AS 116, 8 - AS 117, 9 - Hen 938. Solid line represents intrinsic color-indices for luminosiuty type V from Strajzhys (1977). Dashed line shows the mean reddening vector.

Table 6. Basic information about the objects.
Notes:
a) intrinsic color-indices are taken from Strajzhys (1977)
b) spectral type predicted from the photometry
c) spectral type determined from the spectroscopy (see Gray & Corbally 1998for an explanation of the spectral-type notation)
d) spectral type determined from a co-added spectrum made up of spectra obtained between JD 24501016.87 and JD 24501070.83
e) spectral type determined from a co-added spectrum made from two spectra obtained on JD24501100.84 and JD24501111.83
f) photometric data are taken from Torres et al. (1995)

4.1. Comments on individual objects

4.1.1. MQ Cas

The largest number of photometric observations were obtained for MQ Cas. Our results (see Table 3) for the amplitude of its brightness variations () are in general agreement with the photographic data reported in the literature, although the information about its color-indices is reported for the first time. Analysis of the color-magnitude diagrams (see Fig. 2) shows that the star becomes bluer when it is fainter.

Fig. 2. Color-magnitude diagrams for MQ Cas.

This behaviour can be compared to that of HAEBEs with Algol-type minima, which become redder until they fade by -20 and bluer in deeper minima due to variable circumstellar extinction during obscurations by dust clouds orbiting around the star (Voschinnikov 1989). The "blueing" effect in deep minima is explained by the fact that the contribution of scattered radiation from the circumstellar envelope becomes larger than that of the direct stellar radiation attenuated by the dusty cloud (Grinin et al. 1991). The difference between this general behavior and what we have observed for MQ Cas can be explained if our photometric observations do not include the brightest state of MQ Cas. This is clearly seen in Fig. 2 especially in the behaviour of the color-index.

It might imply that the object's circumstellar envelope is rather dense and scattering on dust plays an important role even if the star is outside a minimum. Therefore, a small additional obscuration of the stellar light is enough to make the optical color-indices bluer. In other words, we observe only the lower part of the whole color-magnitude track. There are some examples of HAEBEs which display a similar behaviour on timescales comparable with that of our observations of MQ Cas (e.g., UX Ori, Grinin et al. 1994).

During our last observing runs in Kazakhstan in August and December 1998 the star was caught in the faintest state ( and 135 respectively). It displayed an almost constant brightness during nearly 2 weeks. In the framework of the hypothesis about the variable circumstellar extinction described above this brightness level is mainly due to the star's radiation being scattered by circumstellar dust, while the contribution of the direct stellar light is much smaller. In such a situation the real amount of the star's obscuration by the dusty cloud may be larger than (e.g., Grinin et al. 1991). Therefore, is probably the absolute minimum state, which can be observed provided the star does not change its luminosity. Since the duration and amplitude of the photometric minima in HAEBEs decrease as the star evolves toward the main sequence, one can assume that MQ Cas is closer to the birthline (Palla & Stahler 1993) than to the zero-age main sequence (ZAMS). At the same time, we should note that this suggestion may not prove to be true because of the rather short period of our observations. For example, an Algol-type HAEBE, UX Ori, which is frequently seen in its brightest state and is thought to be nearly halfway to the ZAMS, recently showed a long-term minimum of nearly the same duration as we detected in MQ Cas (Grinin et al. 1994). Thus, further photometric observations are needed to constrain the brightest level of MQ Cas. Polarimetric data are important to prove its pre-main-sequence nature, as the Algol-type stars show a significant increase of polarization in the minima (Grinin et al. 1991).

All the spectra obtained of MQ Cas have a rather low signal-to-noise ratio because the star was close to its minimum state. In order to derive a reliable spectral classification we studied both single spectra and spectra co-added in various combinations. The resulting spectral type is A0 Vae with an uncertainty of 2 temperature subtypes. The star shows a considerable variation in the H line, which is seen in emission in all our spectra (Fig. 3). Its emission core was found to be essentially centered on the absorption line on 1998 July 22 (JD 2451016), clearly shifted to the red on August 21 (JD 2451046), slightly to the blue on August 23 (JD 2451048), and considerably weaker and slightly shifted to the blue on September 14 (JD 2451070). Such profile variations may be due to an interplay between accretion and the stellar wind, which are both observed in HAEBEs (e.g., Grady et al. 1996). The Balmer decrement appears "normal" in that while H was in emission, H showed only a slight emission core, and H was normal (see Fig. 4d). Within the limits of the noise, no emission in the Fe II series was visible, but weak absorption lines at 4233 Å (Mult. 27) and 5169 Å (Mult. 42) were detected instead. All the spectra show a number of [Fe II ] lines in emission, the strongest of which, 5158 Å [Fe II ] (18F), is shown in Fig. 4d. Other possible forbidden emission lines are as follows: 4244 Å [Fe II ] (21F), 4814 Å [Fe II ] (20F), 5108 Å [Fe II ] (18F), 5262 Å [Fe II ] (19F), 5273 Å [Fe II ] (18F), 5334 Å [Fe II ] (19F), 5376 Å [Fe II ] (19F).

Fig. 3. Variations in the H line of MQ Cas.

Fig. 4a-d. Portions of the classification-resolution spectra obtained with the 0.8-meter telescope of the Dark Sky Observatory. a  HDE 244604: the upper spectrum was obtained on JD 2450818, the lower on JD 2450839. The spectra are shifted by 0.2 Icont with respect to each other; b  BD (solid line) and GSC 1181-0767 (dashed line); c  AS 117; d  MQ Cas: the spectrum shown is the result of co-addition of all four spectra.

Our estimate of the spectral type for MQ Cas, A0 V ae, is in agreement with its large-amplitude variability, which is usually displayed only by late-B - A-type HAEBEs (c.f., Bibo & Thé 1991). Thus, the observed properties of the star strongly favor its PMS nature.

4.1.2. GSC 1811-0767

This star is located close to a pair of HAEBE stars, IP Per and XY Per. Its spectral type, as determined by us, A1 V a^-(er), is in agreement with weak emission at H, which was detected by Gregorio-Hetem et al. (1992). Its visual magnitude observed by these authors is brighter than we found. During our observations the star showed small and slow variations with an amplitude of about in all 5 photometric bands. In general, the color-indices observed by us agree well with those by Gregorio-Hetem et al. (1992), taking into account the difference between the photometric systems (Johnson and Cousins respectively).

Our spectroscopic observations indicate that the hydrogen lines of GSC 1811-0767 are slightly broader than those of the A1 V a standard star, implying that it is quite close to the ZAMS. There is a slight infilling of the H line, and this emission is shifted slightly to the red. The Balmer decrement is normal, i.e. H is also very slightly filled in, but H is normal. No emission in the Fe II (42) series is visible (see Fig 4b). Thus, the spectroscopic determination of the spectral type is in good agreement with the photometric estimate.

Near-IR observations are strongly desirable for GSC 1811-0767 to check whether it has excess radiation in this spectral region, which would be strong evidence of its youth. Nevertheless, at the moment there is no contradiction with the suggestion that this star is a HAEBE candidate.

4.1.3. V1012 Ori

V1012 Ori is known as a large-amplitude variable star (m_pg = 12^m - 16^m, Kholopov et al. 1985-1990). However, its first photoelectric observations were published only recently (Cieslinski et al. 1997). Three observations obtained by these authors in October 1989 and March 1991 caught the star near its brightest state (-). Our more extended data set shows that the star spends most of its time near the brightest state. The color-indices obtained by Cieslinski et al. (1997) and us are in good agreement. However, we detected a short-term (on the order of a few days) deep minimum in November 1996 (see Table 2), which is similar to those of the Algol-type HAEBEs. JHK observations of Nakano (1998) showed that V1012 Ori has a noticeable near-IR excess (, , ). These facts make the suggestion by Thé et al. (1994) that it is a PMS star more likely.

The photometry we have obtained for V1012 Ori is consistent with any spectral type between B8 and F0, depending upon its reddening. We were able to obtain two classification-resolution spectra of V1012 Ori. These spectra, which are essentially identical, were co-added to increase the signal-to-noise ratio (see Fig. 5). The resulting spectral type is A3 II shell?, with no indication of emission, either in the hydrogen lines or in lines of the Fe II (42) multiplet. However, slightly peculiar hydrogen-line profiles may indicate the presence of a shell.

Fig. 5a-d. Portions of the classification-resolution spectra obtained at Dark Sky Observatory. a , b  V1012 Ori (solid line) in comparison with HR 146 (A3 II , dashed line); c , d  HDE 290380.

The luminosity type suggests that this star is still quite far from the ZAMS. A straight-forward calculation of the distance based on the reddening and absolute magnitude () derived from the spectral type implies a distance of nearly 6 kpc, which is highly unlikely considering the moderate reddening and the fact that this star is in the galactic anticenter direction. However, the presence of a shell has probably caused us to overestimate the luminosity of this star.

4.1.4. HDE 290380

The presence of emission lines in the spectrum of HDE 290380 was noted by MacConnell (1982). Later it was considered to be a transition object between AGB and Planetary Nebulae on the basis of its IRAS and near-IR color-indices (Garciá-Lario et al. 1997). Torres et al. (1995) obtained a low-resolution spectrum and photometry and suggested that it is a Herbig Ae/Be star with an F0 spectral type. These authors detected moderate emission in the H line with equivalent width (EW) = 7 Å, showing a single-peaked profile and a weak indication of a P Cygni-type blueshifted absorption. Recently Garciá-Lario et al. (1997) agreed with the PMS nature of HDE 290380.

Our photometry is in excellent agreement with the published data except for the K-band. Our K-magnitude is brighter than that of Garciá-Lario et al. (1997); this can be explained by the difference in the photometric systems used. The object exhibits a significant near-IR excess which is more likely due to radiation from circumstellar dust rather than free-free emission, because the emission-line spectrum is weak. Its optical color-indices are close to the intrinsic ones for mid-F type stars.

The spectrum is very similar to that of the F6 IV standard. However, it shows emission in the Ca II H and K lines: both lines are at only about 70% of normal depth and the K line has a clear reversal in the core. The H line clearly has blueshifted emission in the core, while the core of H is slightly shallow. There are no signs of emission in the Fe II (42) lines.

4.1.5. AS 116

AS 116 was discovered to have an IRAS counterpart by Dong & Hu (1991). Gregorio-Hetem et al. (1992) detected strong emission in the H line (EW = 102 Å). The optical photometry obtained by these authors suggests that AS 116 is a reddened early B-type star. The SAAO near-IR photometry shows that the star has a near-IR excess, which is likely due to hot circumstellar dust. Its brightness in the L-band (), which was measured on JD 2450506.34, gives additional support to this suggestion.

We have obtained spectra of AS 116 both from Brazil and DSO in the United States. The Brazilian low-resolution spectrum (see Fig. 6ab) shows a number of emission lines including a bright H and H, weak Fe II lines of the multiplets 42 and 49, and a forbidden line of O I at 6300 Å. The He I lines at 5876 and 6678 Å appear in absorption, however a weak emission component is seen in the second line. Several weak interstellar features (Na I lines, diffuse interstellar bands at 5780, 6278, and 6283 Å) are consistent with a moderate reddening of the star. The widths of the Balmer lines are about 400 km s^-1, indicating a low terminal velocity of the stellar wind. Absence of a P Cygni-type absorption in the H profile might imply a non-spherical circumstellar envelope. However this must be supported by more high-resolution observations.

Fig. 6a-f. Portions of the low-resolution spectra obtained with the 1.6-meter telescope of the Laboratorio Nacional de Astrofisica at Pico dos Dias. a,b AS 116, c,d AS 117, e,f Hen 938.

The two classification-resolution spectra from DSO agree well with the Brazilian spectrum in the region of overlap. These spectra also show that emission is present in the cores of the hydrogen lines up to H8, the Ca II K-line appears to be in emission, as well as many Fe II lines (see Fig. 7). All the emission lines are slightly stronger in the second spectrum, while the He I absorption lines appear the same in both spectra. The hydrogen-line profiles and the He I line strengths both indicate a spectral type of B7 V e, taking into account the distortion by the shell emission. A spectrum of HR 1029 (B7 V ) obtained at DSO is shown in Fig. 7 for comparison.

Fig. 7. Portions of the classification-resolution spectra obtained at Dark Sky Observatory. AS 116 (solid line) and HR 1029 (dashed line)

Both sets of our photometric observations (Tien-Shan Observatory and SAAO) show that AS 116 is variable. In December 1998 it faded by in comparison with the earlier data. This process was accompanied by an increase in the and color-indices and a decrease of the . While the increases in and are consistent with the variable circumstellar extinction, the decrease in is probably due to a brightening in the Balmer continuum which, in turn, might be caused by an increase in the emission-line activity, such as was observed at DSO.

The discrepancy between the spectral type derived from photometry and that from spectroscopy may be due to such effects as a contribution of the H emission line to the V-band flux and an inverse Balmer jump from free-bound radiation of the circumstellar envelope. The latter is frequently observed in the spectra of classical Be stars which normally have weaker line emission than that of AS 116. It has been shown that the excess, which is introduced by a circumstellar envelope producing the H emission line with EW 100 Å, can be as large as (cf. Doazan 1982). This is actually the shift in the position of AS 116 in the color-color diagram (Fig. 1) in from that of a normal reddened star with a B7 spectral type. Furthermore, the Balmer decrement in AS 116 is much steeper than seen in emission-line stars with an H line of comparable strength such as HD 200775 (B3, Beskrovnaya et al. 1994) and HD 45677 (B2, Israelian et al. 1996). This points to a later spectral type for AS 116. On the other hand, the star may have a spectral type earlier than B7 and a larger rotational velocity than that of HR 1029 (100 km s^-1, Abt & Morrell 1995). Additionally, its helium lines may be partly filled in with emission. Summarizing all the above, one can conclude that the true effective temperature of AS 116 lies between B3 and B7. A more definite answer to this question would come from high-resolution spectroscopy and modeling of the line profiles.

The above mentioned spectral features are common to three types of stars with near-IR excesses: HAEBEs, B[e] supergiants, and LBVs. The latter two have high luminosities (), which would place AS 116 at a distance of at least 10 kpc from the Sun. This seems unlikely, because the star is located towards the galactic anticenter. In addition, such a high luminosity is inconsistent with the spectral type. On the other hand, its rather weak near- and far-IR excess, lack of evidence for accretion and its spectral type suggest that it is a pre-main-sequence star close to the ZAMS. In this case its luminosity would be about (cf. Strajzhys & Kurilene 1981) giving a distance on the order of 1 kpc. The emission-line spectrum of AS 116 favors this hypothesis, because B[e] supergiants and LBVs of a similar temperature usually display a large number of strong metallic emission lines. For example, Lopes et al. (1992) in their spectroscopic study of luminous peculiar B-type stars pointed out that the strength of the emission lines (namely H and Fe II (42) 4924 Å) is possibly luminosity dependent. Our data for these lines in AS 116 suggest that its luminosity is rather low.

4.1.6. AS 117

AS 117 was selected by Dong & Hu (1991) in the same way as AS 116. Gregorio-Hetem et al. (1992) obtained one photometric observation and a spectrum showing weak emission at H (EW = 7 Å). This star is less reddened than AS 116 and has a lower effective temperature. Our JHK photometry shows that its near-IR excess is weak. Again, we have two spectra for this star, one from Brazil, the other from DSO. The Brazilian spectrum shows only H in emission, while the H line is essentially photospheric (Fig. 6cd). On the other hand, the DSO spectrum shows emission in the core of H, and slight emission in the cores of all the hydrogen lines up to H8 (Fig. 4c). The Fe II (42) lines are filled with emission. The wings of the Balmer lines are slightly broader than the A0 V a standard. Thus, we assign this star a spectral type of A0 V a^-e, suggesting that it lies near to the ZAMS. The low reddening suggests that AS 117 is located not far from the Sun and, hence, has a low luminosity. Recently Yudin & Evans (1998) reported that the star shows a polarization of about 0.5% in the V-band. This fact is consistent with the star being relatively nearby.

The weakness of its emission at H points to a weak stellar wind, which, along with the small IR excess, also implies that the star is very close to the ZAMS (e.g., Miroshnichenko et al. 1996). Its observed properties are similar to those of such HAEBEs as HD 37411 (Hu et al. 1989) and V351 Ori (van den Ancker et al. 1996). At the same time, one can not exclude the possibility that the dusty envelope of AS 117 may be clumpy (like those in HAEBEs with Algol-type minima). In December 1998 we observed a slight decrease of its visual brightness which might be an evidence of such a structure. This event coincided with a similar fading of AS 116, which was observed with the same comparison star, HD 40745. The brightness difference of HD 40745 and the check star, HD 45629, was constant within the observational errors in all five photometric bands used. Thus, the observed variations are real. However, more frequent photometric observations are needed to study these two stars in more detail.

Table 7. Spectral lines in the Brazilian spectrum of AS 116.
Notes:
a) blend with 4929.2 Å
b) blend with 5023.9 Å
c) blend of 6278 and 6283 Å

Table 8. Spectral lines in AS 117.
Notes:
a) blend of 6278 and 6283 Å

4.1.7. Hen 938

Hen 938 was selected by Allen & Swings (1976) as a peculiar Be (or B[e]) type star because it displayed forbidden emission lines in the optical spectrum and a noticeable near-IR excess. In its spectrum, which was obtained with an image tube at a resolution of 140 Å mm^-1, these authors found 5 Balmer lines, 7 Fe II lines, and weak [O I ] lines in emission and the Ca I H and K lines in absorption. They also reported the presence of Ti O absorption bands, which would imply the presence of a late-type companion.

As shown in Figs. 6e and 6f, our spectrum of Hen 938 is enriched with metallic emission lines. We found 37 lines of Fe II , 2 of Ti II , 2 of Si II , and 1 of [Fe II ]. There is also a weak line of [O I ] at 6300 Å. No signs of the TiO absorption bands or other features of late-type stars were found. The He I lines at 5876 and 6678 Å have P Cygni profiles indicating a strong stellar wind and a high temperature ( 20000 K). The Balmer lines (H and H) are strongly in emission (see Table 10), and H has an almost symmetric profile without any absorption components, which may be evidence of a non-spherical outflow. Its equivalent width, 150 Å, is almost twice as strong as that in the spectrum taken by Gregorio-Hetem et al. (1992), 77 Å.

Table 9. Spectral lines in Hen 938.
Notes:
a) EW of the emission part of P Cyg profile
b) from Johansson (1978)
c) blend with Fe II (40) 6368 Å
"p" refers to the preceding line, "f" to the following line which are blended.

Table 10. Characteristics of the Balmer lines from the Brazilian spectra.
Notes:
a) H is in absorption

All these facts show that the emission lines in the spectrum of Hen 938 change on short time scales. At the same time, the object did not display any significant brightness changes between the early 1970's (m_V = 13:m 3, Allen & Swings 1976) and the mid 1990's (V = 13:m 52, Torres et al. 1995). The described spectral features suggest that Hen 938 has greater similarities to B[e] supergiants than to HAEBEs. Indeed, there are only a few early B-type HAEBE candidates with such a large number of metallic lines in emission (e.g., MWC 137, MWC 300). However, both these stars still have a controversial evolutionary state (e.g., Wolf & Stahl 1985and Esteban & Fernández 1998).

Another suggestion about the nature of Hen 938 is that it might be a star evolving towards the planetary nebula stage. Its photometric and spectroscopic variations as well as its spectral appearance are similar to those of OY Gem = HD 51585, which is thought to be a post-AGB star (e.g., Arkhipova & Ikonnikova 1992). In this case Hen 938 seems to be less evolved than OY Gem, because it displays only a few weak forbidden emission lines, while OY Gem shows a large number including some of high-excitation (e.g. [O III ]). This implies that we can expect a further increase in the emission line strength in Hen 938.

At the same time, other methods may be applied in order to investigate whether Hen 938 is a young or an evolved star. For example, radial velocity measurements may help to constrain the distance towards the star, while mid- and far-IR spectroscopy may give information about the composition of the circumstellar dust. The low-resolution mid-IR spectrum obtained by IRAS (Olnon et al. 1986) is essentially featureless; this may imply the absence of crystalline structures. Recent results obtained by ISO show that amorphous circumstellar dust is common in both young stars and evolved objects, where it was formed a long time ago (Voors 1998).

4.1.8. HDE 244604 and BD+11°829

As we pointed out above, these two stars are located in a star formation region in upper Orion and are situated within a few degrees of each other. HDE 244604 was identified with an IRAS source by Oudmaijer et al. (1992). Recently Malfait et al. (1998) showed that it also has a near-IR excess. According to our photometric data, the star is rather stable () except for one observation of 1998 December 9. Its brightness level is close to that indicated by Malfait et al. (1998), .

We obtained two spectra of the star in the blue-yellow region on two different nights. In the first spectrum (JD 2450818), HDE 244604 showed a slight filling-in of the H core due to emission. This core emission was slightly shifted to the blue with respect to the absorption line (Fig 4a). A close comparison of this spectrum with that of the A3 V standard shows a slight infilling of the three Fe II (42) lines at 4923, 5018 and 5169 Å which is not readily apparent in the figure. The spectral type was determined to be A3 V a⁺ (eb) Nem1 for this date. In the second spectrum (JD 2450839), emission in the H line had disappeared, as well as the weak emission in the Fe II (42) lines. The strength of the Ca II K-line became stronger indicating a K-line type of A4, while the general metallic spectrum and the hydrogen lines gave an A3 type. In both spectra, the hydrogen lines were slightly more narrow than the A3 V a standard, Leo. Thus, on JD 2450839, the spectral type was kA4 hA3 mA3 V a⁺.

BD is fainter and slightly more reddened than HDE 244604. Its near-IR excess is comparable with that of HDE 244604 ( and respectively). The photometric temperature type is B9, but the variations in the color-indices make this estimate uncertain. The H line in the spectrum of BD shows a fairly strong emission component (Fig. 4b). The emission core is shifted slightly to the red with respect to the absorption line. The Balmer decrement is very weak, meaning that emission is seen up to quite high Balmer lines (indeed, the cores are shallow all the way up to H9 - the highest Balmer line in our spectrum). The K-line is weak, and this is probably due to an emission component, as the core of H (which is blended with the Ca II H-line) also shows an asymmetry (with strongest emission to the blue side), probably due to Ca II H-line emission. The Fe II (42) lines show a noticeable P-Cygni profile. This star has a temperature type of A3.

Both stars show signs of photometric and spectroscopic activity. However, if we assume the same distance to the stars and equal temperatures, then HDE 244604 would be more luminous than BD, farther from the ZAMS, and more massive. At the same time, the difference in temperature derived from the color-indices and the spectrum, which were obtained at different times, might imply that the temperature of BD is not well constrained. For instance, low-resolution spectra of an HAEBE, AB Aur, obtained with the same equipment on different nights, resulted in different spectral types, A0 and A3 (Gray & Corbally 1998). Thus, if BD has an earlier spectral type than HDE 244604, the stars may have almost the same luminosity (given the same distance), because their brightness difference may be due to the difference in the circumstellar extinctions. As a result, BD may have a larger mass than HDE 244604, but still be closer to the ZAMS.

© European Southern Observatory (ESO) 1999

Online publication: June 18, 1999
helpdesk.link@springer.de