We have estimated the distances to the stars as follows. For the early-B-type group, we have assumed a typical spectral type of B2V and a corresponding absolute V magnitude of -2.4 [26, (Schmidt-Kaler 1982).] The largest uncertainty is in the interstellar extinction and here we have just adopted the value of for LS4825 - a star, which is believed to be beyond the galactic centre [25, (Ryans et al. 1997)] and has . Then most of our early-type targets lie in a magnitude range of , which would map onto a distance scale of 6 to 30 kpc. Similarly for the mid-B-type subgroup, adopting an absolute V magnitude of -1.0 and an apparent magnitude range of would lead to distances from 1.3 to 10 kpc. We should emphasise that given the unknown (and possibly highly variable) extinction towards these targets, we consider these distance estimates to be illustrative. However, given that most of the early B-type candidates have galactic coordinates which indicate that they may sit slightly out of the plane (Table7), one would expect that they would have similar reddenings (and hence distances) to the distant objects discussed in [32, Smartt et al. (1997, 1999).] The latter have , similar to the values shown in Table7, and have reddenings in the range 0.3 < E(B-V) < 0.5. If these values are equally applicable to our newly identified stars, then their magnitudes and approximate atmospheric parameters would be compatible with them being near-galactic centre objects.
As discussed above, at least some of our early-type stars may be subluminous and evolved. Indeed it is interesting that the apparent magnitude range for the early-B-type stars is approximately one magnitude fainter () than that for their mid-B-type counterparts (). This is despite the former being intrinsically brighter by one to two magnitudes, if they are young hydrogen burning objects. Hence at least for the early-B-type stars, our sample may be contaminated by post blue-horizontal-branch (post-BHB) and post-AGB stars (see, for example, [10, Hambly et al. 1997,] and references therein). Alternatively it may just reflect the relatively small sample sizes or a different spatial distribution.
Our early-B-type objects have gravities of 3.2-4.0 dex (adopting their lower effective temperature limit of 20000 K for all stars apart from [456,18] and [456,30]); alternatively if a typical effective temperature of 25000 K was assumed their gravity range would be approximately 3.6-4.2 dex. The evolutionary tracks of [28, Schönberner (1993] and references therein) implies that post-AGB stars have gravities in the ranges 2.4-3.4 and 2.8-3.8 dex for effective temperatures of 20000 and 25000 K respectively. By contrast a post-BHB star with a mass of has a gravity of 3.7 dex for an effective temperature 25000 K [4, (Castellani et al. 1994).] Hence although some of our candidates could be post-AGB stars they are more likely to be post-BHB stars. In any case, if they are evolved then independent of their exact evolutionary status, they would have a mass of approximately 0.5-0.6 .
Adopting a typical mass of 10-12 for a main sequence early-B-type star [27, (Schaller et al. 1992),] leads to a mass ratio between the two evolutionary scenarios of approximately 20. Hence if all our early-B-type sample were evolved, their range of distances estimated above would be reduced to approximately 1-6 kpc. Although this possibility cannot be excluded, our identification of normal early-B-type stars towards the galactic centre [32, (Smartt et al. 1997, 1999)] makes it unlikely. Hence we conclude that although some of our early-type candidates are probably evolved, the sample should also contain young hydrogen burning stars.
We wish to thank the staff of the Anglo-Australian Observatory and UKSTU at the Royal Observatory Edinburgh for obtaining additional Schmidt plates and for assistance with the low dispersion spectroscopy. We are grateful to Harvey MacGillivray and Eve Thomson for measuring the plates on SuperCOSMOS. Part of the data reduction was undertaken on the PPARC funded Starlink node at Queen's University. PPARC financial support including that for a visiting fellowship programme at QUB is gratefully acknowledged.
Copyright The European Southern Observatory (ESO)