We now compare the results of our study to those of previous Be star surveys within NGC 330 and NGC 1818 in order to get some idea of our detection efficiency. Although there have been a few reports of detections of Be stars in other clusters (e.g. Kjeldsen & Baade 1994), it is only for NGC 330 and NGC 1818 that any information is available regarding the identity of the Be stars.
Grebel et al. (1992) have searched the region close to NGC 330 for Be stars using narrow-band H photometry. Within 50 of the cluster centre they find 29 Be stars while we find 32. However, the Be star samples found in the two respective searches are not completely identical. Of the 18 Be stars for which Grebel et al. provide identifications (they list names from Robertson 1974 and Arp 1959), we find 16. We suspect that variability of the Be phenomenon may account for the differences in the two Be star samples (see Sect. 4.2).Overall, however, the two searches find similar numbers of stars. Our search area is considerably larger than that of Grebel et al. which will allow us to compare the properties of reasonable samples of both field and cluster Be stars.
NGC 330 and its surroundings have been searched extensively for variable stars by Sebo & Wood (1994) and Balona (1992). A number of the Be stars and one of the H emitting red giants found here are known variables. The variable star identifications are listed in Table 2. A spectrum of the variable red giant was obtained by Sebo & Wood (1994) and it does indeed show a broad H emission line.
Grebel (1997) finds 21 Be stars with V < 17.5 (the magnitude limit of our study), of which we have identified 13 in common. Of the 8 unidentified stars, 7 are near our magnitude limit and have 17.1 < V < 17.5: furthermore, most lie in crowded regions. More interestingly, we detected 6 Be stars that have no counterparts in the survey of Grebel. This is perhaps surprising given the fainter magnitude limit and better seeing of the Grebel survey. These stars, which are all comparatively bright and lie in uncrowded regions, are unlikely to be misidentifications. We thus seem to have strong evidence for the episodic nature of the Be phenomenon in Magellanic Cloud stars: such behaviour is frequently seen in galactic Be stars (Hanuschik et al. 1993). It is also possible that some of the Be stars found by Grebel, but which we failed to detect, were not exhibiting H emission at the time of our observation.
In summary, both the present survey and the surveys of Grebel et al. (1992) and Grebel (1997) find similar numbers of Be stars, although our survey may be incomplete near the magnitude limit of V = 17.5 in crowded regions. Because of the episodic nature of the Be phenomenon, single-epoch surveys may miss a significant fraction of the total Be star population (20% appears to have been missed in the case of NGC 1818).
Copyright The European Southern Observatory (ESO)