New photometric data in the Walraven WULBV system of our programme stars
were obtained during three observing runs, in May 1982, June 1983, and in
August 1985 using the 90 cm Dutch Light Collector at La Silla. Measuring and
reduction procedures were the same as described in Lub & Pel
(1977), except that a 16
5 diaphragm was used. In order to compare
these data with the photometric data from literature in the Johnson UBV
system we transformed V intensities in the Walraven WULBV system to
Johnson V magnitudes using the formula quoted in Brand & Wouterloot
(1988):
These transformed 
magnitudes are listed in the third column of
Table 2.
Photometric data in the Johnson/Cousins 
system of our
programme stars in NGC 6530 were obtained during four observing runs, in May
1984 at the Cerro Tololo Inter-American Observatory (CTIO), and in June
1983, July 1985 and March 1991 at the European Southern Observatory (ESO),
La Silla, Chile. At CTIO the observations were carried out using a
16
diaphragm at the 60 cm Lowell telescope, equipped with an RCA
31034A (Quantacon) photomultiplier. At La Silla the observations were made
with the ESO 50 cm telescope equipped with an identical photomultiplier and
a 15
diaphragm.
Sky subtraction was achieved by subtracting sky measurements 30
east
from the star, unless that position showed a significantly higher than
background flux, in which case a sky measurement 30
west of the star
was obtained and used. About 35 E-region standard stars from the list of
Graham (1982) were observed each night and used for the
determination of the extinction and colour transformation parameters. In
this way the obtained typical errors for stars brighter than 125 are
001 for V, B-V, V-R and V-I, and 002 for U-B. For fainter
stars these results are about 001 less certain. However, since the
observed set of E-region standards for the observing runs at ESO in 1983
and 1985 did not include stars later than F6, transformation errors for
stars with very red colours can be much larger in these data.
In order to compare our new UBV photometric data with those measured by
previous authors we computed the average difference between our measurements
and those by Chini & Neckel (1981), with 
, together with the same difference between
their data set and those by other authors. Omitted from this systematic
difference calculation were 11 stars with differences larger than 030 in
V between the various data sets. For stars which were only measured 2 or 3
times this difference might be due to stellar misidentifications, whereas
for stars which were measured more often these stars are probably real
variables. These 11 suspected variable stars are listed in Table 3 (click here).
The systematic differences between our data set as well as several data sets
from literature with the one by Chini & Neckel (1981) are
listed in Table 4 (click here). In cases where this systematic difference is
significant (i.e. the difference is larger than its standard deviation
listed in Table 4 (click here)), the data have been corrected for this in the
remainder of this paper.
Table 3: Suspected variable stars in NGC 6530 (
)
Near-IR JHK photometric data of stars in NGC 6530 were obtained in July
1986 with the ESO 1 m telescope at La Silla, equipped with an InSb detector.
These observations were made through a 15
diaphragm. Sky subtraction
was achieved by chopping, with a frequency of 8 Hz, in the east-west
direction with a throw of 30
amplitude. About 35 standard stars from
a preliminary version of the list later published by Bouchet et al.
(1989) were observed each night and used for the determination of
the extinction parameters. Typical errors in the thus data are about 005
for the J, H and K magnitudes. All new photometric measurements are
listed in Table 2.
Figure 1: Examples of spectra of programme stars in NGC 6530 taken with the
ESO 1.5 m telescope with IDS detector. The top line in the plot for
NGC 6530-245 shows an enlarged version of the spectrum
Figure 2: Spectra of programme stars in NGC 6530 taken with the ESO 1.5 m
telescope with CCD detector
Table 4: Comparison of NGC 6530 photometric data sets with Chini & Neckel
(1981), with 
Spectroscopic data of our programme stars were obtained during three observing
runs, in June 1983, May 1985, and in June 1992, with the ESO 1.52 m telescope
at the European Southern Observatory, La Silla, Chile. In 1983 and in 1985 the
telescope was equipped with an Image Dissector Scanner (IDS) mounted at the
Boller and Chivens spectrograph, whereas in 1992 the telescope was equipped
with a Ford Aerospace 2048 CCD and an identical spectrograph. The IDS spectra
were obtained through a 
aperture, whereas for the
CCD long-slit spectra a slit with a width of 2
was used. During all
observing runs a grating with a dispersion of 172 Å mm
, centered at
5600 Å, was used. The 1983 and 1985 spectra were reduced using IHAP, whereas
the 1992 spectra were reduced using MIDAS. All three sets of spectra were
reduced at ESO Headquarters, Garching bei München, Germany. In order to
obtain an optimal S/N ratio for our spectra the optimal extraction algorithm
by Horne (1986) was employed for the extraction of the 1992
spectral data from the CCD images. In the IDS spectra, sky subtraction was
achieved by subtracting a simultaneously measured reference spectrum 2
arcminutes from the star, whereas for the CCD spectra sky subtraction was
achieved by subtracting a third degree polynomial fitted to the spectrum
100 pixels (corresponding to roughly 25
) in the spatial direction of
the stellar spectrum. Because of the two-dimensional nature of the 1992 CCD
long-slit spectra the results of the employed correction for the emission from
surrounding nebulosity are clearly superior to the results of this correction
in the 1983 and 1985 spectra. It should be noted here, however, that the sky
subtraction in both cases sometimes leaves some residual emission, especially
in H
, of which it is not clear whether this is real or if this is due
to the surrounding nebulosity. Examples of the 1983 and 1985 IDS spectra are
shown in Fig. 1 (click here). The absolute flux scales in these figures compare
within 010 with the measured photometry. All 1992 CCD spectra are shown
in Fig. 2 (click here).
New spectral classifications for our programme stars, in the MK system, were
made by comparing the spectra with each other as well as with the standard MK
spectra by Jacoby et al. (1984). The results of these
classifications are listed in Table 5. The error in the resulting spectral
types will generally be smaller than 2 subclasses, although for spectra with
a poor S/N ratio the errors may be much larger (indicated by a colon
following the spectral type in Table 5). The stars in which some residual
H
emission of unclear origin is seen are classified as e:. Whether
these small emission components are due to intrinsic H
emission or
are due to imperfect sky subtraction remains unclear. The stars labelled
with e certainly do possess intrinsic H
emission, however.