value as listed in Steenman
& Thé (1991). Although it could be
well possible that other sets of extinction laws, when
available, could produce other 
values (note
that the extinction laws used are normalized to
E(B-V)), however, the important conclusion will
remain that the extinction law changes from star to
star. This is in contradiction with an averaged value
valid for the whole cluster or parts of it as discussed in the introduction and confirms a similar conclusion by Thé et al. (1990).
Figure 6: Histogram of 
values. Part of the first column are the sh
cases, for which we adopted 
3.1, and possible foreground (PF)
objects
As a next step it would be interesting to investigate whether the different extinction laws depend on the star's location in the cluster and to see whether they are connected to the spectral type.
To show the total ``amount'' of anomalous extinction,
a histogram is drawn for the 
values from Table 7 (click here). The first interval
was chosen to be 
, based on the
normal extinction law with 
= 3.1 
0.1 (see also
Savage & Mathis 1979). The following intervals all have a range of 0.3 (a value
slightly larger than the error spread), with
statistics up to 
= 5.0. The results are drawn in Fig. 6 (click here).
More than 50% of our objects have normal extinction.
And besides two exceptions, the 
value is not extreme, in the sense
. Note that we have distinguished
in the first interval of Fig. 6 (click here) the sh cases, for which we have
adopted a normal extinction law, and the (probable) noncluster members from
Table 8 (click here). This reduces the total number of cluster objects with a normal
extinction law from 32 to 11. We also have to notice that some stars have
been counted double, as we treated different options as
being different objects. Because of the different extinction laws found for
these cases, the statistics will not change much.
In this respect, we mention that the two cases with
extreme 
values (W273 and W374) are included here
and not their sh options as discussed in Sect. 5.
Figure 7: Field of NGC 6111 reproduced from Fig. 9 & Table 1 of
Kamp (1974), a). The
central 8
8
area is enlarged in figure
b). The
PMS programme stars (indicated by crosses), as listed in Table 6, and the objects studied
in Thé et al. (1990, diamonds) are indicated by their Walker number and by their individual
value. Objects without a Walker number, the open dots, are other
field stars. The scales are given in
arcmin and centered on star W280
(
)
What remains besides the 11 objects with normal extinction, are 14
objects with a slightly different extinction law and 11 with very
different extinction laws. No peak around some 
value seems to be
present. This would indicate a more or less equal anomalous extinction,
throughout NGC 6611.
To know the anomalous extinction distribution in NGC 6611,
quantitative and qualitatively, we used Figs. 1 (click here)-2 (click here) of Walker (1961) and
Fig. 9 (click here) and Table 1 of Kamp (1974) of the NGC
6611 field. The latter is reproduced in Fig. 7 (click here) in which our
programme stars are indicated. In Fig. 7 (click here) the 
value for each of our
sample stars is also given together with those of Thé et al. (1990),
the main-sequence cases.
- First we consider the objects which exhibit a normal extinction law.
All of those that are indicated as (probably) noncluster members in Table 8 (click here)
occupy the eastern side of the central area and the outskirts of
the cluster. They are all of late spectral type (Group III).
From the sh options 7 out of 9 are located close to the central
part of the central area of the cluster. Note that the sh option can
also include very anomalous extinction.
Five cluster members are scattered throughout the central area of the cluster
(W240, W267, W300, W339 and W455). Of the remaining 7 stars, 6 are
scattered in the eastern outskirts of the cluster and W103 at the
south-west out of the central area.
All of these objects are B-type stars (Group I).
- Most of the objects having a slightly anomalous extinction law, 
,
occupy the central area of the cluster. But W536 lies in the south-east
while W556 and W559 are located in the very west region of the
cluster. All of these stars are (probably) cluster members and are of
spectral type B. Exceptions are W213, A7 or F9, and W396,
F9 and G2.
- The objects having a significant anomalous extinction law, 
, are
also located in the central area of NGC 6611, they are all (probable)
cluster members and are of spectral type B. Exceptions are the B-type star
W290 that is located down south of NGC 6611, but is a confirmed cluster
member, W396 that is of spectral type F9 or G2 and W232 (F8),
for which the membership is uncertain.
Besides a clear distribution of the later type stars with normal extinction
laws to the East of the central area of NGC 6611, which are thought to be
foreground stars, there seems no dependency of the extinction law or the place
in the NGC 6611 areas. For example, as most of the sample stars are located
in the central area, we do have neighbouring stars with normal and anomalous
extinction. The variability in the 
values (from 3.3
to 4.0, and possibly 4.4 or 4.8) suggests that the anomalous extinction law is very variable
in the central area of the cluster. This is not surprising as we have
already mentioned that the intracluster dust is unevenly distributed.
Furthermore, hot stars (the B-types) can influence their direct circumstellar environment
by evaporating and destroying larger grains, causing also large changes in the
values of the intracluster material.
The extinction law of a star located behind such a region will be altered
as well.
In cases of very young objects it is more likely that post-natal
circumstellar material
is the cause of a different than normal extinction. It is then
interesting to note that Group II objects indeed does not have any members
with a normal extinction law.
This could imply that the anomalous extinction law is higher for late spectral types cluster
members as they would occupy an earlier stage of their evolution than, for
example, the averaged B-type members. The Group II objects
could also be lying in a region where the intracluster matter is more
dense, as they are all located in the central region.
Note also that most sh options, being of different spectral type, are located in the central part of the central area.