The open cluster NGC 6611 (see Fig. 8 (click here)) is located in the constellation Serpens, embedded in the Sagittarius spiral arm. It is part of the North-West of M 16 (``The Eagle Nebula''), an extensive bright emission nebulosity with a large gap in the middle of its northern side. It has a system of sharply defined dark markings (dense dust clouds) of which the most conspicuous is located at the southern side, extending inward into the centre. The nebulosity along the northern rim of this great dark marking is brighter than in the other regions.
Besides the ``classical'' North-West population of M 16, which we refer to as the ``centre'', see Figs. 7 (click here)a-b, NGC 6611 is nowadays assumed to range all over the M 16 area. As M 16 is a star forming region, dark clouds obscure the stars lying behind them. When star formation is taking place the young stellar objects (YSOs) are normally first seen at the edges of dark clouds. Many embedded objects were identified in the near-IR by Hillenbrand et al. (1993) in or near the dark central regions of M 16. In a later phase of the young star a bright reflection nebula can appear. In the case of a nebulous region, such as M 16, the emitted light of these newly born stars will heat up and ionize their surrounding circumstellar (CS) material which can be seen as bright nebulae. Because the cluster NGC 6611 is indeed associated with the star forming region M 16, its stellar members must be very young.
It is also possible that these young objects are the origin of a strong radiation field of which the UV part will evaporate parts of the dark nebular regions more efficiently than other parts, causing the existence of the so-called Evaporating Gaseous Globules (EGGs, Hester & Scowen 1995). The impact of the discovery of such EGGs on the star formation (SF) process in NGC 6611 will be discussed in a subsequent paper (Pérez et al. 1997, in preparation).
For the study of pre-main sequence (PMS) stars, very young open clusters are most suitable, because: (a) all cluster members are located at about the same distance. This distance can be well determined employing the spectrophotometric method to early type members, in which the influence of individual anomalous extinction, if any, is taken into account (see Thé et al. 1990); (b) the foreground interstellar extinction can be determined using the early type members (which are undoubtedly true cluster members). This is necessary since the extinction in the direction of PMS objects is due to foreground as well as circumstellar material and the latter often obeys a different extinction law than the former; (c) all cluster members are young, therefore confusion with evolved objects is not to be expected.
One of the first complete studies on NGC 6611 was done by Walker (1961). He constructed the Hertzsprung-Russell Diagram (HRD) of NGC 6611 and found that it consists of a main sequence (MS) extending from O5 to about B5, beyond which stars are lying above the MS. The photometric observations of Walker show that many of the stars in NGC 6611 are reddened by an extremely variable amount, which is roughly correlated with their position in the cluster. Northward the extinction is larger and more irregular, southward the colour excess is less.
After Walker's study many investigations followed, e.g.: Johnson & Borgman (1963), Turner (1974), Sagar & Joshi (1979), Neckel & Chini (1981), Chini & Krugel (1983), Thé et al. (1990). The major conclusion in these studies is that the extinction law in the star forming regions of NGC 6611 is anomalous. The importance of the anomalous extinction (caused by intracluster matter or by circumstellar dust) was recognized because of its influence on the determination of the distance to the cluster, and its effect on the astrophysical parameters of the individual stars. Consequently it will influence the conclusions regarding the evolution of the stellar members, and thus also those of the cluster itself.
Besides Walker (1961), other authors placed many objects several magnitudes above the MS, e.g. Sagar & Joshi (1979), Chini & Wargau (1990) and Hillenbrand et al. (1993). However, Chini & Wargau (1990) critically questioned whether these stars located far above the MS, are true PMS members. From their study they concluded that their 29 programme stars are in reality foreground objects. The problem of recognizing NGC 6611 field stars as true cluster members is already discussed in Thé et al. (1990). The results of a proper-motion survey and the minimum foreground extinction values are the tools for discriminating between cluster members and foreground objects. The results of these studies will be used for our programme stars together with new conclusions based on their location in the HRD of NGC 6611. Given the probable age of NGC 6611, PMS objects are expected to occur. Therefore it is unlikely that all PMS candidates are foreground stars.
The existence of PMS objects in NGC 6611 was shown more recently by
Hillenbrand et al. (1993). They published a study based on UBV and JHK
imaging, down to magnitude 15, over the whole field of M 16.
Placing the objects in HRDs, two cluster populations were identified:
the more massive ones, 
9
,
being in their post-zero-age main sequence (ZAMS) phase, and intermediate
mass stars,
3
8
, being in their pre-ZAMS phase.
From their study the cluster age was determined to be probably less than
(2 
1) Myr. Many objects were found to be located even further
above the MS, meaning that they are just formed.
In the JHK images many objects were found to be embedded.
This could mean that we are actually still observing recent or even
on-going star formation, which indicates a significant age spread
in the cluster population.
This suggestion is supported by the detection of an evolved B2.5 I star of
about 30 
. High mass
star formation must also haven taken place about 6 Myr
ago. As most pre-ZAMS objects studied by Hillenbrand
et al. (1993) do not exhibit
a near-IR excess, which indicates the presence of an (accretion-)disk,
the possible ages will put strong constraints on the disk lifetimes
for intermediate mass stars.
It should be noted, however, that Hillenbrand et al.
(1993) have used the non-individual approach
for the different cluster members. They used colour-colour
diagrams for de-reddening and for investigating the anomalous extinction.
An overall 
value of 3.75 for the anomalous extinction was derived.
Therefore, the results of Hillenbrand et al.
(1993) are not conclusive.
A spectral study of 73 objects was done by Hillenbrand et al. (1993) and they detected several emission line objects. Several are showing properties as Herbig Ae/Be (HAeBe) objects. Other stars have IR- and emission-line properties indicative of classical Ae/Be stars. These authors questioned whether part of the classical Ae/Be-group could actually be young, objects having a gaseous disk left-over from star formation.
In Thé et al. (1990) 52 out of our 89 programme stars with a
high probability of membership in NGC 6611 were studied. The selected stars
are the ones not recognized as PMS candidates.
They confirmed the extreme youth of NGC 6611 and used the location of several
massive stars in the HR-diagram to derive an age not much more than
yrs. They also demonstrated many characteristics
of NGC 6611, such as the above-mentioned points (a) and (b), and
determined some other important parameters.
The distance of 2.6
0.3 kpc and the detection of individual
extinction laws towards each object, rather than an average one for part of
the whole cluster, are the most important ones.
Intrinsic stellar parameters are then easily determined if fundamental
observations, such as low-resolution spectroscopy and photometry,
have been made.
In this study the PMS candidates will be investigated. The selected 49 stars are collected from those listed by Walker (1961), Sagar & Joshi (1979), Chini & Wargau (1990) and given in Thé et al. (1990). We have put no limits for the spectral type, such as the cut-off at A0 in the study of Hillenbrand et al. (1993). In this study we also applied the individual approach. It is more extensive and detailed than the initial PMS search made by de Winter & Thé (1991) in NGC 6611. For all the programme stars low-resolution spectral and photometric observations have been made or collected from previous studies. The spectra and spectral energy distributions are studied and a spectral type and intrinsic magnitudes are derived. This information is used to determine the extinction law for individual objects as well as their astrophysical parameters. Together with the results of our previous study ((Thé et al. 1990), we will study the HRD of NGC 6611, the probability of the detection of Herbig Ae/Be stars in our sample, the distribution of the extinction in the field of NGC 6611 and the membership of the programme stars. Finally we will test the HAeBe candidates on typical characteristics such as the presence of an infrared excess, variability and emission lines.