The environment of Herbig Ae/Be stars (HAeBe) has been a source of debate for many years (see, for example, Waters & Waelkens 1998). It is clear, today, that there are large differences between the most massive stars in this group (corresponding to spectral types B5-B0; HBe in the following), and the less massive ones, that will be called for simplicity in the following HAe stars (Natta et al. 2000a). In HAe stars, we have strong evidence of the presence of circumstellar disks from millimeter interferometry (Mannings & Sargent 1997, 2000) and from direct images in the visual (Grady et al. 1999, 2000). This is not the case for HBe stars. However, even for HAe stars there is great uncertainty on the structure of the disks, with arguments for the existence of a roughly spherically distributed dust component (Di Francesco et al. 1994). As a consequence, the spectral energy distributions (SED) of HAe stars have been interpreted as the emission of extended spherical envelopes of low optical depth (Pezzuto et al. 1997; Miroshnichenko et al. 1997; Il'in & Krivova 2000), of circumstellar disks (Hillenbrand et al. 1995; Chiang et al. 2001), or of a combination of both (Natta et al. 1993; Miroshnichenko et al. 1999). Given the number of free parameters and the fact that the SED is not uniquely defined, all the fits are equally good (or equally bad).
The major difficulty in reproducing the SEDs of HAe stars
with disk models is their
inability to account at the same time for the shape and for the
luminosity of the near-infrared excess, as discussed by
Hartmann et al. (1993). These authors have pointed out that
the high accretion rates required by the large near-infrared
luminosities (Hillenbrand et al. 1995)
are not consistent with the lack of emission
from hot material (dust or gas) which results in the typical
"3 
m bump" shape of the HAe SEDs.
They suggested that the near-infrared emission could be
due
to transiently heated grains and PAHs in the immediate
surroundings of the stars. This, however, was not confirmed
by theoretical models (Natta & Krügel 1995) nor by
ISO spectra of HAe stars (van den Ancker 1999).
The aim of this paper is to rediscuss the disk hypothesis in HAe stars taking advantage of the new and much improved SEDs that have recently become available over a large range of wavelengths. We have selected a small sample of four stars, which includes three highly variable stars, belonging to the subgroup of so-called UX Ori-type HAe stars or UXORs (UX Ori, WW Vul and CQ Tau) and one HAe star with much smaller photometric variability (AB Aur). The properties of UXORs have been summarized in a number of papers (see, for example, Grinin 1994). Most of them can be accounted for by models where a flared, optically thick circumstellar disk seen almost edge-on surrounds the star (Natta & Whitney 2000). Occasionally, the star is occulted by a dust condensation along the line of sight, of unknown origin, which causes deep photometric minima in the visual. It is likely that strong variability is seen only under favorable orientation with respect to the observer.
Our sample was chosen to allow a discussion of the SED in terms of circumstellar disks in objects where the existence and some of the disk properties are independently known. Secondly, we hoped to investigate further the evolutionary stage of UXORs and to shed some light on the location of the dust inhomogeneities that cause the photometric variability of the stars.
The plan of the paper is as follows. In Sect. 2
we describe new interferometric observations
of WW Vul, as well as unpublished ISO
data for WW Vul and CQ Tau and new ground-based 10 m spectra of CQ Tau
and UX Ori.
In Sect. 3 we
present the results of these new observations and the SEDs of the
four stars.
Section 4 is dedicated to a detailed comparison of the SEDs with the predictions
of disk models. In Sect. 5 we discuss some of the implications of the
proposed disk structure and
summarize our conclusions.
Copyright ESO 2001