3. Results and discussion
From a comparison between their appearance in the broad-band (ESO/SERC) images and the corresponding narrow-band images it becomes obvious that both objects are strong emitters in the light of ( + [N II ]) and [O III ]. As a first result the classification as PN can thus be confirmed. Following Acker et al. (1992), the designation for Wei 1-5 will be PN G013.7-15.3 and for KeWe 1 PN G280.5+01.8.
Wei 1-5 shows two pronounced knots with fainter extensions to the NW, Fig. 1. The morphology in both lines is very similar. At low intensity levels the object has an elliptical outline. The extension of the western knot ends in a second intensity maximum that is seen in only. The intensity distribution can best be understood in terms of a ring or torus seen edge-on. A cut through the plane common to both knots reveals the wellknown doubly-peaked signature with a central trough. Fig. 2 also shows that the intensity distribution is virtually identical in [O III ] and ; both have been scaled to the same peak value for comparison of shape. The distance between the knots is smaller in [O III ] by about 15% as can be expected from ionisation stratification. Based on simple geometric considerations a homogeneous torus with a ratio of radius vs. thickness (R/d) of 2 is in good agreement with the observed distribution. More sophisticated models yield generally the same results (e.g. Pascoli 1990 and Figs. 9 - 11 in Mellema 1994). It is quite common for elliptical PNe to have brightness enhancements at the ends of the short axis and for most objects the long axis is perpendicular to the short one. In this case it is tilted by about . The tilted extension of the knots can not be accounted for by the simple torus model. The shaping of these structures may be due to asymmetric expansion caused by the density distribution in the shell or in the surrounding ISM. Spectroscopic investigation is warranted, here.
KeWe 1 is of overall asymmetric appearance at first sight, but closer inspection reveals a number of symmetries (Fig. 3). Comparison of the ESO R film and the image shows almost perfect agreement of the morphology even to small details; in [O III ] comparison is limited by the low S/N ratio of the object on the SERC J. This comparison also helps to distinguish between nebular structure and the possible contribution from faint stars in this rather dense stellar field. For the following discussion we will concentrate on the image (Fig. 3a) because of its high S/N ratio.
KeWe 1 is dominated by two bright cone shaped knots that give it its bipolar appearance, axis (A). In an axis (C) almost perpendicular to axis (A) two fainter knots can be found. This can be interpreted as the signature of an equatorial torus and two polar knots. Examples for such a configuration are He 2-36 and He 2-123. A cut through the equatorial plane (A) shows again the doubly peaked distribution but an extremely deep central trough (Fig. 4). In the intensity in the trough is essentially at the level of the background outside of the object, the central part is practically devoid of emission.
Such a distribution cannot be reconciled with a complete torus. Deep troughs result from large R/d ratios as in thin shells, but this is not compatible with the extension of the cones in this case. Therefore the existence of an incomplete torus or clouds has to be assumed. A more detailed inspection of the nebula gives us a clue of the processes that shape it. Both bright knots are cone shaped, resembling Mach cones with a rarefaction zone in their wake, forming the waist of the nebula. Both polar knots are elongated radially outward from the center of the nebula, the northern knot is particularly interesting because it shows a linear extension resembling a contrail, pointing to the north. This might be the result of wind erosion of a dense knot by a strong stellar wind. Such winds are not uncommon among PNe, velocities ranging from 1000 to 4000 km/s have been reported (Pauldrach et al. 1988; Patriarchi & Perinotto 1991). A dramatic example for such wind erosion can be seen in HST images of the inner part of A 30, see Borkowski et al. (1995).
Other indications for wind action in KeWe 1 can be found. At low intensity levels the object takes on an elliptical contour but the brighter parts form a broken shell. Along a plane orientated almost N-S very little emission can be found forming an empty channel or chimney (B). A similar but less obvious channel exists in the NW-SE direction (D). Both are inclined to the polar axis (C) by 25 to . Axes (A) to (D) intersect each other very close to the center of the nebula. This indicates that the forces shaping these structures have a common origin there, at the central star. Looking at the radial distances to this center it becomes apparent that a high degree of point symmetry can be discerned including not only knots of matter but also regions of very low density.
So while it may not be the only way to explain the variety of features observed, a strong wind from the central star (CS) can be considered a likely mechanism. Unfortunately no CS has been found in KeWe 1 to date, but this is not really surprising. Bipolar nebulae seem to result from relatively massive ( 1.5 ) stars (Peimbert 1978, Corradi & Schwarz 1993, 1995). The resulting CSs have high temperatures. This is consistent with a strong wind, see Pauldrach et al. (1988). Hot CSs are very faint in the optical and extinction towards KeWe 1 is rather substantial, so its CS may well be beyond the limiting magnitude of the SERC. Objects with similar morphology - including point symmetry - exist, for example K 1-10, NGC 2899 or He 2-36. Most of these objects show high velocity features and/or high expansion velocities (Corradi & Schwarz 1995). For very few of them binary CSs have been identified, for others no CS is known, e.g. NGC 2899. A binary star is the most natural explanation for the morphologies observed, see e.g. Livio 1993, Soker & Harpaz 1992. KeWe 1 may be an object developing into a butterfly nebula, its rapid evolution being manifested in the wind shaped features. Spectroscopy to determine the physical properties of the nebula and the velocity field in it, will be performed in the near future.
© European Southern Observatory (ESO) 1997
Online publication: July 8, 1998