One of the important programs in the extragalactic arena for WUPPE was to verify the thick torus model for Seyfert nuclei proposed by Miller and Antonucci. Based on polarization studies they proposed that the only difference between type 1 and type 2 Seyferts was a geometrical one. The non-thermal engine in the nucleus of a type 2 Seyfert is hidden from direct view by a thick disk, whereas the type 1 Seyfert nuclei can be seen directly. The observational basis for this model was measurements of the polarization of NGC1068. The measured polarization increased from 1.9% at 6500A to 8.0% at 3700A. They corrected the polarization for background dilution using a spectrum of M32 to represent the stellar contamination. The corrected polarization was wavelength independent at a value of 16%. In this model the polarized flux is light from the nucleus being scattered off of unobscured electrons surrounding the nucleus. Was the choice of M32 to represent the stellar background valid? Was the fraction of flux from M32 based on matching line strengths appropriate? Was it safe to assume that the internal reddening of the two galaxies is the same? What effect does galactic I.S. polarization have? Is it possible that the scattering is due to solid grains? Ultraviolet spectropolarimetry provides a crucial test. The stellar background is insignificant, i.e. no uncertain background correction is necessary and the discrimination between electron scattering and other processes becomes very large. WUPPE found a constant polarization of approximately 15% for NGC1068 in the spectral interval from 1500A to 3200A. The analysis of Miller and Antonucci was confirmed. Other conclusions could be drawn from the high signal-to-noise ratio spectra. Among these are that no sharp emission line shows in the polarized flux, which is consistent with the model. Other interesting results follow from a comparison between NGC1068 and the type 1 Seyfert NGC4151.
It is important to recognize that WUPPE was designed to measure all four Stokes parameters, I, Q, U, and V, and that I is a high resolution spectrum. Anything that can be done with a low resolution UV spectrograph such as IUE can be done with WUPPE data as well. One of the puzzles about the spectra of galaxies in the ultraviolet is the absence of a 2200A absorption feature. The 2200A bump is observed in individual star spectra in the LMC and SMC, although it is weaker than the average galactic extinction. There have not been any cnvincing observations of this feature in the integrated light of normal galaxies even though there is often marked visual obscuration and far infrared flux suggesting absorption of stellar radiation. While the contributions to the integrated light are complex, the 2200A feature is believed to have a low albedo and was expected to show in some galaxy spectra. It is unlikely that our galaxy and the Magellanic Clouds are unique, which is why it is a puzzle. IUE spectra of some QSO's do show a depression near 2200A in their rest frame. Summing many spectra yields a high signal-to-noise spectrum which shows this UV dip to be similar in shape to the 2200A bump, but it is generally believed that this is an artifact of strong FeII emission lines (although fitting the profile has proven difficult). One of the reasons that the UV dip is not believed to be interstellar is that the far UV extinction would then be too high if a standard extintion curve is adopted. The high signal-to-noise ratio spectra that WUPPE obtained do not show the bump in normal galaxies, but the AGN MKN335 does show a feature centered at the proper wavelength in the rest frame of the galaxy. The band shape is an excellent fit to an interstellar 2200A bump profile and the galactic extinction in the line of sight is too low to distort the profile. Moreover if the albedo of the non-bump particles is reasonably high, multiple scattering radiative transfer can account for the observed bump without significantly reducing the far UV flux. It is an important question because it bears on the heating of the dust which is important for star formation considerations and for cold dust mass considerations.
- A. D. Code