The Red Rectangle exhibits a very strong display of the UIR bands and has turned out to be a remarkable source possessing very strong Extended Red Emission (ERE) as well as some of the diffuse interstellar bands in emission. The new observations were made both on the central star HD 44179 and well off star along the NW axis of the biconical nebula. In contrast to the unidentified optical emission bands which show dramatic changes in width and wavelength with offset from the central star, we find that the 3.3-µm feature shows a width of ca. 37-cm^-1 that does not change between the value on star and at 10" offset, and a change in peak wavelength to shorter wavelength of only ca. 2% between the on-star measurement and our highest offset of 15". The Lorentzian width is most easily interpreted as reflecting the lifetime of the emitting states which for the on-star data corresponds to 140 femtoseconds.

Unlike previous data largely taken with ISO and also fitted by Lorentzians, the 3.3-µm band in the Red Rectangle is of the rarer Type 2 and retains this classification for spectra both on-star and along the bicone axis. The surprising stability of the peak wavelength and the almost perfect Lorentzian shape are more reminiscent of the results expected for a single chemical carrier than that from a superposition of differently sized PAH carriers in a range of possible charge states. Could there really be such a carrier in the interstellar medium? This remains a major puzzle in the interpretation of the spectra and one which we are now seeking to address through modelling of the emission process.

Observing Accretion and Outflow from Protostars on Solar-System Scales.

Chris Davis

Accretion and outflow in star formation

Although much is known about Herbig-Haro (HH) jets on large (parsec) scales, their structure within a few arcseconds (<1000 AU) of the central protostellar engine is only now being explored. Optical HST imaging and high-resolution spectroscopy reveal emission at the base of HH jets from T Tauri (or Class II) protostars. Similar regions have recently been observed at near-IR wavelengths, towards the much younger, more deeply embedded, Class I sources. By adopting spectro-astrometric techniques more typically used to analyse optical data we are now able to probe mass accretion and outflow processes on tens-of-AU scales towards heavily obscured sources. Indeed, extending these studies to sources younger than the T Tauri stars is of considerable importance, since more of the mass accreted by a protostar is done so in the formative "Class I" era.

With our new CGS4 echelle data we find that Br-gamma and H₂ 1-0S(1) emissions are powerful tracers of the orthogonal processes of infall and outflow respectively. The BrGamma profiles observed in Class I protostars are typical of permitted hydrogen line profiles in T Tauri stars (see, for example, the article by Daniel Folha in issue 7 of this Newsletter); the profiles from the Class I stars therefore probably derive from the same magnetospheric accretion flows (i.e. from hot streams of gas that flow along "magnetic funnels" from the accretion disk onto the central protostar).

The H₂ spectra, on the other hand, trace the gas dynamics at the base of the outflow (Fig.1). We refer to these H₂ emission regions as ``Molecular Hydrogen Emission-Line'' regions, or MHELs, since their properties compare precisely to those of Forbidden Emission-Line regions (FELs) observed in T Tauri stars. Like the FELs, both low (5-20 km/s) and high (50-150 km/s) velocity components (LVCs and HVCs) are observed in H₂. LVCs are more common than HVCs in MHEL regions, and like their FEL counterparts, the latter are spatially further offset from the exciting source in each case (spectro-astrometric techniques allow us to measure spatial offsets on sub-arcsecond scales, that is, on scales of a few tens of AU; see Davis et al. 2001, MNRAS, 326, 524, for details).

Subsequent FP imaging with UFTI has revealed the extended regions of some of the MHEL features so-far discovered (Fig. 2). The 1"-5"--long extensions observed are distinct from the well-known HH knots and bow shocks seen on larger-scales further downwind, so we refer to them as ``microjets''. In reality, however, the H₂ must be associated with the base of a parsec-scale molecular outflow in each case.

Origin of the MHEL regions

So it seems that H₂ is accelerated and excited into emission at the base of many outflows from Class I protostars. This is an important discovery, since H₂ is not the most robust tracer of outflow or indeed accretion activity (it is collisionally dissociated in high-excitation shock regions and photo-dissociated in PDRs; presumably we have both at the base of an outflow!).

Although the origin of the MHEL and microjet regions is not immediately obvious, the survival and excitation of H₂ must at least constrain the conditions in the MHEL/microjet regions. It is unlikely that H₂ has time to reform in the flow as it accelerates away from the disk surface, and the line intensities observed are too strong to be explained in terms of formation pumping, so the observed H₂ must be somehow accelerated (while remaining intact) off the disk surface in each MHEL/microjet region.

Could FUV continuum photons from the central protostar, or Ly-alpha photons from hot shocked gas associated with accretion flows, excite H₂ at the base of each jet? FUV photons from the central protostar will penetrate along the jet beam further than they do in the orthogonal disk plane because of the lower gas density along the polar jet axis; fluorescent excitation is possible out to a distance of about 1 Av. However, to produce the strong 1-0S(1) intensities observed a strong local FUV radiation field and a high gas density are needed. UV continuum luminosities for T Tauri stars are typically 2-3 orders of magnitude below the values required.

Instead, shock-excitation seems a more likely scenario. The H₂ emission could be produced in internal shocks, resulting from episodic ejections or flow variability. Rapidly varying jet velocities on time scales of the order of 1-10 years would then be needed. Alternatively, some other gas heating and/or compression mechanism, associated with the collimation and acceleration of HH jets near the source, could be responsible. Clearly, H₂ survival and excitation should be part of any new outflow generation model.

Sakurai's Object - A Star Rises from the Dead

Tom Geballe

All astronomers, professional and amateur, are familiar with the story. An isolated star is born; it exists on the main sequence by converting hydrogen to helium in its core. Then after the hydrogen is gone, contraction and heating of the core, hydrogen shell burning, and helium shell flashes cause the star to swell up and become a red giant, and to eject its outer layers, briefly becoming a planetary nebula ionized by the hot stellar core, which is now becoming a white dwarf. After the nebula is dispersed what remains of the star ends its existence, not as star but as a white dwarf, no longer able to generate thermonuclear energy.

Stars with masses less than several suns (i.e., nearly all of the stars in the universe) evolve in this way. However there can be wrinkles in the evolutionary sequence, and it is possible for a dying star to live again, albeit briefly on an astronomical time scale. Such is the case with Sakurai's Object, found in 1995 by a Japanese amateur astronomer. Originally suspected of being a nova, it is now clear that the object already had reached the white dwarf phase when it underwent its final helium-shell flash.

The evidence for this includes a variety of optical and infrared spectroscopy, showing that the object was cool (thus not a nova) after the outburst occured and was very hydrogen deficient. It also includes deep optical images which revealed a surrounding planetary nebula continuing to glow faintly, despite having at present no source of ionizing photons (thus demonstrating that the central source previously had been very hot). The only other star definitely known to have undergone this transformation is Nova Aquilae 1919 (now known as V605 Aquilae), long before the advent of modern astronomical instrumentation.

UKIRT has made important contributions to the study of Sakurai's Object. A research team comprised of Prof. Nye Evans, Vic Tyne, and Barry Smalley (all of Keele University), Stewart Eyres of the University of Central Lancashire, and I have been undertaking a long term observing program, which has produced many excellent 1-5um CGS4 spectra chronicling the rapid changes in the object since its outburst. An overview of the 1996-2000 spectra in the 1.0-2.5-µm region is shown in Fig. 1. The most obvious changes these spectra show are: (1) the cooling of the photosphere and the formation of molecules there during 1996-1997; (2) the formation of dust and ejection of dust and gas in 1998 and the following years; and (3) the complete obliteration by the dust envelope of our view of the central star in 1999 and 2000. In 2001 the central star remains completely obscured. We also have made detailed measurements at a variety of resolutions of portions of the 1-5um spectrum, most notably of the He I 1.083-µm line. This line first emerged in 1998 with a P Cygni profile, but has been a pure emission line since then. We believe that the helium line emission is collisionally excited in the wind from Sakurai's Object, as newly formed dust accelerates outward by radiation pressure, sweeping up surrounding helium-rich gas. Velocities as high as 500 km/s have been seen in this line, which is now fading out, although still detectable this year.

Our group has already published three refereed papers based on our UKIRT observations, and more are on the way. Last year UKIRT data figured prominently in an international conference at Keele entirely devoted to this remarkable object. While we hope one day to catch a glimpse of the central "star" again, we note that as the dust cools a greater fraction of the radiation is being emitted in the 10-µm and 20-µm windows. As far as Sakurai's Object is concerned, the arrival of Michelle at UKIRT could not be better timed. PATT willing, we look forward to studying Sakurai's Object at UKIRT for many years to come.