Introduction

In the August 1992 JCMT Newsletter, we reported on our 800 micron polarimetric observations of Rho Oph A. These observations demonstrated the feasibility of mapping submillimetre polarisation across regions containing young stellar objects (YSOs), and showed that the magnetic field in the direction of the brightest peak of the Rho Oph A core, known as SM1 (Ward-Thompson et al. 1989), is aligned roughly constantly in the NE-SW direction. Also reported were our observations of W3-IRS5, a high-mass YSO, where we found evidence for a 'pinched-in' magnetic field. These combined results support models of bimodal star formation, as discussed by Greaves, Murray and Holland (1994).

We returned to JCMT in June 1994 to make further polarimetric observations of star formation regions. Observations were carried out of VLA1623, a very young protostellar candidate in Rho Oph A, to test whether the magnetic field direction alters in the vicinity of this object, compared to that in the rest of the cloud core. We also observed a double core system in S106, comprising a massive YSO (S106 IR) and a submillimetre source which may be another protostar candidate (S106 FIR), to investigate if their magnetic fields were related.

Observations

The observations were made in June 1994, using the Aberdeen/QMW polarimeter, in conjunction with the common-user photometer UKT14. We observed at 800 microns, where the instrumental polarisation is minimum (0.5%) and the spatial resolution is high (14" beam). A 16-position waveplate cycle was used, to give 4 independent estimates of percentage polarisation and position angle in a 10 minute integration, and the data were analysed using Ramon Nartallo's SIT software.

The inferred magnetic field directions are perpendicular to the polarisation vectors. This is because the submillimetre polarisation, which arises from thermal emission by aligned non-spherical dust grains, is predominantly along the long axes of the grains. The exact mechanism of alignment is still a matter of some debate, but for all mechanisms the long axes, and hence the polarisation, are perpendicular to the direction of the magnetic field.

Figure 1. Polarisation map of SM1/VLA1623 at 800 microns. The direction of polarisation appears to be roughly constant across the whole region, inferring a magnetic field direction running roughly NE-SW through the cloud, even at the position of VLA1623, the southernmost polarisation vector.

Low-mass star formation: the case of VLA1623

The Rho Oph A core is a region of low-mass star formation, which has recently been shown to contain one of the youngest protostar candidates yet discovered, the prototype Class 0 source VLA1623 (Andre, Ward-Thompson and Barsony, 1993 - hereafter AWB). This source has an extended and extremely highly-collimated bipolar outflow aligned roughly NW-SE (Andre et al. 1990). Recent OVRO interferometer measurements, however, show no evidence for a circumstellar disk around VLA1623 (Andre, priv. comm.), such as was found around HL Tau (Keene & Masson 1990). The question therefore arises as to what is collimating the bipolar outflow. One possible answer to this question is that the large-scale magnetic field of the cloud is responsible for the outflow collimation. In this scenario the large-scale magnetic field in the molecular cloud should lie parallel to the direction of the bipolar outflow.

The results for Rho Oph A are shown in Figure 1. The three most southerly points are those obtained in the recent run. As the figure shows, the magnetic field deduced from the polarisation vectors is roughly parallel throughout the cloud core, lying roughly NE-SW - ie: perpendicular to the direction of the bipolar outflow. There is no evidence for any change in direction of the magnetic field from SM1 to VLA1623, where the field is aligned with the previous points, and also lies perpendicular to the outflow.

The case of VLA1623 therefore remains an enigma. Our results have shown that the large-scale field cannot be collimating the outflow. There is no apparent disk, and AWB have suggested a 'cored-apple' structure through which the outflow emerges. Further modelling is required to see if this structure, threaded by a toroidal or planar magnetic field, can collimate the outflowing gas.

High-mass star formation: the case of S106

The S106 HII region has a bipolar morphology, aligned roughly N-S, bisected by a lane of obscuration running approximately E-W. At the centre of the system lies the near-infrared source S106-IR . The dark lane was originally hypothesised to be a circumstellar disk. However, recent results (Richer et al. 1993) showed that, when observed with the high resolution of the JCMT at 450 microns, the supposed disk breaks up into a number of fragments, the brightest of which, S106-FIR, is a candidate protostar. We therefore

obtained submillimetre polarimetry of S106-IR and S106-FIR to ascertain the magnetic field direction in these two sources, and to look for any correlation between the two.

Figure 2. Polarisation map of S106IR/S106FIR at 800 microns. The direction of polarisation of the two sources infers a magnetic field direction parallel to the broad lane of 800-micron emission.

Figure 2 shows an 800-micron isophotal contour map of S106, with the two submillimetre polarisation measurements at the positions of S106-IR and S106-FIR superposed. The dark lane of optical obscuration appears as an approximately E-W band of emission at 800 microns. It can be seen that the two sources are both polarised in a similar direction, and that the inferred magnetic field (perpendicular to the polarisation vectors) lies roughly along the lane of 800-micron emission. The alignment of the field directions in the two young objects may have interesting implications for our understanding of the formation of binary star systems.

Recent Zeeman observations (Roberts et al. 1994) detected a strong magnetic field in the lobes of the S106 HII region, with a much reduced field in the vicinity of the dark lane. This was interpreted as a large-scale magnetic field, aligned N-S, parallel to the alignment of the bipolar HII region, which is pinched into an hour-glass shape in the vicinity of the central star. Our observations appear to contradict this scenario, by showing that the field is lying roughly E-W in the vicinity of the star.

We therefore suggest a hypothesis to explain both our data, and the Zeeman data as follows: The large-scale magnetic field lies N-S along the bipolar HII region, but close to the star the field may be twisted into a toroidal morphology around the central star. The resultant field lies parallel to the dust lane in the small JCMT beam, but diverges back to the large-scale N-S field on larger scales. At the centre, the two competing field directions are seen simultaneously in the larger beam Zeeman observations, causing an apparent reduction in the observed magnitude of the magnetic field.

References:

Andre et al., 1990, A&A 236, 180.

Andre, Ward-Thompson and Barsony, 1993, ApJ 406, 122.

Greaves, Murray and Holland, 1994, A&A 284, L19.

Keene and Masson, 1990, ApJ 355, 635.

Richer et al., 1993, MNRAS 262, 839.

Roberts et al., 1994, ApJ in press.

Ward-Thompson et al., 1989, MNRAS 241, 119.

W.S. Holland, J.S. Greaves, JAC

D. Ward-Thompson, ROE