Table 1 gives a journal of our VLA observations. The observations in the A configuration were designed to locate any compact radio components in the field. Those in the B, C, and D configurations were intended to image the largest scale emission in enough detail to reveal any relationships and connections between the extended components, as well as to determine their spectral and Faraday rotation/depolarization properties. The BnC configuration data were designed as a sensitive search for connections, such as jets, between the central radio source and the extended features.
Table 1: VLA Observing Log
| Center Frequencies|
| Integration |
|1995 Jul 08||8415, 8465||50||6||5|
|A||1995 Jul 08||4835, 4885||50||6||6|
|B||1995 Dec 23||1365, 1636||12.5||3||13|
|C||1996 Feb 15||1365, 1636||25||3||22|
|D||1996 Sep 02||1365, 1636||25||3||18|
|B||1995 Dec 23||327.5, 333||3.1||1||69|
|C||1996 Feb 15||327.5, 333||3.1||1||30|
|D||1996 Sep 02||327.5, 333||3.1||1||7.5|
|BnC||1997 Jun 17||1365, 1435||50||3||185|
The flux density calibration was based on 3C48 and 3C286. The on-axis instrumental polarization corrections were determined from observations of the unresolved synthesis phase calibrator 2005+778, and the absolute polarization position angle scale from observations of 3C286. Multiple observations of 3C286 and other polarized sources were used to detect problems with ionospheric Faraday rotation, but none was noted in any of the sessions. The data were calibrated using the source 2005+778 as an intermediate phase reference, then self-calibrated using AIPS software developed by W. D. Cotton for the NVSS survey.
Due to the large size of this source, 1.4 and 1.6 GHz observations used three pointings; one on the central source C, and one near the center of each putative lobe. The B, C, and D VLA configuration observations were made at 1.365 and 1.636 GHz to measure rotation measure and spectral index. The data from these frequencies were calibrated and imaged separately. Data taken in the BnC configuration were in two adjacent 50 MHz bands centered on 1.4 GHz. Since the source extent is comparable to that of the antenna pattern and the bandwidth used was relatively large, the deconvolution (CLEAN) and self calibration applied corrections for the frequency dependence of the antenna pattern. Data from each of the three pointings were imaged independently and combined into a single image by interpolating the images onto a common grid, averaging weightings by the square of the antenna power pattern, and correcting for the effects of the antenna pattern. The 0.3 GHz observations were of limited use owing to interference.