Continuum imaging at 2.37 GHz was carried out at the ATCA on 1993 September 5. The east-west array and maximum baseline of 6 km gave a nearly circular beam of approximately 3.8 arcseconds. A bandwidth of 128 MHz, 32 
4 MHz, was used, the middle 50 percent retained for imaging. The data were initially edited in AIPS. A full polarimetric calibration was then made in multi-frequency synthesis mode with the MIRIAD imaging software. The ATCA primary calibrator, PKS 1934-638, was used throughout the observation. PKS 1718-649 itself is listed as a secondary calibrator for the ATCA. It appears that PKS 1718-649 is an excellent calibrator since it appears point-like and has a low fractional polarisation (Figure 7.4). No extended structure was observed at a dynamic range of 1000:1, or 4.6 mJy/beam. The total flux density in the image was 4.6 Jy and the fractional polarisation of the E vector approximately 0.35%. The polarisation position angle is -6 
.
Figure: ATCA image of PKS 1718-649 from 1993 Sep 5. Map intensity peak, 4.60 Jy/beam. Contours, -0.1, 0.1, 0.25, 0.5, 1, 2, 4, 8, 16, 32, 50, 70, 95% of peak. Beam FWHM, 3.''8. Map polarisation peak, 0.017 Jy/beam. Polarisation scale, 0.0022 Jy/beam/mm. The map centre is RA = 17 23 41.270, DEC = -65 00 36.80; J2000.
Figure 7.5 shows the radio spectrum of PKS 1718-649 between 408 MHz and 230 GHz (filled triangles). The data at 1.4, 2.3, 4.8 and 8.4 GHz are from co-eval observations with the ATCA in 1994 (L. Kedziora-Chudzer et al. 1995, private communication). The 408 MHz measurement comes from previously published work [Tzioumis 1987]. The 843 MHz datum was obtained with the Molonglo Observatory Synthesis Telescope (MOST) in 1994 (B. Gaensler et al. 1995, private communication) and the 22 GHz measurement from the 70 m Deep Space Network telescope at Tidbinbilla in 1996.
For the 22 GHz observations, an automated control program measured the on and off source system temperatures and calculated the ratio of source temperature to noise diode temperature for two sources, PKS 1718-649 (the target) and PKS 1934-638 (a calibrator). Both sources were observed in good weather and near the meridian. The 22 GHz flux density of PKS 1718-649 was arrived at by extrapolating the PKS 1934-638 spectrum [Tzioumis et al. 1989] to 22 GHz and using that value to calibrate the source to diode ratio. Correction was made for the small difference in elevation through a gain-elevation curve for the 70 m telescope at 22 GHz (K. Fujisawa 1996, private communication). No beam dilution corrections were required since both sources are point-like.
The 90 and 230 GHz flux density measurements of PKS 1718-649, shown in Figure 7.5, were made at the Swedish ESO Sub-millimetre Telescope (SEST) in Chille (M. Tornikowski 1996, private communication) on 1996 February 26.
In addition to the total flux densities in Figure 7.5, the flux densities of the VLBI source at 2.3, 4.8, and 8.4 GHz are also plotted, with 10% errorbars (see chapter 2). At each of these frequencies the VLBI flux density can account for the entire flux density of the source. This supports the ATCA result that the radio source has no extended structure on the scale of arcseconds. For comparison, the radio spectrum of PKS 1934-638 is shown in Figure 7.5 (filled squares).
Figure: Radio spectrum of PKS 1718-649 (filled triangles) compared to the radio spectrum of PKS 1934-638 (filled boxes). Stars with errorbars are the VLBI fluxes for PKS 1718-649.
Only limited data is available on the variability of the total flux density of PKS 1718-649. No significant flux density variability is apparent at 843 MHz (D. Campbell-Wilson 1996, private communication).