astro-ph0003466/ms.tex
1: % paper.tex  RWH 16-jun-99; 21-jun-99; 24-jun-99; 25-Jun-99; 26-Jun-99; 27-jun-99
2: %                28-Jun-99; 2-jul-99; 7-aug-99; 9-aug-99
3: % Revised following ref comments 24-jan-2000; 25-jan-2000; 27-jan-2000; 
4: % 28-jan-2000
5: 
6: % SAMPLE2.TEX -- AASTeX macro package tutorial paper.
7: 
8: %\documentstyle[12pt,aasms4]{article}
9: \documentstyle[11pt,aaspp4]{article}
10: %\documentstyle[aas2pp4]{article}
11: 
12: %\documentstyle[11pt,eqsecnum,aaspp4]{article}
13: 
14: %\documentstyle[12pt,amssym,aasms4]{article}
15: 
16: %\received{4 August 1988}
17: %\accepted{23 September 1988}
18: %\journalid{337}{15 January 1989}
19: %\articleid{11}{14}
20: 
21: \slugcomment{Submitted to ApJ}
22: 
23: \lefthead{Hannikainen et al.}
24: \righthead{GRO J1655--40 Radio Observations}
25: 
26: \def\groj1655{{\rm GRO\,J1655$-$40}}
27: \def\vlai{{\sc vla1}}
28: \def\vlaii{{\sc vla2}}
29: \def\vlaiii{{\sc vla3}}
30: \def\vlbai{{\sc vlba1}}
31: \def\vlbaii{{\sc vlba2}}
32: 
33: 
34: \begin{document}
35: 
36: \title{Radio Emission from GRO\,J1655--40 during the 1994 Jet Ejection Episodes}
37: 
38: \author{D. C. Hannikainen}
39: \affil{Observatory, PO Box 14, FIN-00014 University of Helsinki,
40:          Finland}
41: 
42: \author{R. W. Hunstead and D. Campbell-Wilson}
43: \affil{School of Physics, University of Sydney, NSW 2006, Australia}
44: 
45: \author{K. Wu}
46: \affil{Special Research Centre for Theoretical Astrophysics, School of Physics,
47:              University of Sydney, NSW 2006, Australia}
48: 
49: \author{D. J. McKay\altaffilmark{1,2}}
50: \affil{University of Manchester, Nuffield Radio Astronomy Laboratories,
51:          Jodrell Bank, \\
52: Cheshire SK11~9DL, United Kingdom}
53: 
54: \author{D. P. Smits}
55: \affil{Department of Mathematics, Applied Mathematics \& Astronomy,
56: University of South Africa, PO~Box 392, Pretoria 0003, South Africa}
57: 
58: \and
59: 
60: \author{R. J. Sault}
61: \affil{Australia Telescope National Facility, CSIRO, PO Box 76, Epping,
62:               NSW 1710, Australia}
63: 
64: \altaffiltext{1}{Joint Institute for VLBI in Europe, Postbus 2, 
65:            7990 AA Dwingeloo, The Netherlands}
66: \altaffiltext{2}{Australia Telescope National Facility, 
67:   CSIRO --- Paul Wild Observatory, 
68:   Locked Bag 194,
69:   Narrabri, NSW 2390,
70:   Australia}
71: 
72: \begin{abstract}
73: 
74: 
75: We report multifrequency radio observations of \groj1655 obtained with
76: the Australia Telescope Compact Array, the Molonglo Observatory
77: Synthesis Telescope and the Hartebeesthoek Radio Astronomy Observatory
78: at the time of the major hard X-ray and radio outbursts in
79: 1994 August-September.  The radio emission reached levels of the order
80: of a few Jy and was found to be linearly polarized by up to 10\%,
81: indicating a synchrotron origin.  The light curves are in good
82: agreement with those measured with the VLA, but our closer time
83: sampling has revealed two new short-lived events and significant
84: deviations from a simple exponential decay.  The polarization data
85: show that the magnetic field is well ordered and aligned at
86: right angles to the radio jets for most of the monitoring period.  The
87: time evolution of the polarization cannot be explained solely in
88: terms of a simple synchrotron bubble model, and we invoke a
89: hybrid `core-lobe' model with a core which contributes both
90: synchrotron and free-free emission and `lobes' which are classical 
91: synchrotron emitters.
92: 
93: \end{abstract}
94: 
95: \keywords{polarization --- stars: individual (\groj1655) --- radio
96: continuum: stars}
97: 
98: \section{Introduction}\label{intro}
99: 
100: The X-ray transient \groj1655 (Nova Sco 1994) was first detected with
101: the {\em Burst and Transient Source Experiment} (BATSE) on board the
102: Compton Gamma-Ray Observatory on 1994 July 27 (\cite{nan94}).
103: Significant flux was observed up to 200 keV and the source reached a
104: luminosity of 1.1 Crab (20--100 keV) on Aug 1. \groj1655 remained in
105: outburst until about Aug 15, and after a period of quiescence flared
106: again on Sept 6 (\cite{harmon95}).  The detection of a steeply-rising
107: radio counterpart was first reported by Campbell-Wilson \& Hunstead
108: (1994a) following observations with the Molonglo Observatory Synthesis
109: Telescope (MOST) at 843~MHz on 1994 Aug 6 and 11.  The flux density
110: continued to increase, reaching 4.2 Jy on 1994 Aug 14 and 5.5 Jy on
111: 1994 Aug 15 (\cite{cw&h94b}).  High resolution radio observations
112: using the Very Large Array (VLA), the Very Long Baseline Array (VLBA)
113: (Hjellming \& Rupen 1995, hereafter HR95) and the Southern Hemisphere
114: VLBI Experiment (SHEVE) array (\cite{tin95}) showed repeated episodes
115: of relativistic ejections. Three major ejection events on TJD~9577.5,
116: 9584 and 9596 (TJD defined as JD $-$ 2440000.5) were observed with the 
117: VLA and three on TJD~9574$\pm$1,
118: 9605$\pm$2 and 9668$\pm$5 were recorded with the VLBA (\cite{h&r95}).
119: The ejection velocity, after correction for inclination to the line of
120: sight, was inferred to be $v\sim0.92c$ in the rest frame of \groj1655,
121: at a position angle of $47^{\circ}\pm1^{\circ}$ (\cite{h&r95});
122: wiggles about the jet axis were interpreted as precession with a
123: period of $3.0\pm0.2$d.  Photometric observations showed \groj1655
124: to be a high-inclination binary system (\cite{bailyn95}). Subsequent 
125: optical observations led to the determination of a precise mass
126: for the primary, $M_1=7.02\pm0.22\,M_{\odot}$ (\cite{o&b97}), which is
127: well above the theoretical upper limit for a neutron star and direct
128: evidence for it being a black hole. In addition, Soria et al.\ (1998)
129: found a 95\% confidence limit of $M_1\,>\,5.1\,M_{\odot}$ for the mass of
130: the primary based on measurements of velocity variations in
131: the He~{\sc ii} disk emission lines, thought to reflect the orbital
132: motion of the primary, and hence confirming \groj1655 to be a
133: black hole binary. The mass of the secondary, which is classified as
134: {\rm F}3 {\rm IV}--{\rm F}6 {\rm IV}, is $M_2=2.34\pm0.12\,M_{\odot}$,
135: and the spectroscopic period of the binary is
136: $P=2^d.62157\pm0^d.00015$ (\cite{o&b97}).
137:  
138: During the 1994 ejection events, the radio outbursts from \groj1655
139: were monitored in the southern hemisphere by the Molonglo Observatory
140: Synthesis Telescope (MOST), the Australia Telescope Compact Array
141: (ATCA), and the Hartebeesthoek Radio Astronomy Observatory (HartRAO).
142: We discuss below the results of these observations.  Sections
143: \ref{obs+dr} and \ref{lightcurves} describe the MOST, ATCA and HartRAO
144: observations and the resulting light curves.  In Section \ref{linpol}
145: we present the linear polarization data from the ATCA, and we discuss
146: the evolution of the radio spectra in Section \ref{spectralstuff}.  We
147: briefly compare our results with those of HR95 in Section
148: \ref{compare}.  Finally, we interpret the polarization data in terms
149: of a `core-lobe' model in Section \ref{physint}.
150: 
151: \section{Observations and data reduction}\label{obs+dr}
152: 
153: \subsection{MOST}
154: 
155: The MOST is an east-west synthesis array comprising two colinear
156: cylindrical paraboloids each 11.6~m wide by 778~m long, separated by
157: a 15~m gap (\cite{mills81}; \cite{rob91}).  The telescope operates at
158: 843~MHz, with a detection bandwidth of 3.25~MHz, and measures right
159: circular polarization (as defined by the IEEE standard). The
160: synthesized beamwidth is $43\arcsec\times43\arcsec {\rm cosec} |\delta|$
161: FWHM (RA $\times$ Dec).  The telescope forms a comb of 64 real-time
162: fan beams spaced by $22\arcsec$; for the observations of \groj1655 the
163: pointing of the beam set was time shared among three adjoining
164: positions to give a field size of $70\arcmin\times70\arcmin$
165: cosec$|\delta|$.  The background noise in a full 12-hour synthesis
166: image was typically 1.5 mJy rms.
167: 
168: Many \groj1655 observations were partial syntheses as the observations
169: were fitted in around a previously scheduled observing program. The
170: choice of field center serendipitously included a strong point source
171: ($\approx$ 900 mJy, J2000 position $16^{\rm h} 53^{\rm m} 55.9^{\rm
172: s}, -40^{\arcdeg} 37^{\arcmin} 23^{\arcsec}$) that has subsequently
173: been used as an internal reference source.  The flux density
174: calibration was based on short fan-beam {\sc scan} observations of
175: strong point sources from the list of Campbell-Wilson and Hunstead
176: (1994c).  For S$_{\rm 843\,MHz}>100$~mJy the dominant source of error
177: in a synthesis image is the scatter in the pre- and post-observation
178: calibrators; a conservative error of $\pm 3$\% has been assigned. At
179: lower flux densities the error becomes dominated by noise and, for
180: partial syntheses, residuals arising from incomplete cancellation of
181: strong out-of-field sources.  This difficulty can be partly overcome
182: by subtracting a reference image (obtained when \groj1655 was
183: quiescent) with the same hour angle coverage.
184: 
185: The data processing used standard MOST reduction software.  When the
186: source was bright we were able to search for sample-to-sample (24~s)
187: variations in flux density.  Such measurements are affected by
188: confusing sources in MOST's fan beams, by the effects of ionospheric
189: refraction and by slow drifts in telescope calibration.  These effects
190: are eliminated or greatly reduced in a full 12-hour synthesis.  For
191: this paper, individual sample flux densities were fitted and
192: averaged in blocks of 200 (80 minutes) and then combined in 4--6-hour
193: blocks for tabulation.  Full synthesis imaging was used when the
194: source flux density fell below 500~mJy, with flux densities measured
195: using IMFIT in {\sc aips}.
196: 
197: In addition to the synthesis-mode observations of \groj1655, spot
198: measurements of flux density were obtained using 4-minute {\sc scan}
199: observations, bracketed by {\sc scan}s of calibrators.  This mode of
200: operation measures the target and calibrator source flux densities
201: using the (same) central beams of the 64-beam block. It was used only
202: when the flux density of \groj1655 was greater than 1~Jy, with the fit
203: parameters serving as a guide to the presence of confusing sources in
204: the fan-beam response.  The final light curve, tabulated in
205: Table~\ref{tbl-most}, contains a mix of flux densities measured from
206: synthesis images, from averaged fits to the synthesis data, and from
207: {\sc scan} observations, and includes data from a third, weaker
208: outburst which occurred in 1994 Nov (\cite{w&h97}; \cite{hann}).
209: 
210: \placetable{tbl-most}
211: 
212: 
213: \subsection{ATCA}
214: 
215: Following the report of the intense radio outburst by the MOST,
216: \groj1655 was monitored as a target-of-opportunity with the 
217: ATCA for a period of about twenty days.
218: VLBI observations and \ion{H}{1} spectral line observations were also
219: %%H~{\sc i} 
220: undertaken (\cite{m&k94}; \cite{tin95}).
221: 
222: The ATCA is an earth-rotation aperture synthesis array, comprising six
223: 22-m antennas which can be moved along an east-west railway track to
224: give baselines up to 6~km (\cite{frater}). 
225: The observations of \groj1655 were made at central frequencies of
226: 1.380, 2.378, 4.800, 5.900, 8.640 and 9.200~GHz with 128~MHz bandwidth
227: in two orthogonal linear polarizations. Each observation was typically
228: 10 minutes duration, with the antenna gain and phase calibration
229: derived from regular observations of the point-source calibrator
230: PKS~B1740$-$517.  The flux density scale was tied to the primary
231: calibrator PKS~B1934$-$638, which we assume has flux densities of
232: 14.96, 11.54, 5.83 and 2.84~Jy at 20, 13, 6 and 3~cm (1.4, 2.3, 4.8,
233: 8.6~GHz) respectively (\cite{rey94}).  The 6.0A configuration was
234: used throughout, giving interferometer spacings from 337--5939~m.
235: 
236: The data were reduced using standard flagging and calibration
237: techniques in the {\sc miriad} package (\cite{bob95}). The
238: polarimetric calibration of ATCA data is described in \cite{bob91}.
239: Polarimetric leakages, which are quite stable with time on the ATCA,
240: were determined simultaneously with the antenna gains using either
241: PKS~B1934$-$638 or PKS~B1740$-$517, both known to be unpolarized.
242: The resultant polarimetric purity of the calibrated data is believed
243: to be better than 0.1\%. At all wavelengths, a single round of phase
244: self-calibration was performed on the \groj1655 data (using a point
245: source model) to eliminate residual phase instability. As the
246: observations placed \groj1655 at the phase center, we simply summed
247: the real parts of the relevant Stokes visibilities to determine the
248: source flux density; this is equivalent to natural weighting.  
249: The errors are dominated by systematic effects. 
250: For the total intensity 
251: measurements, the  errors are smaller than 1\% of Stokes~I,
252: while for Stokes~Q and U they are less than 0.1\% of Stokes~I.
253: The ATCA total flux densities are tabulated in Table~\ref{tbl-atca}. 
254: 
255: \placetable{tbl-atca}
256: 
257: \subsection{HartRAO}
258: 
259: Continuum measurements of the flux density of \groj1655 were made at
260: 3.5 and 6.0~cm using the 26-m telescope of the Hartebeesthoek Radio
261: Astronomy Observatory (HartRAO) situated in Gauteng, South Africa.
262: The observations were made using dual-feed systems operated in a
263: beam-switched Dicke mode.  At 3.5~cm (8.58 GHz) the bandwidth was 400
264: MHz, the beamwidth $5.7'$ FWHM and the feed recorded right circular
265: polarization.  The 6~cm (5 GHz) receiver had a bandwidth of 800 MHz, a
266: beamwidth of $10'$ FWHM and the feed was linearly polarized east-west.
267: 
268: The flux density at each wavelength was measured using a five-point
269: stepping sequence in declination to correct for pointing errors in
270: this coordinate; the stepping positions in terms of beam parameters
271: were first null, half-power, on-source, half-power, first null. The
272: five points were fitted to a Gaussian to determine declination
273: pointing errors.  Right ascension errors were measured at 6~cm by
274: observing midway between the beams. Right ascension pointing
275: corrections at 3.5~cm were assumed to be the same as at 6~cm. The beam
276: separation was sufficiently close ($15'$ or 1.5 times the
277: beamwidth at 6~cm) that the response of the dual-beam Dicke system
278: varied linearly with right ascension in this region.  The observed
279: flux density is an average over both beams, corrected for both
280: declination and right ascension pointing errors. Measurements
281: alternated between the two wavelengths with 10 minutes' integration on
282: each.
283: 
284: Gain curve corrections, determined by G.\ Nicolson (private
285: communication), were applied to the data.  Flux densities were
286: calibrated to an arbitrary scale using a noise diode, and the
287: calibrators 3C\,123, 3C\,161, 3C\,218 (Hydra A), 3C\,274 (Virgo A) and
288: 3C\,348 were observed occasionally to put the flux densities on the
289: Ott et al. (1994) scale.  The HartRAO data are tabulated in
290: Table~\ref{tbl-rao}; we note that on some days the scatter among
291: a closely spaced sequence of measurements is somewhat larger than
292: implied by their formal errors.
293: 
294: \placetable{tbl-rao}
295: 
296: 
297: \section{Radio light curves}\label{lightcurves}
298: 
299: \subsection{MOST and HartRAO}\label{most+rao_lc}
300: 
301: \placefigure{fig-most+batse}
302: 
303: The MOST 843~MHz light curve up to the end of 1994 September is
304: plotted in the top panel of Figure~\ref{fig-most+batse}.  The duration
305: of the ATCA observations is also shown, as are the epochs of
306: specific ejection episodes, as reported by \cite{h&r95} from their VLA
307: and VLBA data; we labelled these epochs {\sc vla1}, {\sc vla2}, {\sc
308: vla3} and {\sc vlba1}, {\sc vlba2}, corresponding to TJD 9577.5, 9584,
309: 9596 and TJD~9574, 9605, respectively.  The BATSE 20--100 keV light
310: curve (provided by S.\ N.\ Zhang and W.\ S.\ Paciesas) is plotted in
311: the bottom panel, showing two major hard X-ray outbursts, beginning on
312: TJD~$\sim 9560$ and TJD~$\sim 9600$.
313: 
314: Two radio outbursts were observed with the MOST in 1994
315: August--September, both apparently associated with the hard X-ray
316: bursts detected by BATSE. The MOST data clearly show the rise of the
317: radio flux density lagging behind the rise of the X-ray intensity in
318: the outbursts. During the first X-ray outburst, the radio flux density
319: was still low (S$_{\rm 843\,MHz}$ $\simeq$ 0.36 Jy) on TJD~9570.60,
320: while the X-ray intensity had already reached its maximum. When the
321: X-ray intensity began to decline, the radio flux density then
322: increased rapidly, reaching a local maximum (S$_1$) of 5.45 Jy on
323: TJD~9579.22.  After a brief decline, the flux density rose again and
324: reached a new peak of 7.62 Jy on TJD~9582.13, about two weeks after
325: the X-ray peak.  At this point the X-ray intensity had already fallen
326: to a much lower level, but was still above the background. 
327: The radio peak in the second outburst (S$_2$) also lagged the X-ray
328: peak, but the delay was shorter.  The maximum intensities in the two
329: X-ray bursts were similar, but the second radio burst was obviously
330: weaker.
331: 
332: 
333: \subsubsection{Short Duration Radio Burst Component}\label{short} 
334: 
335: \placefigure{fig-most+rao}
336: 
337: To emphasize some specific features of the temporal evolution of the
338: radio emission, we plot in Figure~\ref{fig-most+rao} (upper panel) the
339: MOST 843~MHz and HartRAO 5~GHz flux densities on a logarithmic scale
340: as a function of time.  Closely spaced groups of 5~GHz points are
341: plotted as weighted averages.  The 843~MHz flux density curve is
342: characterized by short-duration burst components (marked S$_1$ and
343: S$_2$) superimposed on longer-lived components (L$_1$ and L$_2$) which
344: show an approximately exponential decay.  HartRAO did not observe
345: \groj1655 until after the end of S$_1$, but S$_2$ was clearly recorded
346: at 5~GHz.  The confirmation of the short duration burst component is
347: an important new result, since neither the VLA (\cite{h&r95}) nor the
348: ATCA collected data during these periods.  From the MOST data, the
349: short duration components, S$_1$ and S$_2$, have e-folding rise times
350: of $\leq1.1$ days, and S$_2$ decays with an e-folding time of $\sim
351: 2.2$ days.  The rising time scale for the long duration component
352: cannot be determined from the data, but L$_1$ and L$_2$ have e-folding
353: decay times of $\sim 6.5$ days and $\sim 8.7$ days respectively.  A
354: third short duration event was observed by MOST peaking on TJD~$\sim
355: 9666$ which decayed with an e-folding time of $\sim 2.6$ days
356: (Table~\ref{tbl-most}; \cite{w&h97}).  In addition, the flare observed
357: in 1996 had an initial e-folding decay time of $\sim 1.4$ days,
358: similar in scale to the times quoted above (\cite{hunt97}).  It is
359: tempting to speculate that event S$_2$ may be a prompt response to the
360: brief hard X-ray event at TJD 9610.  A similar association at TJD 9579
361: can be made for S$_1$ but seems less plausible.
362: 
363: 
364: \subsection{ATCA}\label{atca_lc}
365: 
366: \placefigure{fig-atca}
367: 
368: Figure~\ref{fig-atca} shows the ATCA Stokes I light curves at all six
369: observing frequencies, again plotted on a log-linear scale to
370: emphasize the overall exponential decay; the MOST light curve is
371: included for comparison.
372: 
373: The ATCA observations were initiated during the decay phase of the S1
374: short duration burst component.  It is clear from
375: Figure~\ref{fig-atca} that the flux density at all six ATCA
376: frequencies declines simultaneously with the MOST flux density between
377: TJD~$\sim 9579-9580$ suggesting that the short burst component was
378: present at {\it all\/} frequencies.  Following the minimum at
379: TJD~$\sim 9580$ the flux density increases for all six ATCA
380: frequencies, reaching a local maximum of 5.49~Jy at 1.4~GHz on
381: TJD~9581.62; note, however, that the overall maximum flux density at
382: frequencies $\geq4.8$~GHz occurs {\it prior\/} to the minimum.  For
383: TJD~$>9582$ there is a steady decline, with time constants similar to
384: MOST's at the lower frequencies but shorter at the higher frequencies
385: (4.8--9.2 GHz).  Other features, such as the dip at the higher
386: frequencies around TJD~9583 and the rise between 9596 and 9598 at
387: 8.6~GHz, are discussed in later sections.
388: 
389: 
390: \section{Linear polarization}\label{linpol} 
391: 
392: Linear polarization was detected at all six ATCA frequencies
393: throughout the epoch of the observations. The fractional polarization
394: at each frequency, ${\rm P/I}\,=\,\sqrt{({\rm Q}^2+{\rm
395: U}^2)}\,/\,{\rm I}$, expressed as a percentage, is plotted together
396: in Figure~\ref{fig-linpol} as a function of time, and the values are
397: listed in Table~\ref{tbl-linpol}.  Because Figure~\ref{fig-linpol} is
398: rather complex, and difficult to interpret on its own, we also include
399: Figure~\ref{fig-p+i} which shows separate overlaid plots of the polarized and
400: total intensity at each frequency.
401: 
402: The frequency-dependent time evolution of the P/I light curves is
403: reminiscent of the `expanding synchrotron bubble' model (e.g.,
404: \cite{h&j88}; \cite{b&v93}).  This simple model is characterized by an
405: optically thick rising phase in which the flux density peaks first at
406: the highest frequencies, followed by delayed peaks of lower intensity
407: at progressively lower frequencies.  The decay at all frequencies has
408: a power-law time dependence, with an exponent linked to the electron
409: energy distribution.
410: 
411: We have marked three apparent `bubble-type' events in the linearly
412: polarized ATCA data, and labelled them E1, E2 and E3 in
413: Figures~\ref{fig-linpol} and \ref{fig-p+i}, using the 8.6 and 9.2~GHz
414: curves for reference. However, it is worth noting that the beamsize of
415: the ATCA is larger than the angular size of \groj1655 (HR95) at all
416: frequencies, so we are observing effects which encompass both the
417: central object and the ejecta. As a result, we can only hope to
418: establish qualitative comparison with the sychrotron bubble model.  In
419: fact, contrary to the model prediction, Figure~\ref{fig-atca} shows
420: that the source was optically thin at TJD 9582, the time of the major
421: peak in flux density.
422: 
423: \placetable{tbl-linpol}
424: 
425: \placefigure{fig-linpol}
426: 
427: \placefigure{fig-p+i}
428: 
429: Figures~\ref{fig-linpol} and \ref{fig-p+i} show that the first event,
430: E1, is characterized by a series of declining peaks in the 8.6 and 9.2
431: GHz data. The peaks at 4.8 and 5.9~GHz are less well defined and their
432: decay is clearly delayed with respect to the higher frequencies, in
433: qualitative agreement with the expanding bubble model.
434: The delays at 1.4 and 2.3~GHz are sufficiently large that the
435: contributions from successive ejecta blend with one another and
436: produce a broad peak in both P and P/I.  It is interesting to note
437: that the polarized flux density (Fig.~\ref{fig-p+i}) reached
438: essentially the same maximum value (${\rm P}\simeq0.2$~Jy) at all
439: frequencies.
440: 
441: The early stages of event E2 also show qualitative agreement with the
442: expanding synchrotron bubble model, with the rise in P/I at the higher
443: frequencies preceding that at lower frequencies. However, the four
444: higher frequencies all reach a maximum together and then fall to a
445: minimum at the same time, behavior which is not consistent with the
446: model. Event E3 is also inconsistent with the model, as the fractional
447: polarizations at 4.8--9.2~GHz increase and decrease simultaneously.
448: We discuss possible interpretations of this behavior in Section
449: \ref{physint}.
450: 
451: \subsection{Position angle and rotation measure}\label{pa+rm}
452: 
453: Faraday rotation of the plane of polarization has been taken into
454: account by fitting the observed position angles (PA) to the usual
455: equation, ${\rm PA} = {\rm PA}_0 + ({\rm RM}) \lambda^2$, where PA$_0$
456: is the intrinsic position angle, RM is the rotation measure (in
457: rad~m$^{-2}$) and $\lambda$ is the wavelength.  The data were first
458: interpolated to a common epoch, namely that of the 9.2~GHz
459: observations.  The position angle was determined from the equation
460: ${\rm PA}=\case{1}{2}[{\rm arctan}\,{\rm (U/Q)} + n{\pi}]$, the last
461: term reflecting the ambiguities of $n\pi$ that may arise at each
462: frequency, and then plotted against $\lambda^2$.  A weighted
463: least-squares linear fit yielded the intrinsic position angle
464: (y-intercept) and the RM (slope).  Since the position angle is an
465: average over the beamsize, this means that it is also an average over
466: the central source and the ejecta.
467: 
468: \placefigure{fig-pa+rm}
469: 
470: \placetable{tbl-parm}
471: 
472: Figure~\ref{fig-pa+rm} shows the temporal evolution of the intrinsic
473: polarization position angle and rotation measure.  The average
474: position angle between TJD~9579 and 9589 was $\sim 47\arcdeg$, which
475: is consistent with the position angle of the jets (\cite{h&r95}). This
476: implies that the magnetic field is perpendicular to the jets, a
477: phenomenon also seen in the parsec-scale relativistic jets from
478: powerful radio-loud AGN (e.g.  \cite{b&p84}). The overall constancy in
479: position angle suggests that the magnetic field is well ordered, and
480: maintains basically the same orientation. However, on TJD~9589.5 the
481: position angle drops to $\sim 26\arcdeg$, and after a brief increase
482: to $\sim 43\arcdeg$, remains at $\sim 24\arcdeg \pm 4\arcdeg$ during
483: the last three observing epochs.
484: 
485: 
486: The rotation measure increases rapidly from $\sim 6$ rad m$^{-2}$ on TJD~9579.5 
487: to $\sim 104$ rad m$^{-2}$ on TJD~$\sim 9582$. It then levels off at $\sim 60$
488: rad m$^{-2}$ for the rest of the observing period, the only exception
489: being an increase to $\sim 80$ rad m$^{-2}$ on TJD~9598.  Its relative
490: constancy after TJD~9582 suggests that $\sim 60$ rad m$^{-2}$ arises
491: from the path through the interstellar medium, while the rapid
492: increase prior to TJD~9582 probably reflects local effects arising at
493: or near the source.  The low values of rotation measure around
494: TJD~9579.5 imply that the magnetic field at the source is in the
495: opposite direction to that of the interstellar medium, and the
496: increase to $\sim 104$ rad m$^{-2}$ indicates that the orientation of
497: the magnetic field has changed direction and is now more-or-less
498: aligned with that in the interstellar medium.
499: 
500: \placefigure{fig-paplots}
501: 
502: While Faraday effects predict that the rotation of the plane of
503: polarization is proportional to $\lambda^2$, the validity of the
504: $\lambda^2$ dependence has been questioned by O'Dea (1989).  He
505: presents polarization observations of fifteen core-dominated quasars,
506: arguing that poor fits to the PA vs. $\lambda^2$ data for six of the
507: quasars cannot be due to large apparent RM values combined with
508: ambiguities of $n\pi$, but rather can be accounted for by
509: wavelength-dependent polarization structure.  A similar situation may
510: well apply to \groj1655.  The upper panel of Figure~\ref{fig-paplots}
511: shows a least-squares fit to the polarization data on days TJD~9588.60
512: and 9590.53, which resulted in ${\rm PA}_0=47.5^{\circ}$, ${\rm
513: RM}=56.4$ rad~m$^{-2}$ and ${\rm PA}_0=38.2^{\circ}$, ${\rm RM}=63.5$
514: rad~m$^{-2}$ respectively.  Both fits appear secure.
515: 
516: In contrast, the lower panel of Figure~\ref{fig-paplots} shows an
517: example of PA vs. $\lambda^2$ for \groj1655 on TJD~9589.53, where
518: $\pi$ has been added to both the 1.4 and 2.3~GHz data points (open
519: circles). A least-squares fit to the data (solid line) yielded ${\rm
520: PA}_0=26.1^{\circ}$, ${\rm RM}=65.5$~rad~m$^{-2}$, and a large reduced
521: $\tilde{\chi}^2$ of 12.8 for 4 d.o.f.  We tried to accommodate the
522: high frequency points by adding $4\pi$ and $2\pi$ to the 1.4 and
523: 2.3~GHz data respectively (filled circles) and obtained ${\rm
524: PA}_0=-5.8^{\circ}$, ${\rm RM}=282.9$ rad~m$^{-2}$ (dotted line) and
525: an improved $\tilde{\chi}^2$ of 5.3.  However, these fitted parameters
526: are so inconsistent with the values obtained on TJD~9588.60 and
527: 9590.53 (with $\tilde{\chi}^2$ values of 0.31 and 0.05 respectively)
528: that they seem implausible.  Hence, we conclude that the discrepancies
529: seen in the high frequency points on TJD~9589.53, especially those at
530: 8.6 and 9.2~GHz, are due to wavelength dependent polarization
531: structure in the source.  Considering that \groj1655 is a complex
532: time-variable source, and that the observations are integrations over
533: multiple evolving ejecta, a poor fit to a simple $\lambda^2$
534: dependence is perhaps to be expected.
535: 
536: 
537: \section{Spectra}\label{spectralstuff}
538: 
539: \subsection{Spectral indices}\label{specind}
540: 
541: To fully investigate the evolution of the radio spectrum, two sets of
542: spectral indices were considered: the first set covers the entire
543: MOST/HartRAO monitoring period, from TJD~$\sim 9580$--9618, while the
544: second set concentrates on the ATCA data which, due to better
545: sampling, show more detail but for a shorter period of time.
546: Two-point spectral indices $\alpha$ were calculated, assuming spectra
547: of the form $S_{\nu}\propto\nu^{\alpha}$.  To examine the spectral
548: evolution more closely, low- and high-frequency spectral indices were
549: calculated separately.  For the MOST/HartRAO data these consisted of
550: the 843~MHz--5~GHz ($\alpha_{0.8-5}$) and 5--8.58~GHz
551: ($\alpha_{5-8.5}$) bands, while for the ATCA observations the data
552: were divided into the 1.4--2.3~GHz ($\alpha_{1.4-2.3}$) and
553: 4.8--8.6~GHz ($\alpha_{4.8-8.6}$) bands. In addition, a `global'
554: spectral index was obtained from the MOST 843~MHz and HartRAO 8.58~GHz
555: data ($\alpha_{0.8-8.5}$) and the MOST and ATCA 9.2~GHz data
556: ($\alpha_{0.8-9.2}$).  
557: 
558: The single-dish HartRAO flux densities are less accurate than those
559: from the ATCA (and the VLA) and will be affected by confusing sources
560: in the beam; an image of the field of \groj1655 is shown in Hunstead
561: et al. (1997).  On the other hand, the HartRAO data span a wider time
562: interval than the ATCA and fill in vital gaps in the VLA light curves.
563: Furthermore, while confusion will affect the spectral indices
564: systematically, trends with time will be preserved.
565: 
566: \subsubsection{MOST and HartRAO}
567: 
568: The lower panel in Figure~\ref{fig-most+rao} shows the spectral
569: indices from the MOST and HartRAO data plotted as a function of time.
570: The MOST and HartRAO 8.58~GHz data points were interpolated to the
571: HartRAO 5 GHz epochs.  Between TJD~9580--9593 and TJD~9610.5--9616
572: only the global $\alpha_{0.8-8.5}$ spectral index is plotted, while
573: from TJD 9593 to 9610, $\alpha_{0.8-5}$ and $\alpha_{5-8.5}$ are
574: plotted separately. Prior to TJD~9593, 
575: $\alpha_{0.8-8.5}$ ranges between $-0.6$ and $-0.4$, whereas after
576: TJD~9610 the spectrum is slightly flatter overall with
577: $\alpha_{0.8-8.5} \approx -0.3$.  At the time of the {\sc vla3}
578: ejection event (TJD $\sim9596$), the 5--8.58~GHz spectrum inverts to
579: $\alpha_{5-8.5}\approx +0.25$ (with a large uncertainty); this
580: mini-outburst is also seen in the VLA data (\cite{h&r95}), most
581: notably at 15 and 22 GHz.  The 5--8.58~GHz spectral index then returns
582: to the pre-outburst value at TJD $\sim 9600$.  A second
583: outburst, also recorded by the VLA, occurs a few days later, with
584: $\alpha_{5-8.5}$ peaking at $\sim +0.4$ at TJD~$\sim 9607$ 
585: and steepening soon thereafter.  The inversion of the
586: spectrum, indicating the emergence of an optically thick component, is
587: consistent with the {\sc vlba2} ejection event; the precise time
588: correspondence is, however, uncertain because neither the HartRAO nor
589: VLA dataset covers the critical interval TJD 9601--9606.
590: 
591: 
592: 
593: \subsubsection{MOST and ATCA}\label{most_atca_specind}
594: 
595: \placefigure{fig-specind}
596: 
597: Figure~\ref{fig-specind} shows the spectral indices obtained from the
598: MOST/ATCA datasets plotted as a function of time.  To calculate the
599: 843~MHz--9.2~GHz index the MOST data points were interpolated to the
600: 9.2~GHz epochs.  The temporal evolution of the three datasets is
601: similar, with $\alpha_{0.8-9.2}$ and $\alpha_{1.4-2.3}$ ranging from
602: $\sim -0.4$ to $\sim -0.6$, while $\alpha_{4.8-8.6}$ (excluding the
603: last two points at TJD~9596 and 9598) varies between $-0.5$ and
604: $-0.8$, indicating that the high-frequency portion of the spectrum is
605: consistently steeper. There is an overall steepening of the spectrum
606: between TJD~9579 and 9582, as the source becomes optically thin at the
607: lower frequencies.  Immediately after this, the high-frequency
608: spectral index begins to flatten and is followed by the lower
609: frequencies on TJD~$\sim 9583$; the light curves in Figure~\ref{fig-atca} 
610: shows a clear dip and rise at this epoch.  This
611: behavior is consistent with the ejection of an optically thick
612: synchrotron component and points to an ejection epoch TJD~$\sim 9582$,
613: somewhat earlier than inferred for the {\sc vla2} event
614: (\cite{h&r95}).  The spectra are flattest on TJD~$\sim 9586$, after
615: which the spectrum steepens.  Between TJD~9596.33 and 9598.07,
616: $\alpha_{4.8-8.6}$ jumps from $-$0.87 to $-$0.21, coincident with the
617: {\sc vla3} ejection episode; the corresponding increase in the 8.6~GHz
618: flux density can be seen in Figure~\ref{fig-atca}.
619: 
620: 
621: 
622: \subsection{Linear polarization spectra}\label{linpolspectra}
623: 
624: 
625: \placefigure{fig-spectra}
626: 
627: In Figure~\ref{fig-spectra} we show a montage of overlaid plots of the
628: total (I) and polarized (P) intensity spectra of \groj1655.  To
629: construct the spectra, the flux densities at all frequencies,
630: including 843~MHz (but without polarization), were interpolated to the
631: epochs of the ATCA 9.2~GHz observations.  The P spectra, in
632: particular, are especially valuable as diagnostics of ejection events
633: and their evolution with time.  We now discuss these spectra in
634: conjunction with the overlaid fractional polarization plot in Figure~\ref{fig-linpol} 
635: and the light curves in Figure~\ref{fig-atca}.
636: 
637: On the first day, TJD 9579.6 (four observations spanning two hours),
638: the P spectrum is strongly inverted and the I spectrum shows spectral
639: flattening at the low frequencies.  On the following day TJD~9580.50,
640: the I spectrum has become essentially a power law.  The P spectrum
641: shows a marked increase at 2.3~GHz but is still optically thick at 1.4
642: GHz. By TJD~9581.64, near the peak in total intensity, the P spectrum
643: has flattened, with signs of an upturn at the highest frequencies
644: signalling another ejection (E1).  The following four P spectra,
645: spanning TJD~9582.17--9583.59, trace a steady increase in the
646: polarized flux density at the lower frequencies, a pattern which is
647: qualitatively consistent with the expansion of the
648: synchrotron-emitting region(s).
649: 
650: At TJD~9584.63 the P spectrum is still not completely transparent at
651: the lower frequencies but has steepened markedly at the higher
652: frequencies, with signs of a new component (E2) appearing at 8.6 and
653: 9.2 GHz.  The development of this new component over the next three
654: days is again consistent with the expected trend for an expanding,
655: optically-thick synchrotron bubble.  Over this same time interval the
656: low frequency P spectrum has become optically thin.  Between TJD
657: 9587.62 and 9588.62 we interpret the persistently high level of P for
658: frequencies $\geq 4.8$~GHz as indicating yet another ejection event,
659: even though there is no perceptible change in the total flux density.
660: 
661: From TJD 9588.62 to 9589.55 there is an interesting transition.  At
662: all frequencies the total flux density from TJD 9589.55--9591.37 lies
663: well above an extrapolation of the exponential decay seen over the
664: previous four days (Fig.\ \ref{fig-atca}), implying that a
665: flat-spectrum component has been introduced.  A simultaneous flux
666: density increase across the spectrum is {\it not\/} a characteristic
667: of the synchrotron bubble model, so an alternative interpretation must
668: be sought.
669: 
670: Accompanying the flux density increase at TJD 9589.55 is a sharp
671: decrease in the high-frequency polarized flux density and fractional
672: polarization which occurs simultaneously at the four frequencies $\geq
673: 4.8$~GHz.  The high-frequency polarized flux `reappears' on TJD
674: 9590.55 and 9591.37 but is quenched again in the same fashion two days
675: later on TJD 9593.36.  We put forward a speculative but
676: self-consistent explanation of this behavior in Section \ref{physint}.
677: 
678: 
679: 
680: \section{Comparison with the VLA/VLBA}\label{compare}
681: 
682: In Sections \ref{lightcurves}, \ref{linpol} and \ref{spectralstuff} we
683: have described various aspects of the radio behavior of \groj1655 as
684: observed with the MOST, ATCA and HartRAO. We now discuss this behavior
685: in the light of the images and light curves recorded with the VLA and
686: VLBA (\cite{h&r95}).
687: 
688: The light curves presented here (Figs. 1--3) are completely
689: consistent with those produced at the VLA (HR95), with an outburst
690: peaking on ${\rm TJD}\sim 9582$ followed by a more-or-less exponential
691: decay.  Due to sparser temporal coverage the VLA did not record the
692: two short duration outbursts clearly observed with MOST (S$_1$ and
693: S$_2$) and HartRAO (S$_2$ only).  
694: 
695: As discussed in Section \ref{linpolspectra}, the ATCA linear
696: polarization spectra (Fig.~\ref{fig-spectra}) seem to give the
697: clearest indication of plasmon ejection and expansion.  The E1 event
698: is probably composed of several short-lived contributions, beginning
699: at ${\rm TJD} \sim 9580$, and possibly associated with ejection event
700: {\sc vla1} on TJD 9578.  The {\sc vla2} event (\cite{h&r95}),
701: originating at TJD 9584, is not well defined because of the 6-day gap
702: in VLA coverage from TJD 9583 to 9589.  It was noted previously
703: (Section \ref{most_atca_specind}) that the overall spectral indices
704: pointed towards a major outburst beginning near TJD 9582.  This
705: interpretation is supported by our polarization spectra which suggest
706: that events beginning near TJD 9581.6 and 9584.6 may be confusing the
707: interpretation of the component positions recorded later by the VLA.
708: 
709: Both the {\sc vla3} and {\sc vlba2} events are well traced and
710: supported by our spectral index plots in Figure \ref{fig-most+rao} and
711: Figure \ref{fig-specind}.  It is worth noting, however, that the
712: steepness of the high-frequency P and I spectra in the last panel in
713: Figure~\ref{fig-spectra} suggests that the {\sc vla3} event began {\it
714: after\/} TJD 9596.32.
715: 
716: %*******************************************************************
717: 
718: \section{Physical Interpretation}\label{physint} 
719: 
720: The radio emission from candidate black hole binaries is usually
721: attributed to optically thin synchrotron emission or thermal free-free
722: emission. The former explains the steep high frequency part of the
723: power-law spectrum, while the latter is invoked if the spectrum is
724: flat, extending to high frequency.
725: 
726: \subsection{Polarization and the Synchrotron Bubble Model}\label{synchbub} 
727: 
728: In the simple synchrotron bubble model (e.g.\ Hjellming \& Johnston
729: 1988; Ball \& Vlassis 1993) the emission region is a homogeneous
730: spherical, expanding plasma cloud, composed of relativistic electrons
731: and embedded in a magnetic field. As the cloud (bubble) expands
732: adiabatically, the synchrotron emission from it becomes optically thin
733: at progressively lower frequencies.  A natural consequence, therefore,
734: is that the total intensity I and fractional polarization P/I will
735: peak first at high frequency and later, with smaller amplitude, at
736: lower frequencies.
737: 
738: How well does this model describe the observed properties of \groj1655?
739: To begin with, the radio
740: emission is strongly polarized, giving direct evidence for the
741: presence of synchrotron emission (Fig.~\ref{fig-linpol}).  During
742: event E1, P/I peaks at 9.2 and 8.6~GHz ahead of 5.9~GHz and 4.8~GHz.
743: In addition, the maximum values of P/I at the high frequencies are
744: larger than those at the lower frequencies, the only exception being
745: the 1.4~GHz P/I which is explained as a blend of contributions from
746: several outbursts.  Event E2 displays a similar pattern.  While these
747: characteristics are in {\it qualitative\/} agreement with the
748: synchrotron bubble model, after event E2 the P/I values at
749: 4.8--9.2~GHz reached a minimum at almost the same time, which is not
750: predicted.  Thereafter, during event E3, P and P/I rise and fall
751: simultaneously at 4.8--9.2~GHz, again contrary to model predictions.
752: The polarization spectra (Section \ref{linpolspectra}, Figure~\ref{fig-spectra}) 
753: paint a similar picture.  Lack of compliance with
754: the synchrotron bubble model is perhaps not surprising considering the
755: complex collimated structures revealed by the VLBA observations (HR95)
756: and the simplicity of the model itself.
757: 
758: 
759: \subsection{``Core-Lobe'' Model }\label{l+c} 
760: 
761: As polarization is detected, we cannot dismiss the contribution from
762: synchrotron emission. However, we have shown that the synchrotron
763: bubble model cannot provide a satisfactory explanation for the time
764: evolution of the total intensity spectrum or the polarization in \groj1655.  
765: As a modification, we propose a new model which takes
766: into account the fact that VLBI observations of GRO J1655$-$40 have
767: shown multiple, time-varying emission regions (HR 95, Tingay et al.\
768: 1995).  A hybrid model which brings in other emission mechanisms is
769: therefore worth exploring for this very complex system.
770: 
771: There are two examples in the literature where multiple components
772: have been invoked to explain the light curves and/or polarization:
773: V404 Cyg (\cite{han}) and GRS\,1915$+$105 (Fender et al. 1999).
774: 
775: \begin{itemize}
776: 
777: \item The X-ray transient, V404 Cyg (\cite{han}), shows time-dependent
778: polarization properties similar to those seen in \groj1655, including
779: little variation in polarization position angle over the 50-day
780: monitoring period.  However, in contrast to GRO~J1655$-$40, V404 Cyg
781: showed a rapid initial decay, in which the spectrum was
782: optically-thin, followed by a much slower decline during which the
783: radio spectrum remained flat or inverted.  Han \& Hjellming (1992)
784: interpret this behavior as requiring at least two radiating
785: components.
786: 
787: \item
788: During a series of major ejection events in the superluminal jet X-ray
789: binary, GRS\,1915$+$105, recorded at high angular resolution with
790: MERLIN, Fender et al.\ (1999) reported that the (stationary) core did
791: not show any significant polarization, even though the approaching
792: component was significantly polarized.  They note that the core
793: emission was dominated by flat spectrum oscillations and attribute the
794: lack of polarization to either the superposition of multiple
795: components with different polarization position angles, or to large
796: Faraday depolarization close to the binary system.
797: 
798: \end{itemize}
799: 
800: In our hybrid model for \groj1655 we assume that the emission regions
801: consist of a `core' (an extended plasma cloud surrounding the central
802: source) and `lobes' (the expanding plasmons or ejecta). The core
803: represents the ejecta when they are localized near the binary; it is
804: therefore dense and compact, and unresolved in the VLBA image.  For a
805: beamwidth $\sim$15 mas and an estimated distance of $\sim 3$~kpc (HR95),
806:  we estimate that the corresponding size of the
807: emitting region should be $\la 10^{14}$~cm.  Synchrotron emission
808: is inefficient in the core, because of strong self-absorption. During
809: the initial stage of ejection, the core is most compact and the
810: electron density is highest, making free-free emission a very
811: efficient process.  The electrons in the ejecta are presumably
812: accelerated by some non-thermal (magnetic) processes near the black
813: hole, so their energy distribution will have a non-Maxwellian
814: component giving rise to non-thermal free-free emission. If the
815: density in the core is sufficiently high to allow particle collisions,
816: a sub-population of electrons with a Maxwellian distribution will
817: result. These electrons will emit thermal free-free emission.
818: 
819: When the ejecta leave the immediate environment of the binary and
820: expand to form the lobes, which are then resolved in the radio images,
821: they eventually become transparent to synchrotron emission. Moreover,
822: when the density drops, free-free emission becomes less efficient,
823: leaving synchrotron emission, which is no longer self absorbed, as the
824: dominant mechanism for energy loss.
825: 
826: According to this model, events E1 and E2 can be understood as two
827: consecutive episodes of mass ejection. During the onset of the
828: ejection, the lobes and the core are not actually separable. As the
829: emission region is compact, the synchrotron emission is self-absorbed.
830: There is also a free-free component.  Hence the overall spectrum shows
831: a turn-over, with the critical frequency determined jointly by the
832: emissivities of the synchrotron and free-free processes.  When the
833: individual plasma bubbles begin to separate from the core and expand,
834: their free-free emission becomes unimportant. Moreover, they become
835: progressively transparent to synchrotron emission, first at the high
836: frequencies and then at the low. The emission from the compact core,
837: on the other hand, may still be dominated by free-free emission, which
838: is weak in comparison with the synchrotron emission from the expanding
839: lobes.
840: 
841: Free-free emission has a flat spectrum, and so any free-free emission
842: from the core does not contribute significantly to the low frequency
843: part of the overall spectrum. For most of the time, therefore, the
844: spectral properties and polarization at low frequency are
845: characterized by the synchrotron emission from the ejecta, especially
846: when that emission peaks.  However, if the synchrotron component has
847: declined substantially, or there is a substantial brightening in the
848: emission from the core, the spectral index and the fractional
849: polarization will both be affected, becoming observable first at the
850: high frequencies and then at the lower frequencies.  The simultaneous
851: occurrence of the minima in P/I between events E2 and E3, together
852: with the simultaneous rise and fall in P/I at the four higher
853: frequencies during event E3, can be readily explained by dilution of
854: the synchrotron emission by a varying free-free contribution from the
855: core.  The magnetic field disruption resulting from core-brightening
856: episodes may also explain the apparently anomalous polarization
857: position angles (Fig.\ \ref{fig-pa+rm}) at TJD 9589.55, 9593.36,
858: 9596.32 and 9598.07.
859: 
860: 
861: A free-free core component can also explain the polarization spectra
862: for ${\rm TJD} > 9588.6$.  Consideration of the timescales evident in
863: Figure~\ref{fig-spectra}, and the VLBI structures (\cite{h&r95},
864: \cite{tin95}), tells us that the polarized emission at low frequency
865: comes from regions well away from the core, whereas the high-frequency
866: polarized emission is concentrated close to the core.  A sudden
867: increase in the size or electron density in the core will therefore
868: have a significant impact on P at high frequency and little or no
869: effect at low frequency, which seems to explain qualitatively what
870: occurs in the transition between TJD 9588.62 and 9589.55 (Fig.\
871: \ref{fig-spectra}).  We note that a small but significant upturn in P
872: at TJD 9589.55 between 8.6 and 9.2 GHz probably signals a new ejection
873: event. The following days, TJD 9590.55 and 9591.37, see an increase in
874: the polarized flux at high frequency, presumably because the ejecta
875: have moved outside the region of core absorption.  Interpretation of
876: the last two panels in Figure~\ref{fig-spectra} is complicated by the
877: poorer time sampling, but it is plausible that a similar series of
878: events follows the minimum in P and P/I at TJD 9593.36.
879: 
880: Although the hybrid `core-lobe' model has provided a more satisfactory
881: explanation for the radio emission from \groj1655 during the 1994
882: ejection episodes, we emphasize that the model is only qualitative at
883: this stage.  Further work is necessary to quantify the model, so that
884: fits to the data can be carried out.
885: 
886: \section{Summary and conclusions}\label{summary}
887: 
888: As Figure~\ref{fig-most+batse} shows, there is a relationship between
889: the hard X-ray and radio emission from \groj1655. The ejection
890: episodes traced by the VLBA both originated during enhanced activity
891: in the hard X-rays, and preceded the radio outbursts recorded with the
892: MOST, suggesting a connection between activity near the event horizon
893: of the black hole and the production of relativistic electrons. One
894: possible explanation for the decline in intensity of the radio
895: outbursts with time could be that the first ejection occurred in an
896: environment that was relatively undisturbed, whereas the subsequent
897: outbursts will have taken place in an environment already disrupted by
898: previous activity (\cite{hjell96}).  The implications of the
899: X-ray/radio correlations have been discussed elsewhere (e.g.
900: \cite{harmon95}), with a general consensus that the hard X-rays may be
901: indicating enhanced accretion near the black hole which, through
902: processes still not well understood, triggers the formation of
903: relativistic radio jets.
904: 
905: In general the ATCA, MOST and HartRAO flux density light curves agree
906: well with the VLA light curves and ejection epochs reported in
907: HR95. However, the better time sampling in Figures \ref{fig-most+rao}
908: and \ref{fig-atca} reveals the presence of short-lived events that
909: were not recorded by the VLA or VLBA, and shows that the light curves
910: do not decay as smooth exponentials.  The radio spectra, especially
911: the linear polarization spectra, have proved to be valuable
912: diagnostics of the timing of plasmon ejection events and their
913: subsequent evolution.
914: 
915: The ATCA polarization data show that the magnetic field is aligned at
916: right angles to the radio jets (jet ${\rm PA} = 47 \pm 1^{\circ}$), except
917: towards the end of the monitoring period when core contributions may
918: have become important.  The rotation measure initially shows a
919: contribution local to the \groj1655 system, but after TJD 9582.5, the
920: rotation measure is roughly constant at $\sim$60~rad~m$^{-2}$ which
921: must correspond to the interstellar value.
922: 
923: After examining the time evolution of the total and polarized flux
924: density of \groj1655 we conclude that there are specific aspects of
925: the behavior which cannot be explained by the simple
926: synchrotron bubble model.  We therefore invoke a hybrid `core-lobe'
927: model, with a core which emits by non-thermal (or maybe thermal)
928: free-free emission and lobes which are classical synchrotron
929: emitters.  We suggest that a similar model may apply to the other
930: Galactic superluminal jet X-ray binary, GRS 1915+105.
931: 
932: \acknowledgments
933: 
934: The Australia Telescope Compact Array is funded by the Commonwealth of
935: Australia for operation as a National Facility managed by CSIRO.  MOST
936: is operated by the University of Sydney and funded by grants from the
937: Australian Research Council.  KW acknowledges the support of the
938: Australian Research Council through an Australian Research Fellowship.
939: DM acknowledges support for his research by the European Union under
940: contract FMGECT950012, and thanks N.P.F.\ McKay and M.J.\ Kesteven for
941: assisting with the ATCA observations.  DH acknowledges financial
942: support from the Academy of Finland.  The director of HartRAO,
943: G. Nicolson, is thanked for the allocation of observing time that made
944: this project possible. We thank T.\ Ye for assistance with the MOST
945: data reduction, and J-P Macquart for his invaluable help in checking
946: some of the position angle fits.
947: 
948: 
949: 
950: \clearpage
951: 
952: \begin{thebibliography}{}
953: \bibitem[Bailyn et al. 1995]{bailyn95} Bailyn, C.D., Orosz, J.A.,
954: 			McClintock, J.E. \& Remillard, R.A. 1995,
955: 			\nat, 378, 157
956: \bibitem[Ball \& Vlassis 1993]{b&v93} Ball, L. \& Vlassis, M. 1993,
957: PASA, 10, 342
958: \bibitem[Bridle \& Perley 1984]{b&p84} Bridle, A.H. \& Perley,
959: R.A. 1984, \araa, 22, 319
960: \bibitem[Campbell-Wilson \& Hunstead (1994a)]{cw&h94a}
961: Campbell-Wilson, D. \& Hunstead, R.W. 1994a, \iaucirc \ 6052
962: \bibitem[Campbell-Wilson \& Hunstead 1994b]{cw&h94b} Campbell-Wilson,
963: D. \& Hunstead, R.W. 1994b, \iaucirc \ 6055
964: \bibitem[Campbell-Wilson \& Hunstead 1994c]{cw&h94c} Campbell-Wilson,
965: 	D. \& Hunstead, R.W. 1994c, PASA, 11, 33
966: \bibitem[Fender et al. 1999]{fen99} Fender, R.P., Garrington, S.T.,
967: 	McKay, D.J., Muxlow, T.W.B., Pooley, G.G., Spencer, R.E.,
968: 	Stirling, A.M. \& Waltman, E.B. 1999, MNRAS, 304, 865
969: \bibitem[Frater, Brooks \& Whiteoak 1992]{frater} Frater, R.H.,
970: 		Brooks, J.W. \& Whiteoak, J.B. 1992, Journal of
971: 		Electrical and Electronics Engineering, Australia, 12,
972: 		103
973: \bibitem[Han \& Hjellming 1992]{han}Han, X. \& Hjellming, R.M. 1992,
974:                         \apj, 400, 304
975: \bibitem[Hannikainen, Hunstead \& Campbell-Wilson 1998]{hann}
976: 			Hannikainen, D., Hunstead, R. \&
977: 			Campbell-Wilson, D. 1998, NewAR, 42, 601
978: \bibitem[Harmon et al. 1995]{harmon95} Harmon, B.A., Wilson, C.A.,
979: 		 	Zhang, S.N., Paciesas, W.S., Fishman, G.J.,
980: 		 	Hjellming, R.M., Rupen, M.P., Scott, D.M.,
981: 		 	Briggs, M.S. \& Rubin, B.C. 1995 \nat, 374,
982: 		 	703
983: \bibitem[Hjellming \& Johnston 1988]{h&j88} Hjellming, R.M. \&
984: Johnston, K.J. 1988, \apj, 328, 600
985: \bibitem[HR95]{h&r95} Hjellming, R.M. \& Rupen, M.P. 1995, \nat, 375, 464 (HR95)
986: \bibitem[Hjellming et al. 1996]{hjell96} Hjellming, R.M, Rupen, M.P.,
987: 			Shrader, C.R., Campbell-Wilson, D.  Hunstead,
988: 			R.W. \& McKay, D.J. 1996, \apj, 470, L105
989: \bibitem[Hunstead, Wu \& Campbell-Wilson 1997]{hunt97} Hunstead, R.W.,
990: 			Wu, K. \& Campbell-Wilson, D. 1997, in
991: 			Accretion Phenomena and Related Outflows, IAU
992: 			Colloq. 163, eds. D. T. Wickramasinghe,
993: 			L. Ferrario \& G. Bicknell, ASP
994: 			Conf. Ser. 121, 63
995: \bibitem[McKay \& Kesteven 1994]{m&k94} McKay, D.J. \& Kesteven,
996:                         M. 1994, \iaucirc 6062
997: \bibitem[Mills 1981]{mills81} Mills, B.Y. 1981, PASA, 4, 156
998: \bibitem[O'Dea 1989]{odea89} O'Dea, C.P. 1989, \apj, 210, 35
999: \bibitem[Orosz \& Bailyn 1997]{o&b97} Orosz, J.A. \& Bailyn,
1000:                         C.D. 1997, \apj, 477, 876
1001: \bibitem[Ott et al. 1994]{ott94} Ott, M., Witzel, A., Quirrenbach, A.,
1002: 	Krichbaum, T.P., Standke, K.J., Schalinski, C.J. \& Hummel,
1003: 	C.A. 1994, \aap, 284, 331
1004: \bibitem[Reynolds 1994]{rey94} Reynolds, J.E. 1994, ATNF Technical
1005: 			Document AT/39.3/040, Australia Telescope
1006: 			National Facility
1007: \bibitem[Robertson 1991]{rob91} Robertson, J.G. 1991,
1008: Aust. J. Physics, 44, 729
1009: \bibitem[Sault, Teuben \& Wright 1995]{bob95} Sault, R.J., Teuben,
1010:                         P.J. \& Wright, M.C.H. 1995, in Astronomical
1011:                         Data Analysis Software and Systems IV,
1012:                         ed. R. Shaw, H.E. Payne, J.J.E. Hayes,
1013:                         Astronomical Society of the Pacific Conference
1014:                         Series, 77, 433-436
1015: \bibitem[Sault, Killeen \& Kesteven (1991)]{bob91} Sault, R.J.,
1016: 			Killeen, N.E.B. \& Kesteven, M.J. 1991, ATNF
1017: 			Technical Document AT/39.3/015, Australia
1018: 			Telescope National Facility
1019: \bibitem[Soria et al. 1998]{soria} Soria, R., Wickramasinghe, D.T.,
1020: 			Hunstead, R.W. \& Wu, K. 1998, \apj, 495, L95
1021: \bibitem[Tingay et al. 1995]{tin95} Tingay, S.J., Jauncey, D.L.,
1022: 			Preston, R.A., Reynolds, J.E. et al. 1995,
1023: 			\nat, 374, 141
1024: %\bibitem[van der Laan 1966]{vdl66} van der Laan, H. 1966, \nat, 211, 1131
1025: \bibitem[Wu \& Hunstead 1997]{w&h97} Wu, K. \& Hunstead, R.W. 1997, in
1026: 			Accretion Phenomena and Related Outflows, IAU
1027: 			Colloq. 163, eds. D. T. Wickramasinghe,
1028: 			L. Ferrario \& G. Bicknell, ASP
1029: 			Conf. Ser. 121, 835
1030: \bibitem[Zhang et al. 1994]{nan94} Zhang, S.N., Wilson, C.A., Harmon,
1031:                         B.A., Fishman, G.J., Wilson, R.B., Paciesas,
1032:                         W.S., Scott, M. \& Rubin, B.C. 1994, \iaucirc
1033:                         6046
1034: \end{thebibliography}
1035: 
1036: 
1037: \clearpage
1038: 
1039: \begin{figure}
1040: %\plotone{most+batse.eps}
1041: \plotone{figure1.eps}
1042: \caption{The 1994 MOST and BATSE light curves (\cite{w&h97}).  The
1043: open arrows mark the ejection epochs determined by \cite{h&r95} from
1044: the VLA observations (which we label as {\sc vla1}, {\sc vla2} and
1045: {\sc vla3}), while the filled arrows ({\sc vlba1} and {\sc vlba2})
1046: denote their estimated VLBA ejection epochs.  S$_1$ and S$_2$ mark
1047: short duration components discussed in the text.  The duration of the
1048: ATCA observations is marked by a horizontal line in the upper
1049: panel.\label{fig-most+batse}}
1050: \end{figure}
1051: 
1052: \begin{figure}
1053: \epsscale{1}
1054: %\plotone{most+rao.eps}
1055: \plotone{figure2.eps}
1056: \caption{The upper panel shows the MOST 843~MHz and HartRAO 5~GHz
1057: light curves on a logarithmic flux density scale, with the VLA and
1058: VLBA ejection epochs from \cite{h&r95} labelled as in Figure 1.
1059: Labels S$_1$, S$_2$, L$_1$ and L$_2$, discussed in Section
1060: \ref{short}, mark short and long duration components respectively.
1061: The 843~MHz -- 8.58~GHz, 843~MHz -- 5~GHz, and 5 -- 8.58~GHz spectral
1062: indices derived from the combined MOST and HartRAO data are plotted in
1063: the lower panel.\label{fig-most+rao}}
1064: \end{figure}
1065: 
1066: \begin{figure}
1067: \epsscale{1}
1068: %\plotone{atca.eps}
1069: \plotone{figure3.eps}
1070: \caption{The MOST 843~MHz and ATCA light curves shown on a logarithmic flux
1071: density scale. The error bars lie largely within the plotting
1072: symbols. \label{fig-atca}}
1073: \end{figure}
1074: 
1075: \begin{figure}
1076: \epsscale{1}
1077: %\plotone{linpol_color.eps}
1078: \plotone{figure4.eps}
1079: \caption{The fractional polarization (P/I, expressed as a percentage)
1080: at all six ATCA frequencies plotted as a function of time.  Labels E1,
1081: E2 and E3 are used to mark specific events in polarized intensity, as
1082: discussed in the text. \label{fig-linpol}}
1083: \end{figure}
1084: 
1085: \begin{figure}
1086: \epsscale{1}
1087: %\plotone{p+i.eps}
1088: \plotone{figure5.eps}
1089: \caption{Total (solid line) and polarized (dotted line) flux density
1090: versus time at each of the six ATCA frequencies.  E1, E2 and E3 refer
1091: to the same outburst events as in Figure
1092: \ref{fig-linpol}.  Labels on the left-hand y-axes refer to the total
1093: flux density (I), while the right-hand side labels refer to the
1094: polarized flux density (P).
1095: \label{fig-p+i}}
1096: \end{figure}
1097: 
1098: \begin{figure}
1099: \epsscale{1}
1100: %\plotone{pa+rm.eps}
1101: \plotone{figure6.eps}
1102: \caption{Upper panel: the intrinsic polarization position angle
1103: (degrees) as a function of time; the dotted line marks 47$^{\circ}$,
1104: the position angle of the radio jets (\cite{h&r95}).  Lower panel:
1105: rotation measure (rad m$^{-2}$) as a function of time.  The last two
1106: points were obtained using only the 1.4, 2.3, 4.8 and 8.6~GHz
1107: data. \label{fig-pa+rm}}
1108: \end{figure}
1109: 
1110: \begin{figure}
1111: \epsscale{1}
1112: %\plotone{paplots.eps}
1113: \plotone{figure7.eps}
1114: \caption{Upper panel: PA vs. $\lambda^2$ on TJD~9588.60 and
1115: TJD~9590.53 together with the fitted lines. Lower panel: PA
1116: vs. $\lambda^2$ on TJD~9589.53; the open circles and solid line (the
1117: adopted fit) show the result of adding $\pi$ to both the 1.4 and
1118: 2.3~GHz points, while the filled circles and dotted line show the
1119: result of adding $4\pi$ to the 1.4~GHz data and $2\pi$ to the 2.3~GHz
1120: data; see Section 4.1.  
1121: \label{fig-paplots}}
1122: \end{figure}
1123: 
1124: \begin{figure}
1125: \epsscale{1}
1126: %\plotone{specind.eps}
1127: \plotone{figure8.eps}
1128: \caption{The 843~MHz--9.2~GHz, 1.4--2.3~GHz and 4.8--8.6~GHz spectral
1129: indices, from the combined MOST and ATCA data, plotted as a function
1130: of time.\label{fig-specind}}
1131: \end{figure}
1132: 
1133: 
1134: \begin{figure}
1135: \epsscale{1}
1136: %\plotone{spectra.eps}
1137: \plotone{figure9.eps}
1138: \caption{Spectra constructed from the MOST and ATCA data.  The filled
1139: dots and solid lines represent the combined 843~MHz--9.2~GHz MOST and
1140: ATCA total flux density spectra (left-hand y-axis labels).  The open
1141: circles and dashed lines represent the polarized flux from the ATCA
1142: 1.4--9.2~GHz data (right-hand y-axis labels).  The spectra on
1143: TJD~9596.32 are constructed from the 1.4, 2.3, 4.8 and 8.6~GHz data
1144: points alone (in addition to the MOST data for the total flux density
1145: spectrum).  The observing epoch is marked in each
1146: plot. \label{fig-spectra}}
1147: \end{figure}
1148: 
1149: 
1150: 
1151: \clearpage
1152:  
1153: \begin{deluxetable}{lll}
1154: \scriptsize
1155: \tablecaption{Journal of MOST observations. \label{tbl-most}}
1156: \tablewidth{0pt}
1157: \tablehead{
1158: \colhead{UT Date} & \colhead{JD $-$ 2,440,000.5\tablenotemark{a}}   & \colhead{ S$_{0.843}$\tablenotemark{b}}  \\
1159: \colhead{} & \colhead{} & \colhead{(Jy)}
1160: }
1161: \startdata 
1162: 1994 Aug 6  &      \hspace{0.4cm} 9570.60  &   0.362  $\pm$ 0.020\nl
1163: 1994 Aug 11 &      \hspace{0.4cm} 9575.59  &   0.792 $\pm$ 0.054 \nl
1164: 1994 Aug 12 &      \hspace{0.4cm} 9576.33  &   0.950 $\pm$ 0.057$^{\ast}$ \nl
1165: 1994 Aug 14 &      \hspace{0.4cm} 9578.55  &   4.86 $\pm$ 0.15 \nl
1166:            &       \hspace{0.4cm} 9578.61  &   5.33 $\pm$ 0.16  \nl
1167: 1994 Aug 15 &      \hspace{0.4cm} 9579.22  &   5.45 $\pm$ 0.16$^{\ast}$ \nl
1168:             &      \hspace{0.4cm} 9579.38  &   5.30 $\pm$ 0.16  \nl
1169:             &      \hspace{0.4cm} 9579.55  &   4.97 $\pm$ 0.15  \nl
1170: 1994 Aug 16 &      \hspace{0.4cm} 9580.14  &   4.25  $\pm$ 0.13$^{\ast}$  \nl
1171:             &      \hspace{0.4cm} 9580.28  &   4.34 $\pm$ 0.13  \nl
1172:             &      \hspace{0.4cm} 9580.51  &   5.08 $\pm$ 0.15  \nl
1173: 1994 Aug 17 &      \hspace{0.4cm} 9581.13  &   6.99 $\pm$ 0.21$^{\ast}$ \nl
1174:             &      \hspace{0.4cm} 9581.28  &   6.88 $\pm$ 0.21   \nl
1175:             &      \hspace{0.4cm} 9581.50  &   7.08 $\pm$ 0.21   \nl
1176: 1994 Aug 18 &      \hspace{0.4cm} 9582.13  &   7.62 $\pm$ 0.23$^{\ast}$  \nl
1177:             &      \hspace{0.4cm} 9582.27  &   7.14 $\pm$ 0.21   \nl
1178:             &      \hspace{0.4cm} 9582.52  &   6.39 $\pm$ 0.19  \nl
1179: 1994 Aug 19 &      \hspace{0.4cm} 9583.13  &   5.43 $\pm$ 0.16$^{\ast}$  \nl
1180:             &      \hspace{0.4cm} 9583.29  &   5.15 $\pm$ 0.15  \nl
1181:             &      \hspace{0.4cm} 9583.54  &   4.96 $\pm$ 0.15 \nl
1182: 1994 Aug 20 &      \hspace{0.4cm} 9584.21  &   4.39 $\pm$ 0.13$^{\ast}$  \nl
1183:             &      \hspace{0.4cm} 9584.56  &   4.16 $\pm$ 0.12   \nl
1184: 1994 Aug 21 &      \hspace{0.4cm} 9585.18  &   3.83 $\pm$ 0.11$^{\ast}$  \nl
1185:             &      \hspace{0.4cm} 9585.31  &   3.41 $\pm$ 0.10   \nl
1186:             &      \hspace{0.4cm} 9585.53  &   3.19 $\pm$ 0.10  \nl
1187: 1994 Aug 22 &      \hspace{0.4cm} 9586.14  &   2.93 $\pm$ 0.09$^{\ast}$  \nl
1188:             &      \hspace{0.4cm} 9586.29  &   2.89 $\pm$ 0.09$^{\ast}$  \nl
1189:             &      \hspace{0.4cm} 9586.46  &   2.71 $\pm$ 0.08   \nl
1190: 1994 Aug 23 &      \hspace{0.4cm} 9587.08  &   2.43 $\pm$ 0.07$^{\ast}$  \nl
1191:             &      \hspace{0.4cm} 9587.28  &   2.33 $\pm$ 0.07$^{\ast}$  \nl
1192:             &      \hspace{0.4cm} 9587.46  &   2.28 $\pm$ 0.07  \nl
1193: 1994 Aug 24 &      \hspace{0.4cm} 9588.39  &   2.10 $\pm$ 0.06$^{\ast}$ \nl
1194:             &      \hspace{0.4cm} 9588.52  &   2.11 $\pm$ 0.06  \nl
1195: 1994 Aug 25 &      \hspace{0.4cm} 9589.21  &   1.91 $\pm$ 0.06$^{\ast}$ \nl
1196:             &      \hspace{0.4cm} 9589.47  &   1.83 $\pm$ 0.05  \nl
1197: 1994 Aug 26 &      \hspace{0.4cm} 9590.12  &   1.81 $\pm$ 0.05$^{\ast}$ \nl
1198:             &      \hspace{0.4cm} 9590.26  &   1.84 $\pm$ 0.06  \nl
1199:             &      \hspace{0.4cm} 9590.55  &   1.59 $\pm$ 0.05  \nl
1200: 1994 Aug 28 &      \hspace{0.4cm} 9592.10  &   1.40 $\pm$ 0.04$^{\ast}$ \nl
1201:             &      \hspace{0.4cm} 9592.54  &   1.27 $\pm$ 0.04$^{\ast}$ \nl
1202: 1994 Aug 29 &      \hspace{0.4cm} 9593.25  &   1.09 $\pm$ 0.03  \nl
1203:             &      \hspace{0.4cm} 9593.48  &   0.998 $\pm$ 0.03 \nl
1204: 1994 Sept 1 &      \hspace{0.4cm} 9596.32  &   0.720 $\pm$ 0.022$^{\ast}$ \nl
1205: 1994 Sept 4 &      \hspace{0.4cm} 9599.24  &   0.444 $\pm$ 0.016  \nl
1206: 1994 Sept 5 &      \hspace{0.4cm} 9600.46  &   0.413 $\pm$ 0.013  \nl
1207: 1994 Sept 6 &      \hspace{0.4cm} 9601.43  &   0.412 $\pm$ 0.013  \nl
1208: 1994 Sept 8 &      \hspace{0.4cm} 9603.46  &   0.393 $\pm$ 0.014  \nl
1209: 1994 Sept 10 &     \hspace{0.4cm} 9605.51  &   0.457 $\pm$ 0.015 \nl
1210: 1994 Sept 11 &     \hspace{0.4cm} 9606.20  &   0.472 $\pm$ 0.015  \nl
1211: 1994 Sept 13 &     \hspace{0.4cm} 9608.35  &   0.511 $\pm$ 0.052$^{\ast}$ \nl
1212: 1994 Sept 14 &     \hspace{0.4cm} 9609.32  &   1.92 $\pm$ 0.06$^{\ast}$  \nl
1213:              &     \hspace{0.4cm} 9609.46  &   2.18 $\pm$ 0.07  \nl
1214: 1994 Sept 15 &     \hspace{0.4cm} 9610.27  &   1.86 $\pm$ 0.06$^{\ast}$  \nl
1215:              &     \hspace{0.4cm} 9610.42  &   1.84 $\pm$ 0.06  \nl
1216: 1994 Sept 16 &     \hspace{0.4cm} 9611.33  &   1.16 $\pm$ 0.03$^{\ast}$  \nl
1217: 1994 Sept 18 &     \hspace{0.4cm} 9613.19  &   0.565 $\pm$ 0.018 \nl
1218: 1994 Sept 20 &     \hspace{0.4cm} 9615.43  &   0.505 $\pm$ 0.016 \nl
1219: 1994 Sept 22 &     \hspace{0.4cm} 9617.43  &   0.322 $\pm$ 0.014  \nl
1220: 1994 Sept 26 &     \hspace{0.4cm} 9621.41  &   0.248 $\pm$ 0.008  \nl
1221: 1994 Sept 30 &     \hspace{0.4cm} 9625.42  &   0.151 $\pm$ 0.007  \nl
1222: 1994 Oct 3 &       \hspace{0.4cm} 9628.11  &   0.121 $\pm$ 0.006   \nl
1223: 1994 Oct 5 &       \hspace{0.4cm} 9630.38  &   0.077 $\pm$ 0.005  \nl
1224: 1994 Oct 6 &       \hspace{0.4cm} 9631.42  &   0.078 $\pm$ 0.007  \nl
1225: 1994 Oct 10&       \hspace{0.4cm} 9635.40  &   0.047 $\pm$ 0.005   \nl
1226: 1994 Oct 12 &      \hspace{0.4cm} 9637.38  &   0.040 $\pm$ 0.004   \nl
1227: 1994 Oct 18 &      \hspace{0.4cm} 9643.39  &   0.034 $\pm$ 0.006   \nl
1228: 1994 Oct 20 &      \hspace{0.4cm} 9645.21  &   0.024 $\pm$ 0.003   \nl
1229: 1994 Nov 10&       \hspace{0.4cm} 9666.37  &   0.232 $\pm$ 0.014   \nl
1230: 1994 Nov 11 &      \hspace{0.4cm} 9667.29  &   0.205 $\pm$ 0.009  \nl
1231: 1994 Nov 12 &      \hspace{0.4cm} 9668.16  &   0.116 $\pm$ 0.004   \nl
1232: 1994 Nov 13 &      \hspace{0.4cm} 9669.07  &   0.074 $\pm$ 0.002   \nl
1233: 1994 Nov 14 &      \hspace{0.4cm} 9670.32  &   0.056 $\pm$ 0.003   \nl
1234: 1994 Nov 16 &      \hspace{0.4cm} 9672.33  &   0.075 $\pm$ 0.006   \nl
1235: \enddata
1236: \tablenotetext{a}{At mid-observation.}
1237: \tablenotetext{b}{Measurements made in fan-beam SCAN mode are marked with $^{\ast}$.}
1238: \end{deluxetable}
1239: 
1240: 
1241: \clearpage
1242: 
1243:  
1244: \begin{deluxetable}{rcrrrrrr}
1245: \scriptsize
1246: \tablecaption{Journal of ATCA observations\tablenotemark{a}. \label{tbl-atca}}
1247: \tablewidth{0pt}
1248: \tablehead{
1249: \colhead{UT Date} & \colhead{JD $-$ 2,440,000.5\tablenotemark{b}}   & \colhead{S$_{1.4}$}   & \colhead{S$_{2.3}$} & 
1250: \colhead{S$_{4.8}$}  & \colhead{S$_{5.9}$} & \colhead{S$_{8.6}$} & 
1251: \colhead{S$_{9.2}$} \\
1252: \colhead{} & \colhead{} & \colhead{(Jy)} & \colhead{(Jy)} & \colhead{(Jy)} &
1253: \colhead{(Jy)} & \colhead{(Jy)} & \colhead{(Jy)}
1254: } 
1255: \startdata 
1256: 1994 Aug 15 & 9579.35 & \nodata   & \nodata   & 2.79 & \nodata   & 1.85 & \nodata   \nl 
1257:             & 9579.36 & \nodata   & \nodata   & \nodata   & 2.41 & \nodata   & 1.77 \nl
1258:             & 9579.39 & \nodata   & \nodata   & 2.74 & \nodata   & 1.82 & \nodata   \nl
1259:             & 9579.40 & \nodata   & \nodata   & \nodata   & 2.37 & \nodata   & 1.74 \nl
1260:             & 9579.41 & \nodata   & \nodata   & 2.72 & \nodata   & 1.82 & \nodata   \nl
1261:             & 9579.42 & \nodata   & \nodata   & \nodata   & 2.37 & \nodata   & 1.75 \nl
1262:             & 9579.46 & \nodata   & \nodata   & \nodata   & \nodata   & \nodata   & 1.71 \nl
1263:             & 9579.52 & 4.29 & 3.52 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1264:             & 9579.53 & \nodata   & \nodata   & 2.52 & \nodata   & 1.66 & \nodata   \nl
1265:             & 9579.54 & \nodata   & \nodata   & \nodata   & 2.18 & \nodata   & 1.57 \nl
1266:             & 9579.55 & 4.23 & 3.45 & \nodata   & \nodata   & \nodata   & \nodata   \nl 
1267:             & 9579.56 & \nodata   & \nodata   & 2.47 & \nodata   & 1.61 & \nodata   \nl
1268:             & 9579.57 & 4.18 & 3.39 & \nodata   & 2.13 & \nodata   & 1.52 \nl
1269:             & 9579.58 & \nodata   & \nodata   & 2.44 & \nodata   & 1.57 & \nodata \nl
1270:             & 9579.59 & \nodata   & \nodata   & \nodata   & 2.09 & \nodata   & 1.49 \nl
1271:             & 9579.60 & 4.13 & 3.32 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1272:             & 9579.61 & \nodata   & \nodata   & 2.40 & \nodata   & 1.54 & \nodata   \nl
1273:             & 9579.62 & 4.07 & 3.25 & \nodata   & 2.06 & \nodata   & 1.46  \nl
1274:             & 9579.63 & \nodata   & \nodata   & 2.35 & \nodata   & 1.52 & \nodata   \nl
1275: 1994 Aug 16 & 9580.49 & 4.71 & 3.50 & 2.35 & \nodata   & 1.57 & \nodata   \nl
1276:             & 9580.50 & \nodata   & \nodata   & \nodata   & 2.10 & \nodata   & 1.53 \nl
1277: 1994 Aug 17 & 9581.62 & 5.49 & 3.96 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1278:             & 9581.63 & \nodata   & \nodata   & 2.51 & \nodata   & 1.61 & \nodata   \nl
1279:             & 9581.64 & \nodata   & \nodata   & \nodata   & 2.17 & \nodata   & 1.54 \nl
1280: 1994 Aug 18 & 9582.15 & 5.35 & 3.88 & \nodata      &  \nodata     & \nodata      & \nodata      \nl
1281:             & 9582.16 & \nodata   & \nodata   & 2.45 & \nodata   & 1.60 & \nodata   \nl
1282:             & 9582.17 & \nodata   & \nodata   & \nodata   & 2.09 & \nodata   & 1.52 \nl
1283:             & 9582.63 & 4.34 & 3.13 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1284:             & 9582.64 & \nodata   & \nodata   & 2.06 & \nodata   & 1.39 & \nodata   \nl
1285:             & 9582.65 & \nodata   & \nodata   & \nodata   & 1.78 & \nodata   & 1.30 \nl
1286: 1994 Aug 19 & 9583.13 & 3.82 & 2.73 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1287:             & 9583.14 & \nodata   & \nodata   & 1.85 & \nodata   & 1.28 & \nodata   \nl
1288:             & 9583.15 & \nodata   & \nodata   & \nodata   & 1.64 & \nodata   & 1.24 \nl
1289:             & 9583.58 & 3.74 & 2.77 & 1.98 & \nodata   & 1.42 & \nodata   \nl
1290:             & 9583.59 & \nodata   & \nodata   & \nodata   & 1.77 & \nodata   & 1.37 \nl
1291: 1994 Aug 20 & 9584.61 & 3.05 & 2.38 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1292:             & 9584.62 & \nodata   & \nodata   & 1.82 & \nodata   & 1.32 & \nodata   \nl
1293:             & 9584.63 & \nodata   & \nodata   & \nodata   & 1.55 & \nodata   & 1.20 \nl
1294: 1994 Aug 21 & 9585.62 & 2.50 & 2.02 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1295:             & 9585.63 & \nodata   & \nodata   & 1.57 & \nodata   & 1.18 & \nodata   \nl
1296:             & 9585.64 & \nodata   & \nodata   & \nodata   & 1.42 & \nodata   & 1.14 \nl
1297: 1994 Aug 22 & 9586.61 & 2.00 & 1.61 & \nodata   & \nodata   & \nodata   & \nodata   \nl
1298:             & 9586.62 & \nodata   & \nodata   & 1.23 & \nodata   & 0.915 & \nodata  \nl
1299:             & 9586.63 & \nodata   & \nodata   & \nodata   & 1.11 & \nodata    & 0.875 \nl
1300: 1994 Aug 23 & 9587.60 & 1.68 & 1.31 & \nodata   & \nodata   & \nodata    & \nodata   \nl
1301:             & 9587.61 & \nodata   & \nodata   & 0.962 & \nodata    & 0.683 & \nodata  \nl
1302:             & 9587.62 & \nodata   & \nodata   & \nodata   & 0.858  & \nodata   & 0.662 \nl
1303: 1994 Aug 24 & 9588.60 & 1.44 & 1.09 & \nodata   &  \nodata    & \nodata    & \nodata  \nl
1304:             & 9588.61 & \nodata   & \nodata   & 0.787 & \nodata    & 0.555 & \nodata  \nl
1305:             & 9588.62 & \nodata   & \nodata   & \nodata   & 0.689  & \nodata    & 0.519 \nl
1306: 1994 Aug 25 & 9589.53 & 1.39 & 1.03 & \nodata   &  \nodata    & \nodata    & \nodata   \nl
1307:             & 9589.54 & \nodata   &  \nodata  & 0.720 & \nodata    & 0.506 & \nodata   \nl
1308:             & 9589.55 & \nodata   &  \nodata  & \nodata    & 0.638 & \nodata    & 0.483 \nl
1309: 1994 Aug 26 & 9590.53 & 1.24 & 0.918 & \nodata   & \nodata    & \nodata    & \nodata    \nl
1310:             & 9590.54 & \nodata   &  \nodata  & 0.642 & \nodata    & 0.448 & \nodata  \nl
1311:             & 9590.55 & \nodata   & \nodata   &  \nodata   & 0.566 & \nodata    & 0.426 \nl
1312: 1994 Aug 27 & 9591.35 & 1.07 & 0.821 & \nodata   & \nodata    & \nodata    & \nodata   \nl
1313:             & 9591.36 & \nodata   & \nodata   & 0.556 & \nodata    & 0.365 & \nodata   \nl
1314:             & 9591.37 & \nodata   &  \nodata  &  \nodata   & 0.487 & \nodata    & 0.350 \nl
1315: 1994 Aug 29 & 9593.35 & 0.806 & 0.636 & 0.465  & \nodata & 0.318   & \nodata    \nl
1316:             & 9593.36 & \nodata    & \nodata  &  \nodata   & 0.413 & \nodata    & 0.306 \nl
1317: 1994 Sept 1 & 9596.32 & 0.630 & 0.488 & \nodata  &  \nodata   & \nodata    & \nodata    \nl
1318:             & 9596.33 & \nodata    &  \nodata   & 0.324 & \nodata  & 0.195 & \nodata    \nl
1319: 1994 Sept 2 & 9597.61 & 0.493 & 0.378 & \nodata    & \nodata  & \nodata    & \nodata    \nl
1320: 1994 Sept 3 & 9598.07 & \nodata    & \nodata  & 0.241 & \nodata    & 0.212 & \nodata    \nl
1321: \enddata   
1322: \tablenotetext{a}{Flux densities are Stokes I values, with errors of 1\%.}
1323: \tablenotetext{b}{At mid-observation.}
1324: \end{deluxetable}
1325: 
1326: 
1327: \clearpage
1328: 
1329: \begin{deluxetable}{lclcl}
1330: \scriptsize
1331: \tablecaption{Journal of HartRAO observations. \label{tbl-rao}}
1332: \tablewidth{0pt}
1333: \tablehead{
1334: \colhead{UT Date} & \colhead{JD $-$ 2,440,000.5}   & \colhead{S$_{5}$}   & 
1335: \colhead{JD $-$ 2,440,000.5} & \colhead{S$_{8.58}$} \\
1336: \colhead{} & \colhead{} & \colhead{(Jy)} & \colhead{} & \colhead{(Jy)}
1337: } 
1338: \startdata 
1339: 1994 Aug 16 & 9580.86 & 1.88 $\pm$   0.06  & 9580.82 & 1.77 $\pm$  0.02  \nl
1340:             & 9580.88 & 2.05 $\pm$   0.04  & 9580.86 & 1.73 $\pm$  0.12  \nl
1341:             & 9580.90 & 2.00 $\pm$   0.04  & 9580.91 & 1.44 $\pm$  0.02 \nl
1342: 1994 Aug 18 & 9582.74 & 1.88 $\pm$   0.04  & 9582.75 & 1.57 $\pm$  0.03 \nl
1343:             & 9582.77 & 1.77 $\pm$   0.04  & 9582.77 & 1.46 $\pm$  0.02 \nl
1344: 1994 Aug 19 & 9583.52 & 1.59 $\pm$   0.06 & 9583.52 & 1.46 $\pm$  0.02 \nl
1345:             & 9583.53 & 1.76 $\pm$   0.05 & 9583.54 & 1.49 $\pm$  0.06  \nl
1346:             & 9583.55 & 1.85 $\pm$   0.06 & 9583.55 & 1.68 $\pm$  0.06 \nl
1347:             & 9583.57 & 1.82 $\pm$   0.05 & 9583.57 & 1.68 $\pm$  0.07 \nl
1348: 1994 Aug 20 & 9584.53 & 1.72 $\pm$   0.06 & 9584.54 & 1.44 $\pm$  0.05 \nl
1349:             & 9584.55 & 1.68 $\pm$   0.05 & 9584.56 & 1.54 $\pm$  0.06 \nl
1350:             & 9584.57 & 1.69 $\pm$   0.07 & 9584.58 & 1.52 $\pm$  0.07 \nl
1351: 1994 Aug 22 & 9586.52 & 1.14 $\pm$   0.05 & 9586.53 & 1.01 $\pm$  0.05 \nl
1352:             & 9586.54 & 1.12 $\pm$   0.04 & 9586.55 & 0.90 $\pm$  0.03  \nl
1353:             & 9586.56 & 1.23 $\pm$   0.06 & 9586.57 & 1.16 $\pm$  0.09 \nl
1354:             & 9586.59 & 1.20 $\pm$   0.06 & 9586.59 & 1.15 $\pm$  0.08  \nl
1355: 1994 Aug 23 & 9587.54 & 0.91 $\pm$   0.05 & 9587.54 & 0.76 $\pm$  0.02 \nl
1356:             & 9587.56 & 0.94 $\pm$   0.07 & 9587.64 & 0.67 $\pm$  0.02 \nl
1357:             & 9587.63 & 0.83 $\pm$   0.04 & 9587.66 & 0.78 $\pm$  0.01 \nl
1358:             & 9587.65 & 0.77 $\pm$   0.03 & 9587.68 & 0.84 $\pm$  0.04 \nl
1359:             & 9587.68 & 0.91 $\pm$   0.05 & 9587.71 & 0.69 $\pm$  0.02 \nl
1360:             & 9587.70 & 0.77 $\pm$   0.03 & 9587.73 & 0.77 $\pm$  0.07  \nl
1361:             & 9587.72 & 0.88 $\pm$   0.03 & 9587.75 & 0.76 $\pm$  0.04 \nl
1362:             & 9587.74 & 0.82 $\pm$   0.03 & 9587.79 & 0.76 $\pm$  0.02  \nl
1363:             & 9587.77 & 0.88 $\pm$   0.05 & 9587.81 & 0.77 $\pm$  0.02 \nl
1364:             & 9587.79 & 0.82 $\pm$   0.04 & 9587.84 & 0.88 $\pm$  0.10 \nl
1365:             & 9587.80 & 0.88 $\pm$   0.06 & 9587.86 & 0.89 $\pm$  0.05 \nl
1366:             & 9587.83 & 0.88 $\pm$   0.07 & 9587.88 & 0.93 $\pm$  0.02  \nl
1367:             & 9587.85 & 0.85 $\pm$   0.18 & \nodata     & \nodata      \nl
1368:             & 9587.87 & 0.77 $\pm$   0.03 & \nodata     & \nodata      \nl
1369:             & 9587.89 & 0.93 $\pm$   0.07 & \nodata     & \nodata      \nl
1370:             & 9587.92 & 0.91 $\pm$   0.07 &  \nodata    & \nodata      \nl
1371: 1994 Aug 24 & 9588.50 & 0.63 $\pm$   0.06 & 9588.50 & 0.68 $\pm$   0.07 \nl
1372:             & 9588.52 & 0.67 $\pm$   0.05 & 9588.52 & 0.65 $\pm$   0.03 \nl
1373:             & 9588.54 & 0.64 $\pm$   0.05 & 9588.55 & 0.59 $\pm$   0.06 \nl
1374:             & 9588.56 & 0.78 $\pm$   0.05 & 9588.57 & 0.60 $\pm$   0.04 \nl
1375: 1994 Aug 26 & 9590.51 & 0.61 $\pm$   0.04 & 9590.52 & 0.52 $\pm$   0.04 \nl
1376:             & 9590.54 & 0.56 $\pm$   0.04 & 9590.54 & 0.48 $\pm$   0.04  \nl 
1377:             & 9590.58 & 0.65 $\pm$   0.07 & 9590.56 & 0.48 $\pm$   0.03 \nl
1378:             &  \nodata    & \nodata                & 9590.59 & 0.52 $\pm$   0.07 \nl
1379: 1994 Aug 28 & 9592.59 & 0.32 $\pm$   0.13 & 9592.64 & 0.39 $\pm$   0.05 \nl
1380:             & 9592.62 & 0.39 $\pm$   0.03 & 9592.68 & 0.51 $\pm$   0.07 \nl
1381:             & 9592.64 & 0.39 $\pm$   0.04 & 9592.70 & 0.39 $\pm$   0.03 \nl
1382:             & 9592.66 & 0.51 $\pm$   0.10 & 9592.86 & 0.38 $\pm$   0.03 \nl
1383:             & 9592.67 & 0.53 $\pm$   0.03 & 9592.88 & 0.49 $\pm$   0.09 \nl
1384:             & 9592.69 & 0.52 $\pm$   0.03 & \nodata     & \nodata            \nl
1385:             & 9592.87 & 0.54 $\pm$   0.05 & \nodata     & \nodata            \nl
1386: 1994 Aug 31 & 9595.83 & 0.36 $\pm$   0.03 & 9595.83 & 0.35 $\pm$   0.03 \nl
1387:             & 9595.85 & 0.37 $\pm$   0.03 & 9595.86 & 0.43 $\pm$   0.02 \nl
1388:             & 9595.87 & 0.39 $\pm$   0.05 & 9595.88 & 0.49 $\pm$   0.13 \nl
1389: 1994 Sept 2 & 9597.75 & 0.23 $\pm$   0.08 & 9597.75 & 0.27 $\pm$   0.07 \nl
1390:             & 9597.77 & 0.27 $\pm$   0.07 & \nodata     & \nodata           \nl
1391:             & 9597.79 & 0.23 $\pm$   0.25 & \nodata     & \nodata           \nl
1392:             & 9597.81 & 0.24 $\pm$   0.05 & \nodata     & \nodata            \nl
1393:             & 9597.83 & 0.19 $\pm$   0.06 & \nodata     & \nodata             \nl
1394: 1994 Sept 3 & 9598.85 & 0.19 $\pm$   0.16 & 9598.83 & 0.21 $\pm$   0.01 \nl
1395:             & 9598.87 & 0.11 $\pm$   0.10 & 9598.86 & 0.10 $\pm$   0.05 \nl
1396:             & 9598.89 & 0.24 $\pm$   0.10 & 9598.90 & 0.12 $\pm$   0.09 \nl
1397: 1994 Sept 5 & 9600.82 & 0.16 $\pm$   0.06 & 9600.81 & 0.12 $\pm$   0.01 \nl
1398:             & 9600.86 & 0.26 $\pm$   0.07 & \nodata     & \nodata                \nl
1399: 1994 Sept 6 &  \nodata    & \nodata                &  9601.49 & 0.19 $\pm$  0.03 \nl
1400:             &   \nodata   & \nodata                &  9601.53 & 0.18 $\pm$  0.03 \nl
1401:             &   \nodata   & \nodata                &  9601.56 & 0.18 $\pm$  0.03 \nl
1402: 1994 Sept 8 &   \nodata   & \nodata                &  9603.62 & 0.33 $\pm$  0.08 \nl
1403: 1994 Sept 11 & 9606.50 & 0.42 $\pm$   0.05 & 9606.47 & 0.38 $\pm$  0.06 \nl
1404:              & 9606.52 & 0.25 $\pm$   0.06 & 9606.51 & 0.44 $\pm$  0.02 \nl
1405: 1994 Sept 12 & 9607.46 & 0.25 $\pm$   0.06 & \nodata     & \nodata           \nl
1406: 1994 Sept 14 & 9609.46 & 0.65 $\pm$   0.06 & 9609.47 & 0.72 $\pm$  0.07 \nl
1407:              & 9609.80 & 0.69 $\pm$   0.05 & 9609.81 & 0.77 $\pm$  0.06 \nl
1408:              & 9609.82 & 0.89 $\pm$   0.05 & 9609.85 & 0.78 $\pm$  0.04  \nl
1409:              & 9609.84 & 0.79 $\pm$   0.04 & 9609.87 & 0.74 $\pm$  0.10 \nl
1410:              & 9609.87 & 0.81 $\pm$   0.05 &  \nodata    & \nodata              \nl
1411: 1994 Sept 15 & 9610.61 & 0.85 $\pm$   0.03 & 9610.62 & 0.67 $\pm$  0.03 \nl
1412:              & 9610.66 & 0.79 $\pm$   0.03 & 9610.67 & 0.74 $\pm$  0.02 \nl
1413:              & 9610.70 & 0.74 $\pm$   0.03 & 9610.71 & 0.77 $\pm$  0.04 \nl
1414: 1994 Sept 17 & 9612.80 & 0.32 $\pm$   0.03 & 9612.80 & 0.27 $\pm$  0.04 \nl
1415:              & 9612.82 & 0.25 $\pm$   0.08 & \nodata     & \nodata              \nl
1416: 1994 Sept 20 & 9615.79 & 0.19 $\pm$   0.03 & 9615.79 & 0.18 $\pm$  0.04 \nl
1417:              &   \nodata   & \nodata                & 9615.81 & 0.22 $\pm$  0.01 \nl
1418:              &   \nodata   & \nodata                & 9615.85 & 0.26 $\pm$  0.02 \nl
1419: 1994 Sept 21 &   \nodata   & \nodata                & 9616.76 & 0.11 $\pm$  0.02 \nl
1420:              &   \nodata   & \nodata                & 9616.78 & 0.18 $\pm$  0.03 \nl
1421:              &   \nodata   & \nodata                & 9616.80 & 0.26 $\pm$  0.02 \nl
1422: 1994 Sept 24 &   \nodata   & \nodata                & 9619.71 & 0.13 $\pm$  0.05 \nl
1423:              &   \nodata   & \nodata                & 9619.76 & 0.15 $\pm$  0.03 \nl
1424: 1994 Oct 1   &   \nodata   & \nodata                & 9626.75 & 0.08 $\pm$ 0.04 \nl
1425: \enddata
1426: \end{deluxetable}
1427: 
1428: \clearpage
1429: 
1430: \begin{deluxetable}{rcrrrrrr}
1431: \scriptsize
1432: \tablecaption{Fractional linear polarization\tablenotemark{a}. \label{tbl-linpol}}
1433: %%($\sqrt{Q^2+U^2/I^2)\times100$). \label{tbl-2}}
1434: \tablewidth{0pt}
1435: \tablehead{
1436: \colhead{UT Date} & \colhead{JD $-$ 2,440,000.5\tablenotemark{b}}   & 
1437:         \multicolumn{6}{c}{P/I $\times$ 100} \\  \cline{3-8}\\[-3mm]
1438:  & & \colhead{1.4 GHz} & \colhead{2.3 GHz} & \colhead{4.8 GHz} & \colhead{5.9 GHz}
1439:    & \colhead{8.6 GHz} & \colhead{9.2 GHz} 
1440: }
1441: %\colhead{P/I$_{1.4}$}   & \colhead{P/I$_{2.3}$} & 
1442: %\colhead{P/I$_{4.8}$}  & \colhead{P/I$_{5.9}$} & \colhead{P/I$_{8.6}$} & 
1443: %\colhead{P/I$_{9.2}$}
1444: \startdata 
1445: 1994 Aug 15 & 9579.35 &   \nodata       & \nodata         & 4.70$\pm$0.11 &                 & 5.96$\pm$0.12 & \nodata  \nl 
1446:             & 9579.36 &   \nodata       & \nodata         &  \nodata       & 5.15$\pm$0.11 & \nodata         & 6.78$\pm$0.12 \nl
1447:             & 9579.39 &   \nodata       & \nodata         & 4.11$\pm$0.11 & \nodata         & 6.37$\pm$0.12 & \nodata \nl
1448:             & 9579.40 &   \nodata       & \nodata         & \nodata         & 5.23$\pm$0.11 &  \nodata        & 6.99$\pm$0.12 \nl
1449:             & 9579.41 &   \nodata       & \nodata         & 4.11$\pm$0.11 &   \nodata       & 6.22$\pm$0.12 & \nodata \nl
1450:             & 9579.42 &   \nodata       & \nodata         & \nodata         & 4.22$\pm$0.11 &  \nodata        & 6.86$\pm$0.12 \nl
1451:             & 9579.46 &   \nodata       & \nodata         &    \nodata      &  \nodata        &   \nodata       & 7.13$\pm$0.12 \nl
1452:             & 9579.52 & 0.561$\pm$0.100 & 0.834$\pm$0.100 & \nodata         & \nodata         & \nodata         & \nodata \nl
1453:             & 9579.53 &   \nodata       & \nodata         & 3.89$\pm$0.11 & \nodata         & 7.16$\pm$0.12 &\nodata \nl
1454:             & 9579.54 &   \nodata       & \nodata         &   \nodata       & 3.78$\pm$0.11 &   \nodata       & 7.77$\pm$0.13 \nl
1455:             & 9579.55 & 0.510$\pm$0.100 & 0.870$\pm$0.100 & \nodata         & \nodata         &  \nodata        &  \nodata \nl
1456:             & 9579.56 &   \nodata       & \nodata         & 3.76$\pm$0.11 & \nodata         & 7.32$\pm$0.12 & \nodata \nl
1457:             & 9579.57 & 0.311$\pm$0.100 & 0.945$\pm$0.100 &   \nodata       & 3.88$\pm$0.11 &   \nodata       & 7.98$\pm$0.13  \nl
1458:             & 9579.58 &   \nodata       & \nodata         & 3.82$\pm$0.11 & \nodata         & 7.48$\pm$0.12 &  \nodata\nl
1459:             & 9579.59 &   \nodata       & \nodata         &  \nodata        & 3.62$\pm$0.11 &    \nodata      & 7.83$\pm$0.13 \nl
1460:             & 9579.60 & 0.239$\pm$0.100 & 1.02$\pm$0.10 & \nodata         & \nodata         & \nodata         &  \nodata \nl
1461:             & 9579.61 &   \nodata       & \nodata         & 4.03$\pm$0.11 &  \nodata        & 7.49$\pm$0.12 &\nodata  \nl
1462:             & 9579.62 & 0.244$\pm$0.100 & 1.14$\pm$0.10 &   \nodata       & 3.53$\pm$0.11 &   \nodata       & 7.79$\pm$0.13  \nl
1463:             & 9579.63 &   \nodata       & \nodata         & 3.70$\pm$0.11 &  \nodata        & 7.21$\pm$0.12 & \nodata \nl
1464: 1994 Aug 16 & 9580.49 & 0.216$\pm$0.100 & 3.04$\pm$0.10   & 3.60$\pm$0.01 & \nodata         & 3.96$\pm$0.11 & \nodata \nl
1465:             & 9580.50 &   \nodata       & \nodata         & \nodata         & 3.94$\pm$0.11 & \nodata         & 3.96$\pm$0.11 \nl
1466: 1994 Aug 17 & 9581.62 & 1.49$\pm$0.10 & 2.77$\pm$0.10 & \nodata         & \nodata         & \nodata         & \nodata  \nl
1467:             & 9581.63 &   \nodata       & \nodata         & 4.45$\pm$0.11 & \nodata         & 11.34$\pm$0.15 & \nodata  \nl
1468:             & 9581.64 &   \nodata       & \nodata         &  \nodata        & 6.00$\pm$0.12 & \nodata         & 11.51$\pm$0.15 \nl
1469: 1994 Aug 18 & 9582.15 & 2.50$\pm$0.10 & 2.80$\pm$0.10 & \nodata         & \nodata         & \nodata         & \nodata  \nl
1470:             & 9582.16 &   \nodata       & \nodata         & 3.02$\pm$0.10 & \nodata         & 6.75$\pm$0.12 & \nodata \nl
1471:             & 9582.17 &   \nodata       & \nodata         & \nodata         & 6.08$\pm$0.12 &   \nodata       & 5.96$\pm$0.12 \nl
1472:             & 9582.63 & 2.51$\pm$0.10 & 3.23$\pm$0.10 &\nodata          & \nodata         & \nodata         & \nodata   \nl
1473:             & 9582.64 &    \nodata      & \nodata         & 8.92$\pm$0.13 &  \nodata        & 9.93$\pm$0.14 & \nodata \nl
1474:             & 9582.65 &    \nodata      & \nodata         &   \nodata       & 10.11$\pm$0.14 &    \nodata     & 9.68$\pm$0.14 \nl
1475: 1994 Aug 19 & 9583.13 & 2.82$\pm$0.10 & 5.49$\pm$0.11 & \nodata         & \nodata         & \nodata         & \nodata \nl
1476:             & 9583.14 &    \nodata      & \nodata         & 8.89$\pm$0.13 &    \nodata      & 6.95$\pm$0.12 & \nodata  \nl
1477:             & 9583.15 &    \nodata      & \nodata         &   \nodata       & 9.28$\pm$0.14 &     \nodata     & 6.42$\pm$0.12  \nl
1478:             & 9583.58 & 3.22$\pm$0.10 & 6.63$\pm$0.12     & 9.53$\pm$0.14 &  \nodata        & 7.27$\pm$0.12 & \nodata \nl
1479:             & 9583.59 &      \nodata    & \nodata         &     \nodata     & 9.48$\pm$0.14 &   \nodata       & 6.80$\pm$0.12 \nl
1480: 1994 Aug 20 & 9584.61 & 5.57$\pm$0.11 & 7.94$\pm$0.13 & \nodata         &\nodata          & \nodata         &  \nodata \nl
1481:             & 9584.62 &      \nodata    & \nodata         & 5.46$\pm$0.11 & \nodata         & 0.815$\pm$0.100 & \nodata \nl
1482:             & 9584.63 &      \nodata    & \nodata         &  \nodata        & 3.33$\pm$0.10 & \nodata         & 1.22$\pm$0.10 \nl
1483: 1994 Aug 21 & 9585.62 & 6.73$\pm$0.12 & 6.25$\pm$0.12 & \nodata         &  \nodata        & \nodata         & \nodata \nl
1484:             & 9585.63 &       \nodata   & \nodata         & 3.43$\pm$0.11 & \nodata         & 1.76$\pm$0.10 & \nodata   \nl
1485:             & 9585.64 &       \nodata   & \nodata         &  \nodata        & 1.16$\pm$0.10 &  \nodata        & 2.09$\pm$0.10 \nl
1486: 1994 Aug 22 & 9586.61 & 7.38$\pm$0.12 & 5.92$\pm$0.12 &\nodata          &\nodata          & \nodata          & \nodata \nl
1487:             & 9586.62 &        \nodata  & \nodata         & 1.22$\pm$0.10 &   \nodata       & 4.11$\pm$0.11 & \nodata \nl
1488:             & 9586.63 &        \nodata  & \nodata         &   \nodata       & 2.53$\pm$0.10 &       \nodata   & 4.20$\pm$0.11 \nl
1489: 1994 Aug 23 & 9587.60 & 8.30$\pm$0.13 & 4.40$\pm$0.11 & \nodata         & \nodata         & \nodata          & \nodata \nl
1490:             & 9587.61 &        \nodata  & \nodata         & 3.33$\pm$0.10 &   \nodata       & 5.28$\pm$0.11 & \nodata \nl
1491:             & 9587.62 &         \nodata & \nodata         &   \nodata       & 4.21$\pm$0.11 &  \nodata        & 5.26$\pm$0.11 \nl
1492: 1994 Aug 24 & 9588.60 & 8.97$\pm$0.13 & 2.86$\pm$0.10 & \nodata         & \nodata         & \nodata         & \nodata \nl
1493:             & 9588.61 &       \nodata   & \nodata         & 2.69$\pm$0.10 &  \nodata        & 4.63$\pm$0.11 & \nodata  \nl
1494:             & 9588.62 &       \nodata   & \nodata         &  \nodata        & 3.49$\pm$0.11 &   \nodata       & 4.78$\pm$0.11 \nl
1495: 1994 Aug 25 & 9589.53 & 9.45$\pm$0.14 & 6.77$\pm$0.12 & \nodata         & \nodata         & \nodata          & \nodata \nl
1496:             & 9589.54 &       \nodata   & \nodata         & 2.05$\pm$0.10 &  \nodata        & 1.27$\pm$0.10 & \nodata \nl
1497:             & 9589.55 &       \nodata   & \nodata         &   \nodata       & 1.63$\pm$0.10 &  \nodata        & 1.49$\pm$0.10  \nl
1498: 1994 Aug 26 & 9590.53 & 8.48$\pm$0.13 & 7.62$\pm$0.13 & \nodata         & \nodata         &  \nodata        & \nodata  \nl
1499:             & 9590.54 &        \nodata  &  \nodata        & 4.44$\pm$0.11 &  \nodata        & 2.87$\pm$0.10 & \nodata  \nl
1500:             & 9590.55 &       \nodata   &   \nodata       &  \nodata        & 3.69$\pm$0.11 &  \nodata        & 2.92$\pm$0.10 \nl
1501: 1994 Aug 27 & 9591.35 & 8.44$\pm$0.13 & 8.06$\pm$0.13 & \nodata         & \nodata         & \nodata          &  \nodata\nl
1502:             & 9591.36 &    \nodata      &   \nodata       & 4.78$\pm$0.11 &  \nodata        & 3.66$\pm$0.11 & \nodata \nl
1503:             & 9591.37 &    \nodata      &   \nodata       &     \nodata     & 4.20$\pm$0.11 &        \nodata  & 3.41$\pm$0.13 \nl
1504: 1994 Aug 29 & 9593.35 & 7.81$\pm$0.13 & 3.72$\pm$0.11     & 1.97$\pm$0.10 &  \nodata        & 1.40$\pm$0.11 & \nodata \nl
1505:             & 9593.36 &    \nodata      &   \nodata       &    \nodata      & 1.67$\pm$0.10 & \nodata         & 1.53$\pm$0.15 \nl
1506: 1994 Sept 1 & 9596.32 & 6.95$\pm$0.12 & 3.35$\pm$0.10 & \nodata         &\nodata          & \nodata          & \nodata \nl
1507:             & 9596.33 &    \nodata      &   \nodata       & 1.76$\pm$0.10 &    \nodata      & 1.63$\pm$0.14 & \nodata \nl
1508: 1994 Sept 2 & 9597.61 & 6.46$\pm$0.16 & 3.38$\pm$0.17 & \nodata         & \nodata         & \nodata          & \nodata  \nl
1509: 1994 Sept 3 & 9598.07 &    \nodata      &    \nodata      & 2.9$\pm$0.36 &   \nodata       & 0.669$\pm$0.531 &  \nodata \nl
1510: \enddata   
1511: 
1512: % Text for table footnotes follows the tabular data and must be inside the
1513: % deluxetable environment.  Note that it is OK to put \ref's in 
1514: % \tablenotetext's.
1515:  
1516: \tablenotetext{a}{Expressed as a percentage of the total flux density.}
1517: \tablenotetext{b}{At mid-observation.}
1518: 
1519: \end{deluxetable}
1520: 
1521: \clearpage
1522: 
1523: \begin{deluxetable}{cl@{\hspace{1mm}}c@{\hspace{1mm}}l@{\hspace{8mm}}l@{\hspace{1mm}}c@{\hspace{1mm}}l}
1524: \scriptsize
1525: \tablecaption{Position angle and rotation measure. \label{tbl-parm}}
1526: \tablewidth{0pt}
1527: \tablehead{
1528: \colhead{JD $-$ 2,440,000.5} & \multicolumn{3}{c}{PA~~~~}   & \multicolumn{3}{c}{RM} \\
1529: \colhead{ } & \multicolumn{3}{c}{$\arcdeg$~~~~} & \multicolumn{3}{c}{rad m$^{-2}$}
1530: }
1531: \startdata 
1532:       9579.58  &    41.97 &$\pm$   &1.13\tablenotemark{a}  &7.3 &$\pm$  &6.9\tablenotemark{a} \nl
1533:       9580.50  &    54.5  &$\pm$   &4.6   &     31   &$\pm$  &15 \nl
1534:       9581.64  &    53.48 &$\pm$   &0.99  &     65.3 &$\pm$  &4.1 \nl
1535:       9582.17  &    53.7  &$\pm$   &1.6   &    104.1 &$\pm$  &3.7 \nl
1536:       9582.65  &    49.06 &$\pm$   &0.15  &    59.74 &$\pm$  &0.47 \nl
1537:       9583.15  &    53.20 &$\pm$   &0.27  &    56.70 &$\pm$  &0.65 \nl
1538:       9583.59  &    48.95 &$\pm$   &0.55  &    58.66 &$\pm$  &1.25 \nl
1539:       9584.63  &    44.98 &$\pm$   &0.59  &    61.94 &$\pm$  &0.62 \nl
1540:       9585.64  &    45.65 &$\pm$   &1.48  &    60.18 &$\pm$  &1.20 \nl
1541:       9586.63  &    45.47 &$\pm$   &0.66  &    58.87 &$\pm$  &0.59 \nl
1542:       9587.62  &    42.89 &$\pm$   &0.59  &    58.39 &$\pm$  &0.57 \nl
1543:       9588.62  &    47.53 &$\pm$   &0.96  &    56.38 &$\pm$  &0.79 \nl
1544:       9589.55  &    26.1  &$\pm$   &5.0   &    65.5  &$\pm$  &3.2 \nl
1545:       9590.55  &    38.24 &$\pm$   &0.27  &    63.49 &$\pm$  &0.23 \nl
1546:       9591.37  &    42.92 &$\pm$   &0.32  &    61.80 &$\pm$  &0.29 \nl
1547:       9593.36  &    26.86 &$\pm$   &0.88  &    66.03 &$\pm$  &0.51 \nl
1548:       9596.32  &    19.9  &$\pm$   &3.3\tablenotemark{b}  &67.0 &$\pm$ &2.0\tablenotemark{b} \nl
1549:       9598.07  &    28.0  &$\pm$   &2.8\tablenotemark{b} & 79.9 &$\pm$ &2.1\tablenotemark{b} \nl
1550: \enddata   
1551: \tablenotetext{a}{Average of first four interpolated points.}
1552: \tablenotetext{b}{Obtained from the 1.4, 2.3, 4.8 \& 8.6~GHz points alone.}
1553: \end{deluxetable}
1554: 
1555: 
1556: \end{document}
1557: 
1558: 
1559: 
1560: 
1561: 
1562: 
1563: 
1564: