1: % Centaurus high resolution RM paper
2: %
3: % revised in response to comments of the referee
4:
5: \def\about {$\sim$}
6: \def\pks {PKS\,1246{\tt -}410}
7: \def\solmass {M$_\odot$}
8: \def\peryr {yr$^{-1}$}
9: \def\HI {H\kern0.1em{\sc i}}
10: \def\radm {rad m$^{-2}$}
11: \def\etal {{\sl et~al.\ }}
12: % for astro-ph
13: \voffset-.4in
14:
15: \documentclass[useAMS,usenatbib]{mn2e}
16: \usepackage{psfig}
17:
18: %\documentclass[12pt,twocolumn]{mn2e}
19:
20: %\documentstyle[psfig, twocolumn]{mn2e}
21:
22: \begin{document}
23: \title{Fields and Filaments in the Core of the Centaurus Cluster}
24: \author[]
25: {\parbox[]{6.in} {G. B. Taylor$^1$, A. C. Fabian$^2$, G. Gentile$^1$, S. W. Allen$^3$,
26: C. Crawford$^2$ and J. S. Sanders$^2$\\
27: \footnotesize
28: 1. University of New Mexico, Dept. of Physics and
29: Astronomy, Albuquerque, NM 87131, USA; gbtaylor@unm.edu\\
30: 2. Institute of Astronomy, Madingley Road,
31: Cambridge CB3 0HA, acf@ast.cam.ac.uk; csc@ast.cam.ac.uk\\
32: 3. Kavli Institute of Particle Astrophysics and Cosmology,
33: Stanford University, Stanford, CA 94305, USA; swa@stanford.edu\\
34: }}
35:
36: \maketitle
37: \begin{abstract}
38:
39: We present high resolution images of the Faraday Rotation Measure (RM)
40: structure of the radio galaxy \pks\, at the center of the Centaurus
41: cluster. Comparison with H$\alpha$-line and soft X-ray emission
42: reveals a correspondence between the line-emitting gas, the soft X-ray
43: emitting gas, regions with an excess in the RM images, and signs
44: of depolarization. Magnetic field strengths of 25 $\mu$G, organized
45: on scales of $\sim$1 kpc, and intermixed with gas at a temperature of
46: 5 $\times 10^6$ K with a density of $\sim$0.1 cm$^{-3}$ can reproduce
47: the observed RM excess, the depolarization, and the observed X-ray
48: surface brightness. This hot gas may be in pressure equilibrium with
49: the optical line-emitting gas, but the magnetic field strength of 25
50: $\mu$G associated with the hot gas provides only 10\% of the thermal
51: pressure and is therefore insufficient to account for the stability of
52: the line-emitting filaments.
53:
54: \end{abstract}
55:
56: \begin{keywords}
57: galaxies: magnetic fields -- galaxies: active -- galaxies: nuclei --
58: galaxies individual: \pks -- radio continuum: galaxies -- X-rays: galaxies: clusters
59: \end{keywords}
60:
61: \section{Introduction}
62:
63: The Centaurus cluster, Abell 3526, is a nearby (redshift z=0.0104),
64: X-ray bright galaxy cluster. At the center of this cluster is the
65: bright elliptical galaxy NGC 4696, hosting the moderately powerful radio
66: source, \pks. We have been engaged in detailed studies of the X-ray
67: and radio emission from this cluster (Sanders \& Fabian 2002, Taylor,
68: Fabian \& Allen 2002), and have
69: recently obtained a further 200 ksec of Chandra data (Fabian et al. 2005).
70: The new Chandra image reveals a complex structure
71: within the central few kiloparsecs. A plume-like structure swirls
72: clockwise to the NE and wraps around the radio source. There are
73: clear signs of interaction between the X-ray and radio emission
74: including: (1) strong X-ray emission that matches the shape of the
75: radio emission just south of the core; (2) a faint rim of hard X-ray
76: emission along the northern edge of the radio source; and (3) deep
77: cavities in the X-ray emission on both the east and west sides. We
78: also find a compact X-ray component coincident with the radio and
79: optical core (Taylor et al. 2006).
80:
81: \begin{figure*}
82: \centerline{\psfig{figure=fig1.eps,width=1.05\textwidth} }
83: %\vskip12cm
84: %\special{psfile=fig1.eps hoffset=-50 voffset=-100 hscale=60.0 vscale=60.0}
85: \caption{The rotation measure (RM) map for PKS\,1246{\tt -}410
86: with contours from the 5 GHz total intensity image overlaid.
87: Contours start at 0.15 mJy/beam and increase by factors of 2.
88: The angular resolution is $1.2 \times 0.4$ arcsec.
89: \label{fig1}}
90: \end{figure*}
91:
92: \begin{figure*}
93: \centerline{\psfig{figure=fig2.eps,width=1.05\textwidth} }
94: %\vskip12cm
95: %\special{psfile=fig2.eps hoffset=-50 voffset=-100 hscale=60.0 vscale=60.0}
96: \caption{The rotation measure (RM) map for PKS\,1246{\tt -}410
97: with contours from the unsharp-masked H$\alpha$ image
98: overlaid. The color bar ranges from $-$3500 to 3500 \radm.
99: Relative contours are drawn starting at 2.5$\sigma$ and
100: increasing by factors of 4 (Crawford et al. 2005).
101: \label{fig2}}
102: \end{figure*}
103:
104: From VLA data primarily in BnA configuration, Taylor et al. (2002)
105: imaged the rotation
106: measure (RM) distribution across \pks\ at a resolution of 2.1 $\times$
107: 1.2 arcsecond. The RMs appear patchy and are
108: typically in the range $-$500 to +500 \radm, with a maximum absolute value of
109: $\sim$1800 \radm. Assuming that the RMs are produced by cluster
110: magnetic fields as found for other galaxies at the center of dense
111: clusters (see review by Carilli \& Taylor 2002), simple magnetic field
112: topologies lead to estimates of the magnetic field strength of $\sim$11
113: $\mu$G. The Centaurus cluster is exceptional in that the radio source
114: \pks\ provides a well resolved probe of the central 10 kpc of a
115: cooling core cluster.
116:
117: Taylor et al. (2002) suggested that some regions of higher RM lay
118: behind regions of high X-ray surface brightness. This raised the
119: possibility that (1) some of the bright inner X-ray emission is
120: located in front of the radio source; and (2) this thermal gas within
121: a few kiloparsecs of the radio could contribute significantly to the RM.
122: Rudnick \& Blundell (2003) used the observations of Taylor
123: et al. to investigate the possibility of a local cocoon producing the
124: high RMs, and claimed to find evidence for this in a correlation
125: between the intrinsic polarization angles and the RM. However,
126: Ensslin et al. (2003) showed that the relationship was an artifact of
127: the analysis used, and a better formed statistic showed no correlation.
128: Here, we present improved observations of the RM structure in
129: \pks\ and further explore the evidence for local interactions through
130: detailed comparison with X-ray and H$\alpha$ images.
131:
132: The arc-shaped structures seen in the H$\alpha$ filaments appear
133: quite smooth with no indication on subkiloparsec scales of
134: surrounding turbulence (Crawford et al. 2005). This smoothness,
135: and the commonality of structure between optical filaments, X-ray
136: arcs, and the dust lanes suggest that these could all have been
137: shaped by strong magnetic fields. An alternative explanation
138: could be a bulk laminar flow within the intracluster medium.
139: Detailed comparisons between the Faraday rotation measure and
140: the optical filaments should allow us to measure the strength
141: of the magnetic fields in the region of the filaments and
142: thereby discriminate between these two possibilities.
143:
144: Here we present VLA (Very Large Array)\footnote{The National Radio
145: Astronomy Observatory is a facility of the National Science Foundation
146: operated under a cooperative agreement by Associated Universities,
147: Inc.} observations of the radio galaxy PKS\,1246{\tt -}410 at the
148: centre of the Centaurus cluster.
149: In \S 2 we describe the radio observations. In \S 3 we
150: summarize the H$\alpha$ emission, and in \S 4 we summarize
151: the X-ray observations. In \S 5 we discuss the radio properties
152: of \pks\ and in \S 6 we compare the multi-wavelength
153: observations and consider the origin of the observed RMs.
154:
155:
156: We assume $H_0$ = 71 km s$^{-1}$ Mpc$^{-1}$ and $\Omega$
157: = 1 so that 1\arcsec = 210 pc at the redshift (0.0104) of
158: the Centaurus cluster.
159:
160: \section{Radio Observations and Data Reduction}
161:
162: Observations of PKS\,1246{\tt -}410 were made at 4.635,
163: 4.885, 8.115, and 8.485 \,GHz at the VLA in its A configuration
164: on 2006 February 8. The
165: source 3C\,286 was used as the primary flux density calibrator, and as
166: an absolute reference for the electric vector polarization angle
167: (EVPA). Phase calibration was derived from the nearby compact source
168: J1316{\tt -}3338 with a cycle time between calibrators of about 16 minutes.
169: Instrumental calibration of the polarization leakage terms was
170: obtained using OQ\,208 which was observed over a wide range in
171: parallactic angle. The data were reduced in AIPS (Astronomical Image
172: Processing System), following the standard procedures. The AIPS task
173: IMAGR was used with a suitable taper to make Stokes $I$, $Q$ and $U$ images at each
174: of the 4 frequencies observed at the same resolution. Polarized
175: intensity, $P$, images and polarization angle, $\chi$, images were
176: derived from the $Q$ and $U$ images. The Faraday Rotation Measure (RM)
177: image was made from the combination of the $\chi$ images.
178: Pixels in the RM image were flagged when the error in any of the
179: $\chi$ images exceeded 20 degrees.
180:
181:
182: \section{H$\alpha$ Observations}
183:
184: A detailed study of the extended H$\alpha$ filaments surrounding
185: \pks\ has recently been published by Crawford et al. (2005). They find
186: a network of line-emitting filaments extending over the central 50
187: arcsec (10.5 kpc). These filaments correlate with
188: arc-like plumes seen in soft X-ray emission. Crawford
189: et al. speculate that both the H$\alpha$ filaments and the
190: plumes of soft X-ray emitting gas have been drawn out of the
191: center by buoyant bubbles created by the central radio
192: galaxy, \pks.
193:
194: \section{X-ray Observations}
195:
196: Fabian et al. (2005) presented deep (199 ksec) observations
197: of the Centaurus cluster, that showed well defined cavities
198: coincident with \pks. Further evidence of outbursts from
199: the radio galaxy were seen in semi-circular edges to the
200: east and west marked by sharp temperature increases.
201:
202: The electron density derived at the inner 70 kpc (4.2\arcmin)
203: radius region, determined from a deprojection analysis, is
204: shown in Taylor et al. (2002). The best-fitting $\beta-$model for the
205: density has parameter values $n_0 = 0.099\pm0.001$ cm$^{-3}$, $\beta =
206: 0.39\pm0.01$, and a core radius, $r_c = 5.4\pm0.3$ kpc. Sanders
207: \& Fabian (2006) find a central temperature of $\sim$2 $\times 10^7$
208: K in the lobes.
209: Note that
210: the core size is similar to that of the radio source, which has made
211: `holes' in the X-ray emission at radii of 1--5~kpc.
212:
213:
214: \section{Radio Properties of \pks}
215:
216: The radio source PKS\,1246{\tt -}410 is associated with NGC~4696, an
217: elliptical galaxy at the center of the Centaurus cluster (also known
218: as Abell 3526). The radio power of PKS\,1246{\tt -}410 is 1.5 $\times
219: 10^{24}$ W Hz$^{-1}$ at 1.6 GHz, identifying it as a low power
220: radio galaxy for which we would expect an FR~I morphology (edge dimmed
221: extended radio tails). In fact the radio morphology at 5 GHz,
222: shown in Fig.~1, appears quite different from most low power radio
223: galaxies (Parma et al. 1987).
224: The usual well defined core and thin jets feeding into larger lobes
225: are missing. Instead there is a bright point source which we identify
226: as the core based on its central location and slightly flatter
227: spectral index, $\alpha = -0.5$ where $S_\nu \propto \nu^\alpha$
228: (Taylor et al. 2002). The properties of the core were further
229: examined by Taylor et al. (2006) who found it to be radiatively
230: underluminous but powering a jet (detected by the VLBA) that
231: carries off enough energy to inflate the cavities seen in the
232: thermal gas.
233:
234: The bright radio lobes of PKS\,1246{\tt -}410 are moderately well
235: polarized (4 -- 40\%), so it is possible to derive the RM distribution
236: across nearly the entire radio source. This was done by Taylor et
237: al. (2002) at a resolution of 2.1 $\times$ 1.2 arcseconds. The RMs
238: appear patchy and are not correlated with any features in total
239: intensity.
240:
241: We combine our new VLA-A configuration observations at 5 and 8.4 GHz
242: with the previous observations.
243: Our higher angular resolution (1.2 $\times$ 0.4 arcsec) images reveal some fine-scale structure
244: in the Faraday rotation measure image (Fig.~1). In general the
245: western lobe appears to have smoother RMs with typical values of
246: $-$1000 \radm\ over a 5\arcsec\ region. Much of the eastern lobe
247: is patchy with a scale size of $\sim$2\arcsec (0.4 kpc). Typical RM gradients
248: are $\sim$400 \radm/arcsec. Exceptionally large RM gradients are found
249: in the regions spatially coincident with H$\alpha$ filaments. These
250: regions are up to 2 arcseconds (0.4 kpc) wide and 5 arcsec (1 kpc) long.
251:
252: Very little polarization is detected after the radio lobes
253: bend toward the south, but this is most likely due to a lack of
254: sensitivity since the lobes fade at 8.4 GHz to 100-300 $\mu$Jy/beam
255: and below. Given our noise level of $\sim$25 $\mu$Jy we would only
256: detect polarization (at the 2 $\sigma$ level) if it was as high as
257: 20-50\%. There is a hint of an anti-correlation with line emission
258: across the southwest lobe, but more sensitive observations will be
259: required to confirm low fractional polarization there.
260:
261:
262:
263: \section{Contributors to the Faraday Rotation}
264:
265: We can rule out uniform internal Faraday rotation since it deviates
266: from a wavelength-squared law for changes in angle of more
267: than 90 degrees (Burn 1966). Internal Faraday rotation also predicts
268: significant depolarization of the radio
269: source since emission produced at the back interferes with emission
270: produced at the front. The RM of 2000 \radm\ seen in \pks\ corresponds
271: to a change in angle of 400 degrees at 5 GHz.
272:
273: Faraday rotation from an external screen is thus likely and could have
274: contributions from (1) optical line-emitting filaments crossing in
275: front of the radio lobes; (2) the filaments of dense soft X-ray
276: emitting gas on 10-kpc scales; (3) a mixing layer between source
277: fields and the surrounding medium (Rudnick \& Blundell 2003); and
278: (4) the 100-kpc scale cluster gas (Taylor et al. 2002).
279: The latter two explanations do not readily explain fine scale
280: RM structures correlated with the H$\alpha$ filaments so we
281: do not consider them further here.
282:
283: For external Faraday rotation, the RMs are related to the density, $n_{\rm
284: e}$, and magnetic field along the line-of-sight, $B_{\|}$, through
285: the cluster (or line-emitting gas) according to
286: $$ RM = 812\int\limits_0^L n_{\rm e} B_{\|} {\rm d}l ~{\rm
287: radians~m}^{-2}~,
288: \eqno(1)
289: $$
290: where $B_{\|}$ is measured in $\mu$Gauss, $n_{\rm e}$
291: in cm$^{-3}$ ~and d$l$ in kpc.
292:
293: \begin{figure}
294: \centerline{\psfig{figure=fig3.eps,width=0.65\textwidth,angle=-90} }
295: \caption{Fits to the electric vector position angle as a function
296: of wavelength squared in the western filament. The slope of the
297: fit provides the RM and is indicated for each pixel. The transition
298: from $-$2000 \radm\ to $+$2000 \radm occurs over just 0.6\arcsec.
299: \label{fig3}}
300: \end{figure}
301:
302: \begin{figure*}
303: \centerline{\psfig{figure=fig4.eps,width=1.0\textwidth} }
304: \caption{Absolute rotation measure distributions for selected areas in
305: PKS\,1246{\tt -}410. The top three regions were selected by taking
306: a random starting position in the east lobe and then spacing over
307: at regular intervals of 8.8 arcseconds.
308: The bottom two regions correspond to the eastern and western
309: filaments. The width of each bin is
310: 40 \radm, which is about the same as the error in the RM determination
311: for any given point. The number of independent beams is: 11.1, 10.5, and 9.9
312: respectively for the 3 regions plotted on the top row; and
313: 6.1 and 10.7 for the east and west filament, respectively.
314: \label{fig4}}
315: \end{figure*}
316:
317: \subsection{H$\alpha$/Radio Comparisons}
318:
319: Depolarization of the radio source coincident with the line-emitting
320: gas has been seen before (e.g., A1795; Ge \& Owen 1993), and claimed
321: to be the result of large RM gradients within the synthesised beam.
322: Even modest magnetic field strengths, combined with the high thermal
323: densities of the line-emitting gas, can produce large RM gradients.
324: In Fig.~2 we show overlays of the unsharp-masked H$\alpha$ and the RM
325: map. Inspection of Fig.~2 reveals that RM image has values near the
326: extreme ranges of the image and fewer measurements (presumably due to
327: depolarization of the radio source) where the H$\alpha$ filaments
328: cross the lobes. We find values ranging from $-$2000 \radm\ to
329: $+$2000 \radm\ on scales of $\sim$ 0.4\arcsec (the maximum resolution
330: of our observations). These correspond to gradients of 10$^4$
331: \radm/arcsec.
332:
333:
334: A concern about the measurements of the RMs in the filaments is that
335: they are in regions of relatively low SNR. To make sure that the high
336: RMs are not due to low SNR, we show the RM fits in the west filament
337: in Fig.~3. The EVPAs are well fit by a $\lambda^2$-law, and have
338: reasonable reduced $\chi^2$ values of $\sim$1, with a few fits as low
339: as 0.1 and as high as 10. To further test the possibility that the
340: RMs are higher in association with the filaments we compare the
341: absolute RM distribution in three regularly spaced regions with a
342: similar area, with those just coincident with the filaments (Fig.~4).
343: The average $|RM|$ in the east and west filaments is 1224 $\pm$ 342
344: \radm\ and 766 $\pm$ 230 \radm, compared to averages of below 500
345: \radm\ for the three typical regions. We performed a Kolmogorov-Smirnov
346: test (KS-test) to see how significant the difference is, and found
347: that the three typical regions have a $> 99.9\%$ probability of being
348: drawn from a different population compared to the filament regions.
349: We also checked for any correlation between low SNR and high RM. In
350: general regions of low SNR are equally likely to have any RM, and are
351: not biased to large absolute values. We conclude that there is a real
352: association between the H$\alpha$ emission and the regions of enhanced
353: RM.
354:
355: The surface brightness of the filaments in H$\alpha + $[NII] is 1-3
356: $\times$ 10$^{-16}$ erg cm$^{-2}$ s$^{-1}$ arcsec$^{-2}$ (Crawford et
357: al.\ 2005). If we take $2 \times 10^{-16}$ erg cm$^{-2}$ s$^{-1}$
358: arcsec$^{-2}$ for H$\alpha$ in the brightest depolarizing part, 5
359: arcsec ($\sim$1 kpc) west of the nucleus, and adopt a product $nT = 2
360: \times 10^6$ cm$^{-3}$ K for the surrounding gas pressure (taking
361: values from Allen et al. 2006), we find that the depth of the
362: H$\alpha$ emitting region (T $\sim$ 10$^4$K) is 0.01 pc. This assumes
363: a uniform covering fraction. The line-emitting gas is much more
364: likely to be in the form of a network of small filaments, in which
365: case the covering fraction is smaller and the depth proportionately
366: larger, realistically $\sim$1 pc. If the magnetic fields are
367: organized on similar scales then the many field reversals within the
368: synthesized beam (dimensions 252 $\times$ 84 parsec), would scramble
369: the intrinsic polarization vectors, and lead to depolarization. Even
370: if the magnetic fields are for some reason organized on kiloparsec
371: scales such a network of filaments could not provide the observed high
372: RM values as they would still produce larger RM gradients than we
373: observe and would completely depolarize the radio emission due to
374: variations within the synthesized beam. Therefore, even though
375: the H$\alpha$ emission is coincident with regions of enhanced
376: RM, it does not have the proper physical characteristics to
377: produce coherent RMs on the observed scales.
378:
379:
380:
381:
382: \begin{figure*}
383: \centerline{\psfig{figure=fig5.eps,width=0.95\textwidth} }
384: \caption{Overlay of the Faraday RM on the X-ray temperature map.
385: The innermost (highest) contours correspond to temperatures of
386: 0.5 keV and the lowest contours to 0.9 keV. The temperatures over
387: most of the upper part of the radio lobes are around 1.5~keV.
388: \label{fig5}}
389: \end{figure*}
390:
391: \subsection{Soft X-ray/Radio Comparisons}
392:
393: As pointed out by Crawford et al. (2005) the soft X-ray emitting gas
394: is coincident with the optical line filaments. In Fig.~5 we overlay
395: the X-ray temperature map from Crawford et al. on our Faraday RM
396: image. A clear correspondence can be seen between low temperature gas
397: and the regions of the east and west filaments
398: previously noted to have large RM gradients.
399:
400: A contribution to the RM could come from the dense soft X-ray emitting
401: gas, of density 0.1 cm$^{-3}$ (Allen et al 2006). This gas has a
402: covering fraction closer to unity and is much more volume filling than
403: the line emitting gas outside the radio bubble. Then, if its depth is
404: about 1 kpc, the magnetic field required to produce the RMs observed
405: of $\sim$2000
406: \radm\ is about 25 $\mu$G. The magnetic pressure is then 2 $\times
407: 10^5$ cm$^{-3}$ K, or about 10 per cent of the thermal pressure.
408: These values seem reasonable and given the good correspondence with
409: the soft X-ray emitting gas we believe this is the most likely cause
410: for the excess RMs and RM gradients observed.
411:
412: A simple model where tangled magnetic flux is conserved as gas cools
413: at constant pressure would predict that $B \propto T^{-2/3}$. This is
414: in rough agreement with the increase in $B$ from $\sim 11$ to $25\mu$G
415: expected as the temperature drops from $1.5$ to $\sim 0.6$ keV. If
416: this scaling continues to lower temperature the gas at $\sim0.1$ keV
417: ($10^6$ K)
418: would have a magnetic pressure similar to the thermal pressure. The
419: optical filaments may well have significant magnetic pressure, likely
420: needed for maintaining their integrity against shredding by random
421: motions and conduction. Our observations are however unable to comment
422: directly on them due to their very small scale (section 6.1).
423:
424: %\subsection{Cluster X-ray/Radio Comparisons}
425: %
426: %For completeness we consider the large scale X-ray emitting gas
427: %from the Centaurus cluster. This gas likely contributes to the
428: %overall RM of the lobes as was discussed in Taylor et al. (2002)
429: %and leads to estimates for the cluster magnetic field of 11 $\mu$G
430: %for a core radius of 5.4 kpc.
431: %Given the large scales involved it is difficult for the cluster-wide
432: %gas to produce the large RM gradients and consequent depolarization
433: %coincident with the optical filaments.
434:
435:
436: %\renewcommand{\baselinestretch}{1.2}
437: \def\dg{$^{\circ}$}
438: %\begin{center}
439:
440: %TABLE 1 \\
441: %\smallskip
442: %\smallskip
443: \begin{table}
444: \scriptsize
445: \caption{O{\sc bservational} P{\sc arameters}}
446: \begin{tabular}{llrrrr}
447: \hline
448: \hline
449: Source & Date & Frequency & Bandwidth$^a$ & Config. & Duration \\
450: & & (GHz) & (MHz) & & (hours) \\
451: \hline
452: PKS\,1246 & Apr1998 & 1465/1565 & 25 & A & 0.24 \\
453: {\tt -}410 & Jun1998 & 1465/1565 & 25 & BnA & 0.99 \\
454: & Apr1998 & 4635/4885 & 50 & A & 0.09 \\
455: & Jun1998 & 4635/4885 & 50 & BnA & 1.23 \\
456: & Nov1998 & 4635/4885 & 50 & CnB & 0.61 \\
457: & Jun1998 & 8115/8485 & 50 & BnA & 0.89 \\
458: & Nov1998 & 8115/8485 & 50 & CnB & 0.62 \\
459: \hline
460: \end{tabular}
461: \end{table}
462: %\twocolumn
463:
464:
465: \section{Conclusions}
466:
467: New VLA observations of the radio galaxy \pks, providing
468: high angular and spatial resolution,
469: afford us with a rare case where we can sample the RM
470: structure continuously across the core of a cluster.
471: We find distinct
472: regions, of width $\sim$400 pc, where the RM and RM gradients, are
473: enhanced, and the fractional polarization is reduced. This
474: enhancement in the RM most likely originates in dense, soft X-ray
475: emitting gas in front of the radio lobe. The magnetic fields required
476: to produce the observed enhancement are 25 $\mu$G, similar to the
477: equipartition estimates for the magnetic fields interior to the lobes,
478: and a few times larger than the ambient cluster magnetic fields. If
479: the magnetic fields continues to increase as the gas cools according
480: to $B \propto T^{-2/3}$ then the fields may play an important role in
481: shaping the optical filaments. To confirm this scenario would require
482: observations with at least an order of magnitude higher angular
483: resolution, and a hundredfold increase in sensitivity, such as could
484: be provided by the proposed Square Kilometer Array. Observations of
485: the RM structure in other clusters can be obtained now with the VLA
486: and some effort, and will soon be made much easier by the advent of
487: the EVLA. These radio observations can be compared to sensitive X-ray
488: images to look for similar correlations between RM enhancements and
489: soft X-ray emission.
490:
491: \section{Acknowledgments}
492:
493: We thank an anonymous referee for constructive suggestions.
494: GBT acknowledges support for this work from the National Aeronautics
495: and Space Administration through Chandra Award Numbers GO4-5134X and
496: GO4-5135X issued by the Chandra X-ray Observatory Center, which is
497: operated by the Smithsonian Astrophysical Observatory for and on
498: behalf of the National Aeronautics and Space Administration under
499: contract NAS8-03060. ACF thanks The Royal Society for support.
500: This research has made use of the NASA/IPAC
501: Extragalactic Database (NED) which is operated by the Jet Propulsion
502: Laboratory, Caltech, under contract with NASA. The National Radio
503: Astronomy Observatory is a facility of the National Science Foundation
504: operated under a cooperative agreement by Associated Universities,
505: Inc.
506:
507: \bigskip
508:
509:
510: \begin{thebibliography}{}
511:
512: \bibitem []{all01} Allen, S.W., Fabian, A.C., Johnstone, R.M., Arnaud, K.A.,
513: Nulsen, P.E.J.\ 2001, MNRAS, 322, 589
514: \bibitem[Allen et al.(2006)]{all06} Allen, S.~W., Dunn,
515: R.~J.~H., Fabian, A.~C., Taylor, G.~B., \& Reynolds, C.~S.\ 2006, MNRAS,
516: 372, 21
517: \bibitem []{bur66} Burn, B.F.\ 1966, MNRAS, 133, 67
518: \bibitem []{car02} Carilli, C.L., \& Taylor, G.B.\ 2002, ARA\&A, 40, 319
519: \bibitem[Crawford et al.(2005)]{cra05} Crawford, C.~S.,
520: Hatch, N.~A., Fabian, A.~C., \& Sanders, J.~S.\ 2005, MNRAS, 363, 216
521: \bibitem[Ensslin et al.(2003)]{2003ApJ...597..870E} Ensslin, T.~A., Vogt,
522: C., Clarke, T.~E., \& Taylor, G.~B.\ 2003, ApJ, 597, 870
523: \bibitem[Fabian et al.(2005)]{fab05} Fabian, A.~C., Sanders,
524: J.~S., Taylor, G.~B., \& Allen, S.~W.\ 2005, MNRAS, 360, L20
525: \bibitem []{ge93} Ge, J.-P., Owen, F.N. 1993, AJ, 105, 778
526: \bibitem{} Ikebe Y., Makishima K., Fukazawa Y., Tamura T., Xu H.,
527: Ohashi T., Matsushita K., 1999, ApJ, 525, 58
528: \bibitem []{1987A&A...181..244P} Parma, P., de Ruiter,
529: H.~R., Fanti, C., Fanti, R., \& Morganti, R.\ 1987, A\&A, 181, 244
530: \bibitem[Rudnick \& Blundell(2003)]{rud03} Rudnick, L., \&
531: Blundell, K.~M.\ 2003, ApJ, 588, 143
532: \bibitem []{san01} Sanders, J.S., \& Fabian, A.C.\ 2002, MNRAS, 331, 273
533: \bibitem[Sanders \& Fabian(2006)]{2006MNRAS.371.1483S} Sanders, J.~S., \&
534: Fabian, A.~C.\ 2006, MNRAS, 371, 1483
535: \bibitem[Taylor et al.(2002)]{tay02} Taylor, G.~B., Fabian,
536: A.~C., \& Allen, S.~W.\ 2002, MNRAS, 334, 769 \bibitem[Taylor et al.(2006)]{tay06} Taylor, G.~B., Sanders,
537: J.~S., Fabian, A.~C., \& Allen, S.~W.\ 2006, MNRAS, 365, 705
538:
539:
540: \end{thebibliography}
541:
542: \end{document}
543: