1: \documentclass{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: \slugcomment{To appear in The Astrophysical Journal}
4:
5: \shortauthors{Halpern et al.}
6: \shorttitle{Search for Pulsations from 3EG J1835+5918}
7:
8: \begin{document}
9:
10: %
11: % Symbols
12: %
13: \def\xray{RX~J1836.2+5925}
14: \def\source{3EG J1835+5918}
15: \def\ro{{\it ROSAT\/}}
16: \def\chandra{{\it Chandra\/}}
17: \def\glast{{\it GLAST\/}}
18: \def\hst{{\it Hubble Space Telescope}}
19:
20: \title{The Next Geminga: Search for Radio and X-ray Pulsations\\
21: from the Neutron Star Identified with 3EG J1835+5918}
22:
23: \author{J. P. Halpern, F. Camilo, and E. V. Gotthelf}
24: \affil{Columbia Astrophysics Laboratory, Columbia University,
25: New York, NY 10027}
26:
27:
28: \begin{abstract}
29: We report unsuccessful searches for pulsations
30: from the neutron star \xray\ identified
31: with the EGRET source \source.
32: A 24-hr observation with the NRAO Green Bank Telescope at 820 MHz
33: placed an upper limit on flux density of 17 $\mu$Jy for $P\ga 10$~ms,
34: and gradually increasing limits for $1\la P\la 10$~ms. The equivalent
35: luminosity is lower than that of any known pulsar with the possible
36: exception of the radio-quiet $\gamma$-ray pulsar Geminga.
37: A set of observations with the {\it Chandra X-ray Observatory}
38: HRC totaling 118~ks revealed no pulsar with 1~ms~$\leq P \leq 10$~s.
39: The upper limit on its pulsed fraction is 35\% assuming a sinusoidal
40: pulse shape. The position of \xray\
41: in \chandra\ observations separated by 3~years is unchanged within errors,
42: leading to an upper limit on its proper motion of
43: $<0.14^{\prime\prime}$~yr$^{-1}$, or $v_t<530$ km~s$^{-1}$ at $d = 800$~pc,
44: a maximum distance estimated from its thermal X-ray spectrum.
45: With these null results, the properties of \source\ and its
46: X-ray counterpart \xray\ are consistent with a more distant or older
47: version of Geminga, or perhaps a recycled pulsar. Having nearly
48: exhausted the capabilities of current instrumentation at all wavelengths,
49: it will likely fall to the {\it Gamma-ray Large Area Space Telescope}
50: to discover pulsations from \source.
51: \end{abstract}
52:
53: \keywords{gamma rays: observations --- stars: neutron ---
54: X-rays: individual (\xray)}
55:
56: \section{Introduction}
57:
58: The brightest of the ``unidentified'' high-Galactic-latitude EGRET sources
59: \citep{har99},
60: \source\ at $(\ell, b)=(89^{\circ},+25^{\circ})$ has long been
61: associated with the X-ray emitting neutron-star \xray, which
62: remains its most plausible counterpart even though pulsations
63: have not been detected at any wavelength. Unlike blazars,
64: which are highly variable and have steeper $\gamma$-ray spectra,
65: \source\ shows no evidence for long-term variability,
66: and its spectrum can be fitted by a relatively flat power law of
67: photon index $\Gamma = 1.7$ from 70~MeV to
68: 4~GeV, with a turn-down above 4 GeV \citep{rei01}, similar to
69: known $\gamma$-ray pulsars.
70: \xray\ is the only optically undetected X-ray source in the
71: $\gamma$-ray error box, all the others being classified as unlikely
72: $\gamma$-ray emitters \citep{mir00}. The detection of \xray\ as
73: a weak, ultrasoft source in the \ro\ All-Sky Survey \citep{mir01}
74: suggested that it is a thermally emitting neutron star that is either
75: older or more distant than the $\gamma$-ray pulsar Geminga
76: \citep{hal93,big96}.
77:
78: Observations using the \chandra\ Advanced CCD Imaging Camera (ACIS)
79: and the \hst\ ({\it HST\/})
80: further supported this interpretation \citep{hal02}.
81: Two components were required to fit the \chandra\ X-ray
82: spectrum, a blackbody of $T \sim 3.5 \times 10^5$~K
83: and a power law of photon index $\Gamma = 2.2 \pm 0.6$. The non-thermal
84: X-ray component is important evidence of magnetospheric activity that is
85: found in all $\gamma$-ray pulsars, but not in all cooling neutron stars.
86: An optical upper limit of $V > 28.5$ from the {\it HST\/}
87: verified the neutron-star
88: interpretation of \xray\ from its extreme X-ray-to-optical flux ratio,
89: $f_X/f_V > 6000$. We also used the thermal X-ray
90: spectrum and optical limit to bound the
91: distance to \xray\ in the range $250 < d < 800$~pc.y
92: The energetics of \source\ are plausible for a pulsar
93: at $d < 800$~pc since its $\gamma$-ray luminosity
94: (assumed isotropic) is $3.8 \times 10^{34}\,(d/800\,{\rm pc})^2$ ergs~s$^{-1}$,
95: comparable to the spin-down power $\dot E = -I\Omega \dot\Omega$ of Geminga
96: ($3.3 \times 10^{34}$ ergs~s$^{-1}$). Efficiencies approaching 100\%
97: are achieved by those $\gamma$-ray pulsars having the smallest
98: spin-down power.
99:
100: \citet{hal02} also searched for radio pulsations from \xray\
101: using the Jodrell Bank Lovell telescope at a frequency
102: of 1.4~GHz, achieving an upper limit of $S_{1.4} < 0.1$~mJy.
103: Adopting an upper limit of $d < 800$~pc,
104: the pulsed pseudo-luminosity limit of \xray\ is
105: $L_{1.4} \equiv S_{1.4} d^2 < 0.064$~mJy~kpc$^2$.
106: There are only four pulsars known with $L_{1.4} < 0.1$~mJy~kpc$^2$
107: \citep{cam02} in addition to Geminga
108: \citep[][and references therein]{mcl99}.
109:
110: In this paper, we report an even more sensitive search for radio pulsations
111: from \xray\ using the NRAO Green Bank Telescope (GBT),
112: as well as the first X-ray pulsation search with the \chandra\ High
113: Resolution Camera (HRC-S). These investigations approach the sensitivity
114: limits of existing instrumentation to search for the pulsar in \source.
115:
116: \section{GBT Radio Pulsar Search}
117:
118: \xray\ is circumpolar at the GBT, where on 2002 December
119: 6 we observed it for 24\,hr using the BCPM spectrometer \citep{bac97}.
120: Due to the intermediate latitude of the target, with a
121: relatively small predicted dispersion measure (DM) and low Galactic
122: background temperature, we observed at a central frequency of 820\,MHz,
123: with a bandwidth of 48\,MHz in each of two polarizations. The signals
124: from corresponding polarization channels were summed in hardware, and
125: the total-power samples from each of 96 frequency channels were sampled
126: every $72\,\mu$s and recorded to disk for off-line analysis.
127:
128: We analyzed the data using standard pulsar search techniques, implemented
129: in the PRESTO software package \citep{ran02}. First we identified
130: and excised radio-frequency interference. We then dedispersed the data in
131: the DM range 0--110~cm$^{-3}$~pc, with resolution of 0.3~cm$^{-3}$~pc.
132: This is twice the maximum Galactic DM predicted for this line of sight by
133: the \citet{cor02} electron density model; for the distance range
134: of 250--800~pc, within which the neutron star likely lies (\S 1), the
135: predicted DM range is 2--9~cm$^{-3}$~pc. Each time series contained 1.2
136: billion points, and would have been needlessly challenging to analyze.
137: We therefore down-sampled the data to a resolution of 0.288~ms,
138: resulting in much more reasonable time series of 300 million points.
139: For a DM of 40~cm$^{-3}$~pc, the smearing caused by dispersion across
140: one of the frequency channels is 0.3~ms, so that for smaller DMs the
141: effective time resolution of the search was about 0.3~ms.
142:
143: We performed
144: the periodicity search using fast-Fourier techniques, and were sensitive
145: to periods $P \ga 1$~ms. In order to improve sensitivity for a range of
146: pulse duty cycles, we incoherently summed up to 16 harmonics of the data;
147: for $P \ga 10$~ms, we maintained good sensitivity for duty cycles down
148: to about $0.03P$. The sensitivity at long periods decreases gradually
149: due to red noise, but this is not a significant concern for $P \la 5$~s.
150: Unfortunately, we found no convincing pulsar candidates in this search.
151:
152: Using the standard modification to the radiometer equation, assuming
153: a pulse shape with 10\% duty cycle, we were sensitive to pulsars with
154: $P \ga 10$\,ms having a period-averaged flux density at 820\,MHz $\ga
155: 17\,\mu$Jy, with gradually decreasing sensitivity for smaller periods.
156: Converting this to the more usual 1.4\,GHz pulsar search frequency
157: with a typical spectral index of --1.6 \citep{lor95} gives
158: $S_{1.4} \la 7\ \mu$Jy. For a distance of 800~pc, this corresponds
159: $L_{1.4} \la 5\ \mu$Jy~kpc$^2$.
160: All known radio pulsars have a greater $L_{1.4}$ \citep{cam02}.
161:
162: Based on these results, it is therefore unlikely that \xray\
163: is a radio pulsar beaming toward the Earth. More likely, it is either
164: beaming away, or is not an active radio pulsar at all.
165:
166: \section{Chandra HRC Pulsar Search}
167:
168: \begin{deluxetable}{llrc}
169: %\tablewidth{210pt}
170: \tablecaption{Chandra HRC Observations of \xray }
171: \tablehead{
172: \colhead{ObsID} & \colhead{Start Time} & \colhead{Live Time} & Counts\\
173: \colhead{} & \colhead{(UT)} & \colhead{(s)} &
174: }
175: \startdata
176: 6182 & 2005 Feb 28 19:27 & 45,018 & 284\\
177: 4606 & 2005 Mar 9 15:30 & 28,030 & 189\\
178: 6183 & 2005 Mar 11 08:19 & 44,999 & 256\\
179: \hline
180: Total & & 118,047 & 729
181: \enddata
182: \label{logtable}
183: \end{deluxetable}
184:
185: Three observations of \xray\ totaling 118~ks of live time
186: were performed between 2005 February 28
187: and 2005 March 11 using the \chandra\ HRC-S in timing mode, which
188: has time resolution of $16\ \mu$s, and provides absolute accuracy of
189: $42\ \mu$s as determined from observations of the Crab
190: pulsar\footnote{http://cxc.harvard.edu/cal/}.
191: Table~\ref{logtable} lists the basic parameters of the observations,
192: including the number of photons extracted from a
193: $1.\!^{\prime\prime}2$ radius aperture.
194: The live time is $99.6\%$ of the elapsed time, limited by event processing.
195: The total number of photons detected from \xray\
196: is 729, of which 20 are estimated to be background (using nearby,
197: source-free regions of the image). The count rates from
198: the three observations are within $9\%$ or less of the average
199: count rate, $0.006$~s$^{-1}$,
200: consistent with no variation. The spatial
201: distribution of photons is consistent with a point source.
202: Originally conceived as a contiguous, 135~ks
203: pointing, the observation could not be executed as planned because of
204: new operational restrictions that came into effect.
205: The resulting large gaps
206: in time diminish the sensitivity to pulsations
207: because the increased number of independent frequencies
208: and frequency derivatives that have to be searched
209: yield more false candidates.
210:
211: The total time span of the observations
212: from beginning to end is $T=9.6 \times 10^5$~s,
213: which determines the independent intervals for
214: frequency and frequency derivative, $\Delta f = 1/2T$
215: and $\Delta \dot f = 1/T^2$, respectively. We chose
216: a frequency step size $\delta f = 2 \times 10^{-7}$~Hz
217: to oversample the independent frequency interval by a factor
218: of $\approx 2.5$, and $\delta \dot f = 4 \times 10^{-13}$ Hz~s$^{-1}$
219: to achieve a similar oversampling factor for frequency derivative.
220:
221: \begin{figure}
222: \centerline{
223: \includegraphics[width=3.3in,angle=0.]{f1.eps}
224: }
225: \caption{The parameter space ($f,\dot f$) searched
226: for X-ray pulsations from \xray\ is shaded.
227: The step size $\delta \dot f = 4 \times 10^{-13}$
228: Hz s$^{-1}$, and the search includes $\dot f = 0$.
229: Lines of constant $\dot E$ and $\tau_c$ are shown.
230: The star indicates the parameters of Geminga.
231: }
232: \label{search}
233: \end{figure}
234:
235: We transformed the photon arrival times to Barycentric Dynamical Time,
236: and did coherent searches of the full data set
237: using the discrete Fourier transform technique, also known
238: as the Rayleigh or $Z^2_1$ test \citep{buc83}.
239: We searched the entire range of ($f,\dot f$) parameter space
240: for a pulsar that has $0.1 \leq f \leq 1000$~Hz,
241: characteristic age $\tau_c \equiv -f/2\dot f > 20$~kyr,
242: as well as $\dot E = -4\pi^2If\dot f\leq 10^{37}$ ergs~s$^{-1}$.
243: In practice, we found it simplest to confine the search to
244: the region shaded in Figure~\ref{search}, bounded by
245: $\dot f \leq 1.4 \times 10^{-11}$ Hz~s$^{-1}$
246: for $f < 17.9$~Hz, and $\dot f \leq 2.5 \times 10^{-10}f^{-1}$ Hz~s$^{-1}$
247: for $f \geq 17.9$~Hz.
248: The total number of trials thus defined
249: is $\approx 1.6 \times 10^{10}$, an oversampling
250: of the independent trials by a factor of $\approx 7$.
251:
252: This is a liberal search range that is justified as follows:
253: \xray\ would have a blackbody temperature
254: greater than its observed $\sim 3.5 \times 10^5$~K if it
255: were younger than 20~kyr; an independent
256: argument for an older age is given in \S 4.
257: It would have a
258: wind nebula or stronger nonthermal X-ray component if
259: $\dot E > 10^{37}$ ergs s$^{-1}$. In view of
260: the distance limit $d < 800$~pc, the ratio
261: $L_x/\dot E \leq 10^{-6}$ if $\dot E = 10^{37}$ ergs~s$^{-1}$,
262: a smaller ratio than all other pulsars, which typically
263: have $10^{-4} < L_x/\dot E < 10^{-2}$. So it is unlikely
264: that $\dot E$ is as large as $10^{37}$ ergs~s$^{-1}$.
265:
266: The range of
267: parameters searched also includes all of the known
268: millisecond (recycled) pulsars, and is effective as
269: long as the pulsar does not have a neutron star binary companion.
270: The deep {\it HST\/} limit, corresponding to absolute
271: magnitude $M_V > 19$ at $d=800$~pc, rules out the more
272: common white dwarf companions.
273:
274: No significant pulsed signal was detected. The largest values
275: of $Z^2_1$ found were $\approx 44$. The theoretical
276: distribution of $Z^2_n$ follows that of $\chi^2$
277: with $2n$ degrees of freedom. For $n=1$, the distribution
278: is an exponential with a mean of 2,
279: so the single-trial probability that
280: $Z^2_1 \geq 44$ by chance is $2.8 \times 10^{-10}$. This
281: is expected to arise randomly in $\approx 10^{10}$ independent trials.
282: \citet{lea83} showed that to detect sinusoidal pulsation with
283: amplitude (pulsed fraction) $f_p$ at a power level $S = Z^2_1$
284: with 50\% probability,
285: the number of photons needed is $N=2S/f_p^2$. In our case, $N = 729$
286: (709 corrected for background), so
287: we find $f_p \le 0.35$ corresponding to $Z^2_1 \le 44$.
288: We verified this analytic expression by examining folded light curves
289: at the periods associated with the maximum values of $Z^2_1$.
290:
291: It is not surprising that the pulsed fraction of a $\gamma$-ray pulsar
292: in X-rays should be less than $35\%$. This is not a very sensitive
293: limit consdering that other $\gamma$-ray pulsars
294: whose soft X-rays are predominantly thermal (Geminga, PSR B1055--52,
295: possibly PSR B0656+14) have even smaller pulsed fractions
296: \citep[30\%, 21\%, and 12\%, respectively,][]{del05}.
297: In these cases, the pulse profiles are quasi-sinusoidal.
298: The $Z^2_1$ test is therefore a good one for
299: \xray, whose X-rays are also dominated by a soft thermal source
300: within the response of the \chandra\ HRC, although the result is
301: not very restrictive in this case.
302: We also note that the Vela pulsar has a more complicated pulse shape,
303: while its pulsed fraction is only 9\% \citep{man07}.
304:
305: \section{Chandra Astrometry}
306:
307: In order to test for proper motion of the neutron star,
308: we compared the position of \xray\ on the HRC with the ACIS
309: position obtained 3 years earlier \citep{hal02}.
310: First, we updated the ACIS position to the USNO-B1.0
311: system using optically identified sources in the ACIS
312: image. The result
313: is R.A. = $18^{\rm h}36^{\rm m}13.\!^{\rm s}685$,
314: decl. = $+59^{\circ}25^{\prime}29.\!^{\prime\prime}95$ (J2000.0),
315: and differs from the previously used USNO-A2.0 system
316: by $0.\!^{\prime\prime}3$. Then we registered the HRC image
317: to the ACIS image using seven X-ray sources in the vicinity of
318: \xray, which required a zero-point shift of the HRC image of
319: $+0.\!^{\rm s}052$ in R.A. and $+0.\!^{\prime\prime}89$ in decl..
320: The resulting HRC position of \xray\ is
321: R.A. = $18^{\rm h}36^{\rm m}13.\!^{\rm s}674$,
322: decl. = $+59^{\circ}25^{\prime}30.\!^{\prime\prime}15$ (J2000.0).
323:
324: The difference between the positions of \xray\ in 2002 and 2005
325: is therefore only
326: $0.\!^{\prime\prime}21$, which is comparable to their combined
327: statistical errors. In order to bound the possible proper
328: motion, we adopt an upper limit that is twice this difference,
329: or $0.\!^{\prime\prime}42$, corresponding to
330: $\mu < 0.\!^{\prime\prime}14$~yr$^{-1}$.
331: This allows the neutron star to have traveled from a birth in
332: the Galactic plane to its present position in $6.4 \times 10^5$~yr
333: or longer, which is not an unreasonable age for its X-ray temperature.
334: At the maximum distance of 800~pc, the limit on $\mu$
335: corresponds to a tangential velocity limit
336: $v_t<530$ km~s$^{-1}$, which is typical for radio pulsars.
337:
338: \section{Discussion and Conclusions}
339:
340: Even in the absence of detectable pulsations, \xray\ remains the
341: leading (and only) candidate for identification with \source.
342: Its soft X-ray spectrum, and absence of optical and radio emission,
343: support the hypothesis that it is an older and possibly more distant
344: cousin of the Geminga pulsar. The upper limit on
345: proper motion that we derived is a significant new observational
346: constraint, allowing an age of at least $6.4 \times 10^5$~yr
347: for \xray\ if it was born in the Galactic plane. This compares
348: favorably with the characteristic age $\tau_c = 3.4 \times 10^5$~yr
349: of Geminga, and is consistent with a surface temperature of only
350: $\sim 3.5 \times 10^5$~K, compared with $4.8 \times 10^5$~K for
351: Geminga \citep{car04,jac05,kar05}.
352:
353: Assuming that the age and
354: distance of \xray\ are both larger than those of Geminga,
355: it is likely that \source\ is a maximally efficient $\gamma$-ray
356: pulsar close to its death line \citep{che93},
357: following the trend of $\gamma$-ray efficiency
358: increasing with decreasing spin-down power \citep{tho97,tho99}.
359: Alternatively, we cannot rule out that \xray\ is a millisecond
360: pulsar with similar spin-down power and magnetospheric gap
361: voltage as Geminga. If so, its thermal X-ray emission may be due to
362: surface reheating by a magnetospheric accelerator.
363:
364: \section{Prospects for GLAST}
365:
366: It is not now possible to make a significantly more sensitive
367: search for X-ray pulsations from \xray, as the operational
368: limitations of \chandra\ no longer allow long, contiguous pointings.
369: While {\it XMM-Newton} is in principle more sensitive, it has
370: not been used to observe \xray, which lies near
371: the Earth-avoidance zone of the satellite orbit, with maximum
372: allowed pointing durations of $\approx 20$~ks.
373:
374: Finally, the upcoming
375: {\it Gamma-ray Large Area Space Telescope} (\glast)
376: may be the best hope of detecting
377: pulsations from \source. {\it GLAST}
378: will certainly reduce the positional uncertainty of \source,
379: which will be an independent test of association with \xray,
380: and additional motivation for a dedicated pulsar search.
381: For this, a sufficiently long, pointed
382: observation is best, as argued by \citet{ran07}.
383:
384: We estimate the expected \glast\ count rate
385: from \source\ using its photon flux of $5.9 \times 10^{-7}$
386: cm$^{-2}$~s$^{-1}$ between 100~MeV and 4~GeV \citep{rei01}.
387: Assuming an average on-axis effective area of 5000~cm$^2$
388: for the \glast\ Large Area Telescope (LAT) in this energy
389: range, the predicted count rate is 0.003~s$^{-1}$.
390: A more detailed calculation, convolving the spectrum of
391: \source\ from Figure~3 of \citet{rei01} with the LAT on-axis
392: effective area curve\footnote{http://www-glast.slac.stanford.edu/software/IS/glast\_lat\_performance.htm},
393: predicts a similar rate of 0.0027~s$^{-1}$ in the 100--4000~MeV range.
394: This is almost half the count rate from the \chandra\ HRC,
395: and holds great promise for a pulsar discovery.
396:
397: It is likely that \glast\ will succeed where \chandra\
398: failed because, unlike thermal X-rays, $\gamma$-rays
399: have sharp peaks with pulsed amplitude approaching 100\%.
400: In 1 week of elapsed time, including Earth blockage,
401: \glast\ can collect as many photons from \source\ as
402: we have obtained using \chandra. On-source time
403: in a pointed observation
404: of \source\ can be higher than average because
405: the source is close to the ecliptic pole. Also,
406: Galactic background is negligible at these coordinates
407: $(\ell, b)=(89^{\circ},+25^{\circ})$.
408: If \glast\ is successful, the
409: \chandra\ photons analyzed here may yet be used to find a
410: coincident period and period derivative with confidence,
411: thus establishing the identity of \source\ with \xray,
412: and determining its physical parameters.
413:
414:
415: \acknowledgements
416:
417: Support for this work was provided by NASA through {\it Chandra} Award
418: SAO GO4-5058X issued by the {\it Chandra} X-ray Observatory Center,
419: which is operated by the Smithsonian Astrophysical Observatory
420: for and on behalf of NASA under contract NAS8-03060.
421: The GBT is operated by the National Radio Astronomy Observatory,
422: a facility of the National Science Foundation operated under cooperative
423: agreement by Associated Universities, Inc.
424:
425:
426: \begin{thebibliography}{}
427:
428: \bibitem[Backer et al.(1997)]{bac97}
429: Backer, D.~C., Dexter, M.~R., Zepka, A., D., N., Wertheimer, D.~J.,
430: Ray, P.~S., \& Foster, R.~S. 1997, \pasp, 109, 61
431:
432: \bibitem[Bignami \& Caraveo(1996)]{big96}
433: Bignami, G. F., and Caraveo, P. A. 1996, \araa, 34, 331
434:
435: \bibitem[Buccheri et al.(1983)]{buc83}
436: Buccheri, R., et al.\ 1983, \aap, 128, 245
437:
438: %\bibitem[Camilo et al.(2002a)]{cam02a}
439: %Camilo, F., et al. 2002a,\apj, 571, L41
440:
441: \bibitem[Camilo et al.(2002)]{cam02}
442: Camilo, F., Manchester, R. N., Gaensler, B. M., Lorimer, D. R.,
443: \& Sarkissian, J. 2002, \apj, 567, L71
444:
445: \bibitem[Caraveo et al.(2004)]{car04}
446: Caraveo, P. A., De Luca, A., Mereghetti, S., Pellizzoni, A.,
447: \& Bignami, G. F. 2004, Science, 305, 376
448:
449: \bibitem[Chen \& Ruderman(1993)]{che93}
450: Chen, K., \& Ruderman, M. 1993, \apj, 402, 264
451:
452: \bibitem[Cordes \& Lazio(2002)]{cor02}
453: Cordes, J.~M., \& Lazio, T.~J.~W. 2002, preprint (astro-ph/0207156)
454:
455: \bibitem[De Luca et al.(2005)]{del05}
456: De Luca, A., Caraveo, P. A., Mereghetti, S., Negroni, M.,
457: \& Bignami, G. F. 2005, \apj, 623, 1051
458:
459: \bibitem[Halpern et al.(2002)]{hal02}
460: Halpern, J. P., Gotthelf, E. V., Mirabal, N., \& Camilo, F. 2002,
461: \apjl, 573, L41
462:
463: \bibitem[Halpern \& Ruderman(1993)]{hal93}
464: Halpern, J. P, \& Ruderman, M. 1993, ApJ, 415, 286
465:
466: \bibitem[Hartman et al.(1999)]{har99}
467: Hartman, R. C., et al. 1999, \apjs, 123, 79
468:
469: \bibitem[Jackson \& Halpern(2005)]{jac05}
470: Jackson, M., \& Halpern, J. P. 2005, \apj, 633, 1114
471:
472: \bibitem[Kargaltsev et al.(2005)]{kar05}
473: Kargaltsev, O. Y., Pavlov, G. G., Zavlin, V. E., \& Romani, R. R.
474: 2005, \apj, 625, 307
475:
476: \bibitem[Leahy et al.(1983)]{lea83}
477: Leahy, D. A,. Elsner, R. F., \& Weisskopf, M. C. 1983, \apj, 272, 256
478:
479: \bibitem[Lorimer et al.(1995)]{lor95}
480: Lorimer, D.~R., Yates, J.~A., Lyne, A.~G., \& Gould, D.~M. 1995,
481: \mnras, 273, 411
482:
483: \bibitem[Manzali et al.(2007)]{man07}
484: Manzali, A., De Luca, A., \& Caraveo, P. 2007, ApJ, in press
485:
486: \bibitem[McLaughlin et al.(1999)]{mcl99}
487: McLaughlin, M. A., Cordes, J. M., Hankins, T. H.,\& Moffett, D. A.
488: 1999, \apj, 512, 929
489:
490: \bibitem[Mirabal \& Halpern(2001)]{mir01}
491: Mirabal, N., \& Halpern, J. P. 2001, \apj, 547, L137
492:
493: \bibitem[Mirabal et al.(2000)]{mir00}
494: Mirabal, N., Halpern, J. P., Eracleous, M., \& Becker, R. H. 2000, \apj, 541,
495: 180
496:
497: \bibitem[Ransom(2007)]{ran07}
498: Ransom, S. M. 2007, in Proc First GLAST Symposium
499: (arXiv:0704.113)
500:
501: \bibitem[Ransom et al.(2002)]{ran02}
502: Ransom, S.~M., Eikenberry, S.~S., \& Middleditch, J. 2002, \aj, 124, 1788
503:
504: \bibitem[Reimer et al.(2001)]{rei01}
505: Reimer, O., Brazier, K. T. S., Carrami\~nana, A., Kanbach, G., Nolan, P. L.,
506: \& Thompson, D. J. 2001, \mnras, 324, 772
507:
508: \bibitem[Thompson et al.(1997)]{tho97}
509: Thompson, D. J., Harding, A. K., Hermsen, W., \& Ulmer, M. P. 1997,
510: in AIP Conf. Proc. 410, Proc. Fourth Compton Symp., ed. C. D. Dermer,
511: M. S. Strickman, \& J. D. Kurfess (New York: AIP), 39
512:
513: \bibitem[Thompson et al.(1999)]{tho99}
514: Thompson, D. J., et al. 1999, \apj, 516, 297
515:
516: \end{thebibliography}
517:
518: \end{document}
519:
520:
521:
522:
523:
524:
525:
526:
527:
528:
529:
530:
531:
532:
533: