astro-ph0605395/ms.tex
1: \documentstyle[12pt,aasms4]{article}
2: %\documentclass[12pt,preprint]{aastex}
3: %\documentclass{aastex}
4: %\shorttitle{More Periods... }
5: %\shortauthors{Schmidtke \& Cowley }
6: \begin{document}
7: 
8: \title{Photometric Variability of Be/X-ray-Pulsar Binaries in the SMC } 
9: 
10: \author{ P.C.\ Schmidtke \& A.P.\ Cowley }
11: 
12: \affil{Dept.\ of Physics \& Astronomy, Arizona State Univ., Tempe, AZ,
13: 85287-1504, USA; paul.schmidtke@asu.edu, anne.cowley@asu.edu } 
14: 
15: \begin{abstract}
16: 
17: We have studied the photometric variability of ten SMC Be/X-ray pulsars
18: using MACHO and OGLE-II data.  For some of these systems we have found
19: periodic behavior, including orbital outbursts and/or nonradial pulsations
20: (NRP) of the Be star.  For others we were unable to identify any clear
21: photometric periodicity, although their longterm light curves show
22: significant structure.  We present periodograms, phase dispersion
23: minimization (PDM) variances, and folded light curves for the systems
24: which exhibit periodic photometric variability. 
25: 
26: \end{abstract}
27: 
28: \keywords{X-rays: binaries -- stars: Be -- (stars:) pulsars -- stars:
29: variable -- stars: individual: 
30: XMMU~J005152.2$-$731033 
31: RX~J0051.8$-$7231 
32: RX~J0052.1$-$7319 
33: CXOU~J005323.8$-$722715 
34: XMMU~J005605.2$-$722200 
35: XMMU~J005920.8$-$722316 
36: CXOU~J010102.7$-$720658 
37: SAX~J0103.2$-$7209
38: RX~J0103.6$-$7201
39: RX~J0105.1$-$7211 
40: }
41:                                     
42: \section{Introduction}
43: 
44: The Be/neutron-star systems are an important subclass of high-mass X-ray
45: binaries.  These systems are particularly numerous in the Small Magellanic
46: Cloud where about four dozen are known (Coe et al.\ 2005).  Many are
47: transient X-ray sources, and a large fraction have been shown to be X-ray
48: pulsars (e.g. Coe et al.\ 2005, Liu et al.\ 2005, Haberl \& Pietsch 2004,
49: and references therein). 
50: 
51: The optical identifications for these sources depend on accurate X-ray
52: positions which have been greatly improved with data from {\it Chandra}
53: and {\it XMM-Newton}.  Two valuable sources of optical data for Magellanic
54: Cloud systems are the MACHO and OGLE-II surveys (e.g. Alcock et al.\ 1999,
55: Udalski et al.\ 1997, Zebrun et al.\ 2001, Szymanski 2005) that provide
56: regular photometric data in $V,R$ (MACHO) and $I$ (OGLE-II) over many
57: years.  For SMC sources the identification with Be stars is more secure if
58: they are also included in the catalogue of H$\alpha$ emission-line stars
59: of Meyssonnier \& Azzopardi (1993; hereafter designated as [MA93]). 
60: 
61: In previous papers we have studied the photometric properties of a sample
62: of Magellanic Cloud X-ray binaries (e.g. Schmidtke et al.\ 2004, Schmidtke
63: \& Cowley 2005a). In this paper we present results for ten more Be/X-ray
64: pulsars in the SMC.  The sources studied were selected both on the
65: basis of availability of longterm photometry and because they lack
66: previous optical studies.
67: 
68: \section{Analysis of Optical Data from the MACHO and OGLE-II Projects} 
69: 
70: As in our earlier work, we have relied on photometric data from the MACHO
71: and OGLE-II surveys.  We have transformed the `blue' and `red'
72: instrumental MACHO magnitudes into standard $V$ and $R$ colors (see Alcock
73: et al.\ 1999).  Each data set, or portion there of, was flattened
74: (prewhitened) to remove longterm changes in the mean brightness level by
75: fitting a low order polynomical to the data.  In sources which show large
76: longterm variations (ones we call ``swoopers") the light curve can only be
77: flattened in segments, and there are often parts that can't be flattened
78: at all because the slope of variation is too steep or irregular.  The
79: resulting light curve segments are then analyzed using the method of Horne
80: \& Baliunas (1986).  Often these systems show periodic ``outbursts'',
81: thought to be caused by interaction between the neutron star and the Be
82: star's disk near periastron passage.  In such cases the Horne \& Baliunas
83: method may not be the best way to search for periods, since it assumes
84: sinusoidal variations.  Thus, for some systems we also used the phase
85: dispersion minimization (PDM) technique described by Stellingwerf (1978).
86: This method can identify periods in a light curve whose shape is not
87: sinusoidal, but instead is characterized by recurring outbursts. 
88: 
89: Table 1 lists the X-ray sources studied and includes the MACHO coordinates
90: of the optical counterpart, the MACHO and OGLE-II identification numbers,
91: and other information about the systems.  In the text the sources we studied
92: are listed in order of right ascension for ease in locating the discussion 
93: about them.  Coe et al.\ (2005) have
94: introduced a useful identification number for Small Magellanic Cloud
95: pulsars which avoids the confusion of multiple X-ray names.  They use SXP
96: followed by the pulse period in seconds (e.g. SXP15.3 or SXP202), and they
97: give a convenient cross listing of names (their Table 1).  For all the
98: systems studied here we have included the SXP number.  In addition, Coe
99: presents finding charts for most of the SMC pulsars on his web
100: site\footnote{http://www.astro.soton.ac.uk/$\sim$mjc/smc}.  For the
101: systems we have studied, we have confirmed his identifications are the
102: same as ours. 
103: 
104: In an earlier paper on SMC pulsars (Schmidtke \& Cowley 2005a), we showed
105: that some Be/pulsars exhibit nonradial pulsations (NRP) with periods
106: typically less than a day.  Therefore, in all cases we have not only
107: searched for orbital periods, but also for pulsation periods.  Our
108: searches covered the period range from 0.2 to 1000 days using a variety of
109: techniques and subsamples of the data. 
110: 
111: Table 2 lists the periods found for some of the Be/pulsar systems.  For a
112: few systems possible periods are only given in the text, but not included
113: in the table, because we think they need further verification.  There are
114: other sources for which we found no clear photometric periods in spite of
115: detailed searches of the entire data set or subsections of it.  In the
116: section below we give details about the analysis of each source. 
117: 
118: \section{Individual Be/X-ray Pulsar Systems}
119: 
120: \subsection{ XMMU~J005152.2$-$731033 = RX~J0051.9$-$7311 = SXP172 }
121: 
122: This X-ray source was identified with a Be star in the SMC by Cowley et
123: al.\ (1997; their ``Star 1"; also [MA93]504).  Yokogawa et al.\ (2000)
124: later found this source (then named AX~J0051.6$-$7311) to show 172.4 s
125: X-ray pulsations.  Haberl \& Pietsch (2004) also detected the pulsar and
126: gave its pulse period as 172.21$\pm$0.13 s. 
127: 
128: The longterm OGLE-II light curve (Fig.\ 1) shows the source undergoes slow
129: dips and rises.  The MACHO data have several problems.  The west-pier $V$
130: data have many `dropouts' and hence are not usable.  $R$ data are only
131: available when the telescope was east of the pier.  Hence, the longterm
132: MACHO light curves are not continuous, making periods difficult to
133: identify.  When only east-pier data were studied, no clear periods were
134: found in either $V$ or $R$, and additionally none were found in any
135: subset of the MACHO data. 
136: 
137: However, OGLE-II $I$ data reveal a period of 69.9$\pm$0.6 days (Schmidtke
138: \& Cowley 2005b).  We assume this is the orbital period, although one
139: might have expected a somewhat longer orbital period based on Corbet's
140: (1984) P$_{pulse}$/P$_{orb}$ relation.  Fig.\ 1 shows the PDM and
141: periodogram, which are similar to those seen in some other Be/X-ray
142: systems.  Although the strongest peaks in the periodogram are at 1/2 and
143: 1/3 the true period, the PDM variance is minimized at 69.9 days.  The
144: 69.9-day folded light curve (Fig.\ 1) shows ``outbursts" lasting a few
145: tenths of the period, with an average amplitude of $\sim$0.02 mag. 
146: However, at some epochs the outbursts are more prominent.  In Fig.\ 1 we
147: have superimposed observations from the strongest outburst on the mean
148: folded light curve to show how much the amplitude varies.  We have also
149: searched for short periods (i.e. less than 2 days), but we found nothing
150: significant. 
151: 
152: Laycock et al.\ (2005) report a period of P$_X$=67$\pm$5 days based on
153: several X-ray outbursts.  This 67-day period is in acceptable agreement
154: with our optical period, considering their analysis is based on weekly
155: monitoring.  We note that their Table 6 lists a value of P(orb)=147$\pm$24
156: days, rather than the 67-day value given in their text. 
157: 
158: \subsection{ RX~J0051.8$-$7231 = SXP8.88 }
159: 
160: This variable X-ray source, also known as 1E0050.1$-$7247, was found to be
161: a pulsar by Israel et al.\ (1997).  Although there is some uncertainty
162: about its exact position, the optical counterpart is thought to be the Be
163: star [MA93]506 (which is also Star \#2 of Stevens et al.\ 1999).  Fig.\ 2
164: shows its longterm optical light curve from both MACHO and OGLE-II data.
165: Its irregular ``swooping" behavior continued during OGLE-III observations,
166: as shown by Coe et al.\ (2005).  They suggest that this source may be the
167: same as SXP16.6, since neither source has a well-defined position. 
168: However, Laycock et al.\ (2005) argue against this interpretation. 
169: 
170: From MACHO $R$ data, Coe et al.\ (2005) derived P$_{orb}$=185$\pm$4 days.
171: However, since the calibration of east- and west-pier data is not always
172: the same, this can introduce a half-year pseudo-period.  Hence, we have
173: processed data from the two pier configurations separately and merged the
174: data sets after prewhitening.  Using this technique, we did not find the
175: 185-day period.  We suspect that it is an artifact of slightly different
176: calibrations for the east- and west-pier data. 
177: 
178: We have also subdivided both OGLE-II and MACHO photometry into time
179: segments.  In the segment near the start of OGLE-II observations (MJD
180: 50600-50800) there is a weak periodicity at P=33.4$\pm$1.0 days.  This
181: appears to be caused by several small outbursts (perhaps near periastron
182: passage, with amplitudes varying from orbit to orbit).  These outbursts
183: are marked on the inset of Fig.\ 2.  Although these high points are from
184: several different orbital cycles, there is no evidence of this period in
185: any of the MACHO data.  New data would be needed to confirm this possible
186: periodicity.  Corbet et al.\ (2004) analyzed X-ray outbursts from this
187: source and found P$_X$=28.0$\pm$0.3 days, which is in general agreement
188: with the possible 33-day optical period. 
189: 
190: \subsection{ RX~J0052.1$-$7319 = SXP15.3} 
191: 
192: This transient source was found to be a pulsar by Finger et al.\ (2001).
193: It was optically identified with a Be star by Covino et al.\ (2001; their
194: ``Star A" which is also [MA93]552).  The longterm MACHO and OGLE-II light
195: curves for this star are shown in Fig.\ 2.  In spite of a detailed
196: investigation of these data and various subsets of it, we were unable to
197: find any significant periods in $V$, $R$, or $I$. 
198: 
199: \subsection{ CXOU~J005323.8$-$722715 = SXP138  }
200: 
201: Using $Chandra$ data, Edge et al.\ (2004) found this source to be a 138-s
202: pulsar.  From the accurate X-ray position they identified the optical
203: counterpart as the Be star [MA93]667.  Edge et al.\ derived a photometric
204: (orbital) period of P$_{orb}$=125$\pm$1.5 days from MACHO data, although
205: details of their analysis are not given. 
206: 
207: The longterm $R$ light curve for this star is plotted in Fig.\ 2.  There
208: is a slow downward trend with occasional small ``outbursts'' (e.g. near MJD
209: 50400 and 50620), but these are not periodic.  We have re-examined both
210: the $V$ and $R$ data, studying the east-pier and west-pier data
211: separately.  We find that there is a significant seasonal effect in which
212: the mean brightness from each pier differs.  Hence, it was necessary to
213: prewhiten the data from each pier configuration before combining them. 
214: 
215: We have searched for periods in both MACHO colors and in a variety of data
216: subsets.  Although the first half of the west-$R$ data shows a weak
217: periodicity near 122-123 days (perhaps the same period as reported by Edge
218: et al.), other time segments show different possible periods in the range
219: of 220-260 days.  It appears that each of these periods is related either
220: to the length of the seasonal data train in a pier configuration or gaps
221: between these data sets.  No single long period is consistent with all of
222: the photometry.  Each of these ``periods" produced a low-amplitude
223: sinusoidal light curve, rather than the outburst behavior usually
224: displayed in orbital periods. 
225: 
226: The periodograms for this source also show some power near $\sim$1 day,
227: but in each case these are aliases of the long ``periods'' discussed above.
228: We conclude that none of the peaks in the periodograms represents a true
229: orbital period, but rather they are artifacts of the data. 
230: 
231: \subsection{ XMMU~J005605.2$-$722200 = 2E0054.4$-$7237 = SXP140}
232: 
233: This 140-s pulsar was discovered by Sasaki, Pietsch, \& Haberl (2003)
234: using $XMM-Newton$ data.  They note that it may be the same source as
235: 2E~0054.4$-$7237.  Its position is coincident with the Be star [MA93]904
236: whose longterm MACHO light curve is shown in Fig.\ 3.  The source is
237: clearly a ``swooper''.  Unfortunately it lies outside of the OGLE-II
238: fields, so $I$ photometry is not available.  Coe et al.\ (2005) note a
239: blending problem in the crowded field of this source (see image at:
240: ~http://www.astro.soton.ac.uk/$\sim$mjc/smc/Finders/sxp140.jpg). 
241: 
242: The MACHO data were analyzed mainly in Seg A (MJD 49650-50500), since this
243: is the only part of the longterm light curve that could be prewhitened
244: reasonably well.  A period near 197$\pm$5 days is present in both $R$ and
245: $V$ photometry.  The folded light curve shows a typical orbital outburst,
246: with a rapid rise ($\sim$0.04 mag) and a more gradual decline.  In other
247: time segments of the longterm light curve, where it was impossible to
248: flatten the data, we examined the region where outbursts were expected. 
249: In most cases there was evidence for a brightening at the predicted times.
250:  The 197-day period is in reasonable agreement with Corbet's (1984)
251: P$_{pul}$/P$_{orb}$ relationship.  We conclude that this is probably the
252: orbital period since in similar systems the amplitude of outbursts varies
253: or may be absent. 
254: 
255: \subsection{ XMMU~J005920.8$-$722316 = SXP202 } 
256: 
257: This X-ray source was discovered to be a 202-s pulsar by Majid et al.\
258: (2004) using archival $XMM-Newton$ data.  Its X-ray position is coincident
259: with both an OGLE-II and a MACHO star.  The longterm light curve for this
260: system (Fig.\ 4) shows it to be another ``swooper''. 
261: 
262: We have searched the $I$, $R$, and $V$ data for possible periods.  The
263: photometric behavior is similar to SXP138, with various subsets of the
264: data showing weak power at periods related to the length of the data train
265: and its aliases.  In $I$ there is some power at $\sim$203 days and its
266: alias at $\sim$0.99 days, but light curves folded at these periods merely
267: bunch the observations into a few clumps and do not appear to be
268: reasonable.  We conclude that there are no clear periodities in the
269: photometric data.  Coe et al.\ (2005) came to the same conclusion. 
270: 
271: \subsection{ CXOU~J010102.7$-$720658 = RX~J0101.0$-$7206 = SXP304 }
272: 
273: Macomb et al.\ (2003) discovered this 304.5-s pulsar in $Chandra$
274: observations of the SMC.  Its position clearly identifies it with the Be
275: star [MA93]1240, which is the same star Edge \& Coe (2003) found to have
276: strong H$\alpha$ emission. 
277: 
278: Although no OGLE-II data are available, this star is included in the MACHO
279: data set.  The longterm MACHO $V$ light curve is shown in Fig.\ 5.  The
280: mean magnitude is relatively flat over a long stretch of time (MJD
281: 49000-51000) allowing it to be prewhitened quite easily.  The periodogram
282: shows very prominent power at P=0.26 days, which we associate with
283: nonradial pulsations of the Be star.  When the data are subdivided into
284: four time segments, there is evidence that the period changes slightly
285: between segments A-B and C-D (as shown in Fig.\ 5).  The folded Seg C data
286: show that the amplitude is greatest in $V$ and least in $I$. 
287: 
288: We have also searched for longer periods which might arise from orbital
289: interactions between the pulsar and the Be star's disk.  Using the same
290: data in which the NRP were found, there is moderate power at P$\sim$520
291: days.  To further investigate this long period, we subtracted a sinusoid
292: with the period and amplitude of the NRP.  Analysis of the residuals still
293: reveals a periodicity at 520$\pm$12 days.  The light curve folded on this
294: period shows a steep rise and slower decline, typical of orbital
295: outbursts.  However, the significance of this period is low, and further
296: data are needed to confirm this possible orbital period. 
297: 
298: \subsection{ SAX~J0103.2$-$7209 = 2E~0101.5$-$7225 = SXP348 }
299: 
300: Israel et al.\ (2000) showed this X-ray source to be a 348-s pulsar.  Its
301: position identifies SXP348 with the Be star [MA93]904.  The MACHO data
302: have a significant number of `dropouts' and other problems, probably
303: related to its proximity to a much brighter star.  Although unlisted in
304: OGLE-II's Difference Image Analysis, photometry for the star is available
305: in the complete OGLE-II data set.  The $I$ light curve is shown in Fig.\
306: 4. 
307: 
308: Analysis of the $I$ photometry shows a weak period at 93.9 days.  The
309: folded light curve shows a steep rise and slower decline, but additional
310: data would be needed to confirm this as the orbital period.  A system with
311: a 348-s pulse period would be expected to have an orbital period
312: considerably longer than 94 days. 
313: 
314: \subsection{ RX~J0103.6$-$7201 = SXP1323 }
315:  
316: Using $XMM-Newton$ data Haberl \& Pietsch (2005) discovered 1323-s X-ray
317: pulsations from this source, making it the slowest known SMC pulsar.  Its
318: optical counterpart is the Be star [MA93]1393.  Only OGLE-II data are
319: available for this star, but this photometry reveals a wealth of
320: information about the source's variability. 
321: 
322: The periodogram shows extremely significant power at P=26 days as well as
323: at other periods (see Fig.\ 6 and Schmidtke \& Cowley 2006).  When the
324: data are folded on the 26-day period, the variation is approximately
325: sinusoidal with a full amplitude of $\sim$0.04 mag.  From the known
326: relation between pulse period and orbital period (Corbet 1984), one would
327: expect the orbital period to be much longer than 26 days.  In addition,
328: orbital light curves usually show outbursts rather than sinusoidal
329: variations.  Hence, we conclude that the alias period at P=0.96 days (the
330: second strongest peak in the periodogram) is likely to be the true primary
331: period, with the variation due to nonradial pulsations of the Be star. 
332: 
333: The periodogram also shows many other peaks, some resulting from other
334: pulsational periods and others being aliases of the true periods.  The
335: next strongest peak is at P=0.88 days, which we identify as a second
336: pulsational mode of the Be star.  A third pulsational period is present at
337: 0.42 days, as can be seen in Fig.\ 6. 
338: 
339: We further analyzed these periods using a variety of techniques.  Although
340: all three of the NRP periods given above can be seen in the original
341: periodogram, we experimented with the data by subtracting the 0.96-d
342: period and examining the periodogram of the residuals.  This showed the
343: 0.88-day period to be highly significant.  The 0.96-day light curve must
344: be a nearly perfect sine curve since no evidence of it was found in the
345: residuals.  We then subtracted the 0.88-day period, and examined the new
346: residuals.  This showed periods at both 0.42 and 0.28 days (aliases of
347: each other), with the 0.42-day period being slightly more significant.
348: Finally, we note that when all three NRPs (0.96, 0.88, and 0.42 days) are
349: subtracted from the data, the resulting residuals show another weak peak
350: at 1.05 days, suggesting that a fourth NRP period may be present.  Thus,
351: this source has many pulsational modes, all present at the same time.  We
352: also looked for changes in these periods by subdividing the longterm light
353: curve into 4 time segments.  No significant changes in any of the periods
354: were found. 
355: 
356: \subsection{ RX~J0105.1$-$7211 = AX~J0105$-$722 = SXP3.34 }
357: 
358: This 3.34-s pulsar was identified by Coe et al.\ (2005) with the Be star
359: [MA93]1506.  The longterm MACHO light (Fig.\ 7) shows a fairly steady mean
360: light level with considerable scatter.  Coe et al.\ found an optical
361: period of 11.09 days from these data, but he pointed out that the strength
362: of H$\alpha$ emission (EW=$-$54\AA) was not consistent with the
363: P$_{orb}$/EW relationship of Reig, Fabrigat, \& Coe (1997).  We have
364: searched to higher frequencies and find that the fundamental period is at
365: 1.099 days, with the 11-day period being an alias of that period
366: (Schmidtke \& Cowley 2005c).  The power at the 1.009-day period is
367: exceptionally strong (Fig.\ 7).  This period is extremely stable in both
368: frequency and amplitude.  The folded $V$, $B$, and $I$ light curves show
369: the greatest amplitude in $I$ and the smallest in $V$, although the
370: differences are small.  This short period clearly is another example of
371: prominent nonradial pulsations from the Be star. 
372: 
373: When the 1.099-day variation is subtracted from the data, analysis of the
374: residuals shows a periodogram with a second significant peak at P=0.980
375: days.  The folded light curve reveals a very low-amplitude sinusoid in
376: each color (Fig.\ 7).  When the data are divided into 4 time segments, we
377: find that the amplitude is weakest in the last segment (starting about MJD
378: 50950). 
379: 
380: \section{ Summary}
381: 
382: We have looked for general conclusions regarding the Be/X-ray pulsars
383: systems based on their photometric behavior.  We have considered not only
384: the sources in this paper, but also results from previous studies (e.g.
385: Coe et al.\ 2005, Schmidtke \& Cowley 2005a, Schmidtke et al.\ 2004).  We
386: see no correlation between the length of the X-ray pulse period and
387: whether or not orbital periods are seen in the photometric data.  Sources
388: with short pulse periods (e.g. 7.8 s for SMC X-3), long pulse periods
389: (e.g. SXP755 with P$_{orb}$=394 days), and many intermediate values may or
390: may not show orbital outbursts.  The presence of such outbursts must
391: depend on the orbital eccentricity and the extent of the Be star's disk at
392: the time of periastron passage.  Perhaps the systems which do not show
393: outbursts have nearly circular orbits. 
394: 
395: We also see that NRP are found in sources with a wide range of X-ray pulse
396: periods (from 3.34 s to 1323 s).  However, we note that sources with
397: ``swooping'' light curves generally do not show NRP, even in the flattest
398: portions of their light curves (e.g. SXP8.88).  The only possible
399: exception to date is SXP82.4 whose OGLE-III light curve (Coe et al.\ 2005)
400: fades considerably, suggesting it may be a ``moderate swooper".  Its
401: 1.33-day NRP were found only in the early part of the longterm light curve
402: when its the mean brightness was almost level (see Fig.\ 4 in Schmidtke \&
403: Cowley 2005a). 
404: 
405: In summary, we have analyzed the longterm optical light curves of ten SMC
406: Be/X-ray pulsars.  We determined orbital periods for four sources and a
407: possible orbital period for another system.  Overall, the orbital periods
408: found in this study fit reasonably well with the Corbet relation between
409: orbital and X-ray pulse period.  We have identified strong nonradial
410: pulsations for three sources, and two of these sources show multiple NRP. 
411: 
412: \acknowledgments
413: 
414: This paper utilizes public domain data obtained by the MACHO Project,
415: jointly funded by the US Department of Energy through the University of
416: California, Lawrence Livermore National Laboratory under contract No.\
417: W-7405-Eng-48, by the National Science Foundation through the Center for
418: Particle Astrophysics of the University of California under coopertative
419: agreement AST-8809616, and by the Mount Stromlo and Siding Spring
420: Observatory, part of the Australian National University. 
421: 
422: The OGLE-II database, as described by Udalski et al.\ (1997), Zebrun et
423: al.\ (2001), and Szymanski (2005), was also extensively used for this
424: project. 
425: 
426: \newpage
427: 
428: \begin{references}
429: 
430: \reference{} Alcock, C., et al. 1999, PASP, 111, 1539
431: 
432: \reference{} Coe, M.J., Edge, W.R.T., Galache, J.L., \& McBride, V.A. 
433: 2005, MNRAS, 356, 502
434: 
435: \reference{} Corbet, R.H.D. 1984, A\&A, 141, 91 
436: 
437: \reference{} Corbet, R.H.D., Coe, M.J., Edge, W.R.T., Laycock, S.,
438: Markwardt, C.B., \& Marshall, F. 2004, ATel, \#277
439: 
440: \reference{} Covino, S., Negueruela, I., Campana, S., Israel, G.L., Polcaro,
441: V.F., Stella, L., \& Verrecchia, F. 2001, A\&A, 374, 1009
442:  
443: \reference{} Cowley, A.P., Schmidtke, P.C., McGrath, T.K., Ponder, A.L., 
444: Fertig, M.R., Hutchings, J.B., \& Crampton, D. 1997, PASP, 109, 21
445: 
446: \reference{} Edge, W.R.T., \& Coe, M.J. 2003, MNRAS, 338, 428 
447: 
448: \reference{} Edge, W.R.T., Coe, M.J., Galache, J.L., McBride, V.A., 
449: Corbet, R.H.D., Markwardt, C.B., \& Laycock, S. 2004, MNRAS, 353, 1286
450: 
451: \reference{} Finger, M.H., Macomb, D.J., Lamb, R.C., Prince, T.A., Coe, 
452: M.J., \& Haigh, N.J. 2001, ApJ, 560, 378
453: 
454: \reference{} Haberl, F., \& Pietsch, W. 2004, A\&A, 414, 667
455: 
456: \reference{} Haberl, F., \& Pietsch, W. 2005, A\&A, 438, 211
457:  
458: \reference{} Horne, J.H., \& Baliunas, S.L. 1986, ApJ, 302, 757
459: 
460: \reference{} Israel, G.L., Stella, L., Angelini, L., White, N.E., Giommi, 
461: P., \& Covino, S. 1997, ApJ, 484, L141
462: 
463: \reference{} Israel, G.L., et al. 2000, ApJ, 531, L131
464: 
465: \reference{} Laycock, S., Corbet, R.H.D., Coe, M.J., Marshall, F.E.,
466: Markwardt, C., \& Lochner, J. 2005, ApJS, 161, 96 
467: 
468: \reference{} Liu, Q.Z., van Paradijs, J., \& van den Heuvel, E.P.J. 2005,
469: A\&A, 442, 1135
470: 
471: \reference{} Macomb, D.J., Fox, D.W., Lamb, R.C., \& Prince, T.A. 2003, 
472: ApJ, 584, L79
473: 
474: \reference{} Majid, W.A., Lamb, R.C., \& Macomb, D.J. 2004, ApJ, 609, 133
475: 
476: \reference{} Meyssonnier, N., \& Azzopardi, M. 1993, A\&AS, 102, 451 
477: ~ ([MA93])
478: 
479: \reference{} Reig, P., Fabregat, J., \& Coe, M.J. 1997, A\&A, 322, 193
480: 
481: \reference{} Sasaki, M., Pietsch, W., \& Haberl, F. 2003, A\&A, 403, 901
482: 
483: \reference{} Schmidtke, P.C., \& Cowley, A.P. 2005a, AJ, 130, 2220
484: 
485: \reference{} Schmidtke, P.C., \& Cowley, A.P. 2005b, AAS, 20718319
486: 
487: \reference{} Schmidtke, P.C., \& Cowley, A.P. 2005c, ATel, 648 
488: 
489: \reference{} Schmidtke, P.C., \& Cowley, A.P. 2006, ATel, 716 
490: 
491: \reference{} Schmidtke, P.C., Cowley, A.P., Levenson, L., \& Sweet, K. 
492: 2004, AJ, 127, 3388
493: 
494: \reference{} Stellingwerf, R.F. 1978, ApJ, 224, 953
495: 
496: \reference{} Stevens, J.B., Coe, M.J., \& Buckley, D.A.H. 1999, MNRAS, 
497: 309, 421
498: 
499: \reference{} Szymanski, M.K. 2005, Acta Astron., 55, 43
500: 
501: \reference{} Udalski, A., Kubiak, M., \& Szymanski, M. 1997, Acta Astron., 
502: 47, 319
503: 
504: \reference{} Yokogawa, J., Torii, K., Imanishi, K., \& Koyama, K. 2000, 
505: PASJ, 52, L37
506: 
507: \reference{} Zebrun, K., et al. 2001, Acta Astron., 51, 317
508: 
509: \end{references}
510: 
511: 
512: \newpage
513: 
514: \begin{figure}
515: %fig 1  
516: \caption{ (top) Longterm $I$ light curve of XMMU~J005152.2$-$731033
517: (SXP172).  ~(middle) Variance calculated by the PDM method (upper, with
518: scale on right) and periodogram (lower) of $I^*$ data from Seg A. The 99\%
519: confidence level is also shown.  ~(bottom) $I^*$ data folded on the
520: 69.9-day period and binned (filled squares), with the individual points
521: from the outburst near MJD 50790 superimposed (open circles). } 
522: \end{figure}
523: 
524: \begin{figure}
525: %fig 2 
526: \caption{ (top) Longterm $V$ (upper) and $I$ (lower) light curves of
527: RX~J0051.8$-$7231 (SXP8.88).  The inset shows an enlargement of the region
528: between MJD 50600 and 50800, with four outbursts marked.  ~(middle)
529: Longterm $R$ (upper) and $I$ (lower) light curves of RX~J0052.1$-$7319
530: (SXP15.3).  ~(bottom) Longterm $R$ light curve of CXOU~J005323.8$-$722715
531: (SXP138).  } 
532: \end{figure}
533: 
534: \begin{figure}
535: %fig 3 
536: \caption{ (top) Longterm $R$ light curve XMMU~J005605.2$-$722200
537: (SXP140).  ~(middle) Periodogram of $R^*$ data from Seg A showing the
538: orbital period of 197 days.  ~(bottom) $R^*$ light curve folded on 197
539: days and averaged in 20 phase bins.  } 
540: \end{figure}
541: 
542: \begin{figure}
543: %fig 4 
544: \caption{(top) Longterm $R$ (upper) and $I$ (lower) light curves of 
545: XMMU~J005920.8$-$722316 (SXP202).  ~(bottom) Longterm $I$ light curve of
546: SAX~J0103.2$-$7209 (SXP348).} 
547: \end{figure}
548: 
549: \begin{figure}
550: %fig 5 
551: \caption{ (top) Longterm $V$ (upper) and $I$ (lower) light curves of 
552: CXOU~J010102.7$-$720658 (SXP304).  ~(middle) Periodograms of $V^*$ data
553: from 4 time segments (A -- D) showing the variation in period and power.
554: ~(bottom) Folded light curve of $V^*$, $R^*$, and $I^*$ data from Seg C
555: showing strong nonradial pulsations at P=0.26 days.  The amplitude appears
556: to be greatest in $V^*$.  } 
557: \end{figure}
558: 
559: \newpage
560: 
561: \begin{figure}
562: %fig 6 
563: \caption{ (top) Longterm $I$ light curve of RX~J0103.6$-$7201
564: (SXP1323).  ~(second) Periodogram of $I^*$ data showing several
565: significant periods and their aliases (see text). ~(third) $I^*$ data,
566: folded on P=0.96 days, after signals from P=0.88 and P=0.41 days have been
567: subtracted.  ~(fourth) $I^*$ data, folded on P=0.88 days after the
568: variation of the other two NRP have been subtracted.  ~(bottom) Similar to
569: the above two plots, but the $I^*$ data have been folded on P=0.41 days
570: after subtraction of the variation due to the two other short periods. } 
571: \end{figure}
572: 
573: \begin{figure}
574: %fig 7 
575: \caption{ (top) Longterm $R$ (upper) and $I$ (lower) light curves of 
576: RX~J0105.1$-$7211 (SXP3.34).  ~(second) Periodogram of $R^*$ data showing
577: the exceptionally strong peak at P=1.099 days.  ~(third) Folded and binned
578: light curves in $V^*$, $R^*$, and $I^*$ at P=1.099 days.  The amplitude is
579: largest in $I^*$.  ~(fourth) Periodogram of $R^*$ data after the 1.099-day
580: variation has been subtracted.  The strongest remaining power is at
581: P=0.980 days.  ~(bottom) $V^*$, $R^*$, and $I^*$ light curves, folded on
582: P=0.980 days, after removal of the 1.099-day signal.  See text for further
583: discussion of this source.} 
584: \end{figure}
585: 
586: \newpage
587: 
588: \begin{deluxetable}{llll}
589: \tablenum{1}
590: \tablecaption{Small Magellanic Cloud Be/Neutron-Star Systems Studied}
591: \tablehead{  
592: \colhead{System Name} &
593: \colhead{R.A.(2000)\tablenotemark{a}} &
594: \colhead{Dec.(2000)\tablenotemark{a}} &
595: \colhead{SXP \#} \\
596: \colhead{~~~~~~~OGLE-II \#} &
597: \colhead{~~~~~MACHO \#} &
598: \colhead{~~~~~~~[MA93] \#\tablenotemark{b}} & 
599: \colhead{ }
600: }
601: \startdata
602: XMMU~J005152.2$-$731033 &  00:51:51.86 & $-$73:10:33.4 &SXP172  \nl
603:   ~~~~~005152.02$-$731033.7 &~~~~~212.16077.13 & ~~~~~~~~504 & \nl
604: RX~J0051.8$-$7231 &  00:51:53.00 & $-$72:31:48.1 & SXP8.88 \nl
605:  ~~~~~005153.12$-$723148.3 & ~~~~~208.16087.9 & ~~~~~~~~506 & \nl
606: RX~J0052.1$-$7319 & 00:52:13.98 & $-$73:19:18.1  & SXP15.3\nl
607:  ~~~~~005213.99$-$731918.3 & ~~~~~212.16075.13 & ~~~~~~~~552  & \nl
608: CXOU~J005323.8$-$722715 & 00:53:23.86 & $-$72:27:15.1 & SXP138  \nl
609:  ~~~~~~~~~~- - - - -  &~~~~~207.16202.50 & ~~~~~~~~667 & \nl
610: XMMU~J005605.2$-$722200 & 00:56:05.62 & $-$72:22:00.2  & SXP140 \nl
611:  ~~~~~~~~~~- - - - -  &~~~~~207.16374.21 & ~~~~~~~~904 & \nl
612: XMMU~J005920.8$-$722316 & 00:59:20.94 & $-$72:23:17.4 & SXP202  \nl
613:  ~~~~~005921.03$-$722317.1 &~~~~~207.16545.12 & ~~~~~~~~--- & \nl
614: CXOU~J010102.7$-$720658 & 01:01:02.56 & $-$72:06:56.5  & SXP304 \nl
615:  ~~~~~010102.87$-$720658.7  & ~~~~~206.16663.16 &~~~~~~~~1240 & \nl
616: SAX~J0103.2$-$7209 & 01:03:13.93 & $-$72:09:13.8 & SXP348 \nl
617:  ~~~~~010313.89$-$720914.0 &~~~~~206.16776.17 & ~~~~~~~~904 & \nl
618: RX~J0103.6$-$7201 & 01:03:37.50 & $-$72:01:32.9  &  SXP1323 \nl
619:  ~~~~~010337.50$-$720132.9  & ~~~~~- - - - - &~~~~~~~~1393 & \nl
620: RX~J0105.1$-$7211 & 01:05:02.40 & $-$72:10:55.9  & SXP3.34  \nl
621:  ~~~~~010502.51$-$721055.9  & ~~~~~206.16890.17 &~~~~~~~~1506 & \nl
622: \enddata
623: 
624: \tablenotetext{a}{Coordinates are the MACHO position of the optical star,
625: except for RX~J0103.6$-$7201. }
626: 
627: \tablenotetext{b}{From catalogue of SMC H$\alpha$ emission-line stars
628: by Meyssonnier \& Azzopardi (1993). } 
629: 
630: \end{deluxetable}
631: 
632: \newpage
633: 
634: \begin{deluxetable}{lccl}
635: \tablenum{2}
636: \tablecaption{Photometric Periods for 
637: SMC Be/Neutron-Star Systems}
638: \tablehead{  
639: \colhead{System Names} &
640: \colhead{P$_{orb}$} &
641: \colhead{NRP\tablenotemark{a}} &
642: \colhead{References} \\
643: \colhead{} &
644: \colhead{(days)} & 
645: \colhead{(days)} & 
646: \colhead{}
647: }
648: \startdata
649: XMMU~J005152.2$-$731033 (SXP172)\tablenotemark{b} & 69.9$\pm$0.6 & --- & 
650: Laycock et al.\tablenotemark{c} \nl
651: RX~J0051.8$-$7231 (SXP8.88) & 33.4$\pm$1.0: & --- & Corbet et 
652: al.\tablenotemark{d} \nl
653: XMMU~J005605.2$-$722200 (SXP140) & 197$\pm$5: & --- & \nl 
654: CXOU~J010102.7$-$720658 (SXP304) &  520$\pm$12: & 0.26 & Schmidtke \& 
655: Cowley\tablenotemark{e} \nl
656: RX~J0103.6$-$7201 (SXP1323) & --- & 0.98; 0.88; 0.41 & \nl
657: RX~J0105.1$-$7211 (SXP3.34) & --- & 1.099; 0.980 & Coe et 
658: al.\tablenotemark{f} \nl
659: \enddata
660: 
661: \tablenotetext{a}{NRP = nonradial pulsations of the Be star. } 
662: 
663: \tablenotetext{b}{Optical counterpart is Star 1 in field of SMC~25 
664: (Cowley et al.\ 1997). }
665: 
666: \tablenotetext{c}{Laycock et al.\ (2005) find P$_{orb}$=67$\pm$5 days from
667: X-ray outbursts.}
668: 
669: \tablenotetext{d}{Corbet et al.\ (2004) give P$_{orb}$=28.0$\pm$0.3 days from
670: X-ray outbursts.}
671: 
672: \tablenotetext{e}{Schmidtke \& Cowley (2006) give further information about
673: this source.}
674: 
675: \tablenotetext{f}{Coe et al.\ (2005) give P$_{orb}$=11.09 days, which
676: is an alias of the strong NRP at 1.099 days (Schmidtke \& Cowley 2005c).} 
677: 
678: \end{deluxetable}
679: 
680: \end{document}
681: 
682: