CAL. 7T62 ALARM & CHRONOGRAPH

TIME/CALENDAR

Hour, minute and small second hands Date displayed in numerals

STOPWATCH

Measures up to 60 minutes in 1/5 second increments. Split time measurement

SINGLE-TIME ALARM

Rings only once at a designated time within the coming 12 hours.
DISPLAY AND CROWN/BUTTONS
STOPWATCH minute hand Hour hand STOPWATCH 1/5-second hand Minute hand

a Small second hand

ALARM hour hand ALARM minute hand b: First click c: Second click

a: Normal position

SCREW DOWN CROWN
[for models with screw down crown]

Unlocking the crown

1 Turn Crown counterclockwise until you no longer feel the threads turning. 2 Crown can be pulled out.

Locking the crown

1 Push Crown back in to normal position. 2 Turn Crown clockwise while pressing it lightly until tight.
SETTING THE TIME AND ADJUSTING THE STOPWATCH HAND POSITION
This watch is so designed that the following are all made with the crown at the second click position: 1) main time setting 2) alarm hand adjustment 3) stopwatch hand position adjustment Once the crown is pulled out to the second click, be sure to check and adjust 1) and 2) at the same time. If needed, 3) should also be adjusted then.
Pull out to second click when the second hand is at the 12 oclock position.

1) MAIN TIME SETTING

Hour hand Minute hand
Turn to set the hour and minute hands.

Small second hand

* It is recommended that the hands be set to the time a few minutes ahead of the current time, taking into consideration the time required to set the ALARM hands and to adjust the STOPWATCH hand position if necessary.

2) ALARM HAND ADJUSTMENT

Set the ALARM hands to the time the main time hands indicate.

Hour hand

Minute hand
Press repeatedly to set ALARM hands to the time indicated by the main time hands.
* ALARM hands move quickly if button B is kept pressed.
ALARM hour hand ALARM minute hand
3) STOPWATCH HAND POSITION ADJUSTMENT
If the STOPWATCH hands are not in the 0 position, follow the procedure below to set them to the 0 position.

Press for 2 seconds.

* STOPWATCH minute hand turns a full circle.
STOPWATCH 1/5-second hand

STOPWATCH minute hand

Press repeatedly to set STOPWATCH minute hand to the 0 position.
* The hand moves quickly if button B is kept pressed.
* STOPWATCH 1/5-second hand turns a full circle.
Press repeatedly to set STOPWATCH 1/5-second hand to the 0 position.
ALARM and STOPWATCH hands can be readjusted in the following order by pressing button A for 2 seconds.

ALARM hands

Advance ( 12 hours. )

( Turns a ) full circle.

* After all the adjustments are completed, check that the main time and alarm hands indicate the same time.
Push back in to normal position in accordance with a time signal.

SETTING THE DATE

Before setting the date, be sure to set the main time.
1 Pull out to first click. 2 Turn clockwise until the desired date appears. 3 Push back in to normal position.
The stopwatch can measure up to 60 minutes in 1/5-second increments. After 60 minutes, it will start counting again from 0 repeatedly up to 12 hours.
STOPWATCH minute hand STOPWATCH 1/5-second hand

A Start / Stop / Restart

B Reset /

Split /Split release

Before using the stopwatch, be sure to check that the crown is set at the normal position and that the STOPWATCH hands are reset to the 0 position.
* If the STOPWATCH hands do not return to the 0 position when the stopwatch is reset to 0, follow the procedure in SETTING THE TIME AND ADJUSTING THE STOPWATCH HAND POSITION.

Split time measurement

SPLIT RELEASE
Measurement of two competitors

2ND COMPETITOR FINISHES

FINISH TIME OF 1ST COMPETITOR
The alarm can be set to ring only once at a designated time within the coming 12 hours. The alarm time can be set in one minute increments.

ALARM TIME SETTING

Before using the alarm, check that the ALARM hands are adjusted to the current time. (See SETTING THE TIME AND ADJUSTING THE STOPWATCH HAND POSITION)
Pull out to first click. Press repeatedly to set the desired alarm time.

HOW TO STOP THE ALARM

At the designated time the alarm rings for 20 seconds, and it is automatically disengaged as it stops. To stop it manually, press button A or B.
HOW TO CANCEL THE ALARM TIME YOU HAVE SET
CROWN Pull out to first click. Press and hold until ALARM hands stop and indicate the current time. Push back in to normal position. 13
Push back in to normal position.
* The alarm is automatically engaged.

TACHYMETER

[for models with tachymeter scale on the dial]
To measure the hourly average speed of a vehicle

[ Ex. 1 ]

STOPWATCH second hand: 40 seconds Tachymeter scale: (tachymeter scale figure) x 1 (km or mile) = 90 km/h or mph
1 Use the stopwatch to determine how
many seconds it takes to go 1 km or 1 mile.
2 Tachymeter scale indicated by
STOPWATCH second hand gives the average speed per hour.
Tachymeter scale can be used only when the time required is less than 60 seconds. Ex. 2: If the measuring distance is extended to 2 km or miles or shortened to 0.5 km or miles and STOPWATCH second hand indicates 90 on tachymeter scale:
90 (tachymeter scale figure) x 2 (km or mile) = 180 km/h or mph 90 (tachymeter scale figure) x 0.5 (km or mile) = 45 km/h or mph
To measure the hourly rate of operation
[ Ex. 1 ] 1 Use the stopwatch to measure the STOPWATCH second hand: time required to complete 1 job. 20 seconds 2 Tachymeter scale indicated by
Tachymeter scale: (tachymeter scale figure) x 1 job = 180 jobs/hour Ex. 2: If 15 jobs are completed in 20 seconds: 180 (tachymeter scale figure) x 15 jobs = 2700 jobs/hour
STOPWATCH second hand gives the average number of jobs accomplished per hour.
NOTES ON OPERATING THE WATCH
When the stopwatch is or has been measuring, if the crown is pulled out to the second click, it will automatically reset the STOPWATCH hands to 0. If the alarm has been set and the crown is pulled out to the second click, the ALARM hands will turn to indicate the current time. [MAIN TIME SETTING] When setting the hour hand, be sure to check that AM/PM is correctly set. The watch is so designed that the date changes once in 24 hours. When setting the minute hand, first advance it 4 to 5 minutes ahead of the desired time and then turn it back to the exact minute.
It is necessary to adjust the date at the end of February and 30-day months. Do not set the date between 9:00 p.m. and 1:00 a.m. Otherwise, the date may not change properly. Do not press button B when the crown is at the first click position, as this will move the ALARM hands.
Restart and stop of the stopwatch can be repeated by pressing button A. Measurement and release of split time can be repeated by pressing button B.

The single-time alarm cannot be set for a time more than 12 hours ahead of the current time. While you keep button B pressed to advance the ALARM hands quickly, they stop when they indicate the current time and the alarm is disengaged. In that case, release button B, and then, press and hold the button again to set the ALARM hands to the desired time. While the crown is at the normal position, the ALARM hands indicate the current time when the alarm is disengaged and the designated alarm time when it is engaged. While the stopwatch is measuring, the alarm rings differently than usual. However, this is not a malfunction. While the alarm is ringing, pressing button A or B will only stop the alarm, and no stopwatch operation can be made. To correct the alarm time you have set, follow the procedure described in ALARM TIME SETTING.

BATTERY CHANGE 3

Battery life : Approx. 3 years Battery : SEIKO SR927W
If the stopwatch is used for more than 2 hours a day and/or the alarm rings for more than 20 seconds a day, the battery life may be less than the specified period. As the battery is inserted at the factory to check the function and performance of the watch, its actual life once in your possession may be less than the specified period. When the battery expires, be sure to replace it as soon as possible to prevent any malfunction. After the battery is replaced with a new one, set the time/calendar and alarm and adjust the stopwatch hand position.

Battery life indicator

When the battery nears its end, the small second hand moves at two-second intervals instead of normal one-second intervals. In that case, have the battery replaced with a new one as soon as possible. * While the small second hand is moving at two-second intervals, the alarm will not ring even if it is engaged. This is not a malfunction.
* The watch remains accurate while the small second hand is moving at two-second intervals.

WARNING

Do not remove the battery from the watch. If it is necessary to take out the battery, keep it out of the reach of children. If a child swallows it, consult a doctor immediately. Never short-circuit, tamper with or heat the battery, and never expose it to fire. The battery may burst, become very hot or catch fire.

CAUTION

The battery is not rechargeable. Never attempt to recharge it, as this may cause battery leakage or damage to the battery.
TO PRESERVE THE QUALITY OF YOUR WATCH

WATER RESISTANCE

Non-water resistant
If the watch becomes wet, have it checked by an AUTHORIZED PULSAR DEALER or SERVICE CENTER.
Water resistant 5/10/15/20 bar
Before using in water, be sure the crown is pushed in completely. Do not operate the crown and buttons when the watch is wet or in water. If used in sea water, rinse the watch in fresh water and dry it completely. When taking a shower with the water resistant 5 bar watch, or taking a bath with the water resistant 10, 15 or 20 bar watch, be sure to observe the following: * Do not operate the crown or push the buttons when the watch is wet with soapy water or shampoo. * If the watch is left in warm water, a slight time loss or gain may be caused. This condition, however, will be corrected when the watch returns to normal temperature.

10/15/20

bar WR
* Pressure in bar is a test pressure and should not be considered as corresponding to actual diving depth since swimming movement tends to increase the pressure at a given depth. Care should also be taken on diving into water. ** We recommend that you wear a PULSAR Divers Watch for scuba diving.

TEMPERATURES

Your watch works with stable accuracy within a temperature range of 5 -10C C and 35 C (41 F and 95 F). Temperatures over 60 C (140 F) may cause battery leakage or shorten the battery life. Do not leave your watch in very low temperatures below 10 C (+14 F) for a long time since the cold may cause a slight time loss or gain. However, the above conditions will be corrected when the watch returns to normal temperature.

MAGNETISM

Your watch will be adversely affected by strong magnetism. Keep it away from close contact with magnetic objects.
CARE OF CASE AND BRACELET
To prevent possible rusting of the case and bracelet caused by dust, moisture and perspiration, wipe them periodically with a soft dry cloth.

SHOCKS & VIBRATION

PERIODIC CHECK
Light activities will not It is recommended that the affect your watch, but be watch be checked once 2-3 careful not to drop your every 2 to 3 years. Have Years watch or hit it against hard your watch checked by an surfaces, as this may AUTHORIZED PULSAR cause damage. DEALER or SERVICE CENTER to ensure that the case, crown, buttons, gasket and crystal seal remain intact. CHEMICALS Be careful not to expose the watch to solvents, mercury, cosmetic spray, detergents, adhesives or paints. Otherwise, the case, bracelet, etc. may become discolored, deteriorated or damaged.
PRECAUTION REGARDING CASE BACK PROTECTIVE FILM
If your watch has a protective film and/or a sticker on the case back, be sure to peel them off before using your watch.

The Astrophysical Journal, 627:910914, 2005 July 10
# 2005. The American Astronomical Society. All rights reserved. Printed in U.S.A.
THE X-RAY POSITION AND OPTICAL COUNTERPART OF THE ACCRETION-POWERED MILLISECOND PULSAR XTE J1814338
Miriam I. Krauss, Zhongxiang Wang, Allyn Dullighan,1 Adrienne M. Juett,2 David L. Kaplan, and Deepto Chakrabarty
Department of Physics and Center for Space Research, Massachusetts Institute of Technology, Cambridge, MA 02139; miriam@space.mit.edu, wangzx@space.mit.edu, allyn@space.mit.edu, ajuett@space.mit.edu, dlk@space.mit.edu, deepto@space.mit.edu

Marten H. van Kerkwijk

Department of Astronomy and Astrophysics, University of Toronto, 60 St. George Street, Toronto, ON M5S 3H8, Canada; mhvk@astro.utoronto.ca
Danny Steeghs and Peter G. Jonker3,4
Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138; dsteeghs@cfa.harvard.edu, p.jonker@sron.nl

and Craig B. Markwardt5

Department of Astronomy, University of Maryland, College Park, MD 20742; craigm@milkyway.gsfc.nasa.gov Receivvd 2005 January 28; accepted 2005 March 29 e
ABSTRACT We report the precise optical and X-ray localization of the 3.2 ms accretion-powered X-ray pulsar XTE J1814338 with data from the Chandra X-Ray Observatory as well as optical observations conducted during the 2003 June discovery outburst. Optical imaging of the eld during the outburst of this soft X-ray transient reveals an R 18 star at the X-ray position. This star is absent (R > 20) from an archival 1989 image of the eld and brightened during the 2003 outburst, and we therefore identify it as the optical counterpart of XTE J1814338. The s best source position derived from optical astrometry is R:A: 18h 13m 39:04, decl: 0 22B3 (J2000). The featureless X-ray spectrum of the pulsar in outburst is best t by an absorbed power law (with photon index 1:41 0:06) plus blackbody (with kT 0:95 0:13 keV) model, where the blackbody component contributes approximately 10% of the source ux. The optical broadband spectrum shows evidence for an excess of infrared emission with respect to an X-ray heated accretion disk model, suggesting a signicant contribution from the secondary or from a synchrotron-emitting region. A follow-up observation performed when XTE J1814338 was in quiescence reveals no counterpart to a limiting magnitude of R 23:3. This suggests that the secondary is an M3 V or later-type star and therefore very unlikely to be responsible for the soft excess, making synchrotron emission a more reasonable candidate. Subject headingg: binaries: close pulsars: individual (XTE J1814338) stars: neutron X-rays: binaries s
1. INTRODUCTION It has long been believed that millisecond radio pulsars are the spun-up products of sustained mass transfer onto neutron stars in low-mass X-ray binaries (e.g., Bhattacharya & van den Heuvel 1991). Their presumed immediate progenitors, accretion-powered millisecond X-ray pulsars, proved elusive for many years, but six such systems are now known: SAX J1808.43658 (Wijnands & van der Klis 1998; Chakrabarty & Morgan 1998), XTE J( Markwardt et al. 2002), XTE J0929314 (Galloway et al. 2002), XTE J1807294 ( Markwardt et al. 2003a), XTE J1814

1 Current address: Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA 02420. 2 Current address: Department of Astronomy, University of Virginia, Charlottesville, VA 22903. 3 Chandra Fellow. 4 Current address: SRON National Institute for Space Research, Sorbonnelaan 2, 3584 CA Utrecht, Netherlands. 5 Also at: Laboratory for High Energy Astrophysics, NASA Goddard Space Flight Center, Greenbelt, MD 20771.
314 (Markwardt & Swank 2003; Strohmayer et al. 2003), and IGR J00291+5934 (Galloway et al. 2005). In addition, 13 accreting neutron stars also show millisecond oscillations during thermonuclear X-ray bursts (see Strohmayer & Bildsten 2005 for a review). These systems are now understood as nuclear-powered millisecond pulsars, with the burst oscillations tracing the pulsar spin (Strohmayer & Markwardt 2002; Chakrabarty et al. 2003). The soft X-ray transient XTE J1814314 (l 358N7, b 7N6) was discovered in outburst on 2003 June 5 during scans of the central Galactic plane with the Rossi X-Ray Timing Explorer (RXTE; Markwardt & Swank 2003). The outburst lasted for approximately 55 days and had a peak 210 keV ux of around 13 mcrab. RXTE observations also established the source as a 314 Hz (3.2 ms) accretion-powered X-ray pulsar (Markwardt & Swank 2003) in a 4.3 hr binary with a main-sequence companion of at least 0.17 M (using the mass function of 0.002016 M in Markwardt et al. 2003b, assuming a neutron star mass of 1.4 M). Over two dozen thermonuclear X-ray bursts with millisecond oscillations at the spin frequency were detected from XTE Jduring the 2003 June outburst (Strohmayer et al. 2003). These 910
POSITION AND COUNTERPART OF XTE J1814338 burst oscillations had the particularly interesting characteristic of containing signicant harmonic content, which allowed Bhattacharyya et al. (2005) to constrain the neutron star as well as orbital parameters. One of these bursts showed evidence for photospheric radius expansion, allowing Strohmayer et al. (2003) to infer a source distance of 8:0 1:6 kpc.6 We obtained a brief observation of the source on 2003 June 20 with the Chandra X-Ray Observatory for the purpose of measuring its position, and we used this position to identify the optical counterpart (Krauss et al. 2003). Subsequent optical spectroscopy revealed strong emission lines of H and He, including double-peaked H emission, indicative of an interacting binary (Steeghs 2003). In this paper, we present a detailed report on our Chandra and optical observations of XTE J1814338. In x 2, we present X-ray and optical imaging and the precise localization and ux measurements of the optical counterpart. We analyze the X-ray spectrum in x 3, and in x 4 we discuss the implications of the current data in our understanding of the physical parameters of XTE J1814338. 2. X-RAY AND OPTICAL IMAGING We observed XTE J1814338 with Chandra for 9.7 ks on 2003 June 20 using the High Energy Transmission Grating Spectrometer (HETGS) with the spectroscopic array of the Advanced CCD Imaging Spectrometer (ACIS-S). HETGS comprises two sets of transmission gratings: the medium-energy gratings (MEGs), with a range of 2.(0.45.0 keV), and the high-energy gratings (HEGs), with a range of 1.(0.810 keV). The HETGS spectra are imaged by ACIS-S, an array of six CCD detectors. The HETGS/ACIS-S combination provides both an undispersed (zeroth-order) image and dispersed spectra from the gratings. The spatially overlapping spectral orders are sorted using the intrinsic energy resolution of the ACIS-S CCDs. The rst-order MEG (HEG) spectrum has a spectral resolution of k 0:023 (0.012) 8 FWHM. All data processing was done with the CIAO analysis package7 (ver. 2.3). We summed the dispersed rst-order events in 500 s time bins to create an X-ray light curve, which we searched for signs of orbital modulation. We note that the observation spanned only 60% of the 4.3 hr orbital period, but did cover the time period during which the neutron star was behind the secondary. The data were consistent with a constant count rate of 4:6 0:1 counts s1, and we did not detect any evidence of an X-ray eclipse. We corrected the observation aspect to be consistent with the calibration available as of 2004 March 22. No sources other than XTE J1814338 were detected in the eld. Due to the high source count rate, the zeroth-order image is over 75% piled up, which suppresses counts in the image core and results in a characteristic doughnut-shaped point-spread function (see, e.g., Davis 2001). However, this did not affect our ability to use the zeroth-order image to obtain a precise positional measurement with the CIAO tool wavdetect. In order to account for the pileup, we used large wavelet scales (8 and 12 pixels), which are not sensitive to the core of the point-spread function. The best-t s X-ray position of XTE J1814338 was R:A: 18h 13m 39:02, decl: 0 22B3 (J2000.0) with a 90% condence radius of 0B6.

This distance estimate should be viewed with some caution, since the burst it is based on has some features that suggest it may not have been Eddington limited. If the burst was not actually Eddington limited, Lburst < LEdd , and the distance to XTE J1814338 may be less than 8:0 1:6 kpc. We therefore consider 8.0 kpc to be an upper limit for the source distance. 7 See http://cxc.harvard.edu /ciao /.
TABLE 1 Optical Magnitudes of XTE J1814338 Observation Time ( UT ) June 6 10:25.. June 7a 09:57.. June 21b 03:32. June 21 05:06. June 21 05:13. June 21 06:26. June 24b 06:39. March 15 07:47.
B 18.96 18.71 18.77 18.64 18.62 18.61 18.75.
V 18.59. 18.48 18.39 18.34 18.33 18.48.
R 18.26 18.05 18.33 18.16 18.21 18.16 18.29 >23.3c
I.. 17.47 17.37 17.42 17.35 17.38.
a Calibration errors are 0.06 mag; relative uncertainty between these two nights is 0.01 mag. b Calibration errors are 0.02 mag. c XTE J1814338 was not detected; this value is the 3 limiting magnitude.
The precise X-ray source position facilitated the identication of an optical counterpart. We obtained BVR images (Harris broadband lter set, which closely approximates Johnson-Cousins) of the XTE J1814338 eld on 2003 June 6 and BR images on 2003 June 7 using the low-dispersion survey spectrograph (LDSS-2) camera on the 6.5 m Magellan/Clay telescope at the Las Campanas Observatory in Chile (eld of view of 2A5 ; 2A5 with a scale of 0B38 pixel1; the seeing on both nights was approximately 0B8). All frames were debiased and at-elded using the IRAF package. An astrometric solution was derived using the USNOB1.0 catalog (Monet et al. 2003), giving a standard deviation of 0B35. Using this solution, we nd that there is one optical source within the Chandra error circle. This source is not present in a 1989 Digitized Sky Survey8 image to a limiting magnitude of R $ 20; we therefore identify this source as the optical counterpart. Further renement of the optical position is described below. Flux calibration for the rst night was done using the photometric standard star Mark A (Landolt 1992), and calibration for the second night was made by tting several in-eld stars to the previous nights measurements; these values are shown in Table 1. Note that these uxes replace the incorrectly calibrated values reported in Krauss et al. (2003). The optical counterpart brightened slightly over the course of the two days, paralleling the increase in X-ray ux over the same time period as measured by RXTE. On 2003 June 21 and 24 we obtained additional BVRI images (Johnson-Cousins lter set), again with the 6.5 m Magellan / Clay telescope, this time using the Magellan Instant Camera ( MagIC; eld of view of 2A4 ; 2A4 with a scale of 0B069 pixel1; the seeing on both nights was approximately 0B6; see Fig. 1). Again, the frames were debiased and at-elded with the IRAF package. We derived astrometric solutions using 75 sources from the 2MASS catalog (Cutri et al. 2001) and 49 sources from the USNO-B1.0 catalog. The t using the 2MASS sources is signicantly better than the one using the USNO-B1.0 sources and yields rms residuals of around 0B07 in each coordinate. From s this, we derive an optical position of R:A: 18h 13m 39:04, decl: 0 22B3 (J2000.0) with a 90% condence radius of 0B2 (the uncertainty is based on the astrometric accuracy of 2MASS; see, e.g., Cutri et al. 2001). This position is 0B25 from the Chandra-derived X-ray position, well within the 0B6 Chandra error circle. This coincidence, the optical sources long-term as well as night-to-night variability, its blue color, and the emission lines seen by Steeghs (2003) argue strongly that it is the counterpart of XTE J1814338.

See http://archive.stsci.edu /dss/index.html.

KRAUSS ET AL.

Vol. 627
Fig. 1.MagIC I-band image of the XTE J1814338 eld from 2003 June 21. The I 17:4 optical counterpart is marked near the center of the image. North is up, and east is to the left.
The MagIC images were ux calibrated using the standard star Mark A2 (Landolt 1992), and the magnitudes of the counterpart are presented in Table 1. The counterpart on these later dates is on average a bit brighter than in the earlier observations, again agreeing with the observed X-ray ux of XTE J1814 338, which brightened from around 10 mcrab to around 12 mcrab before the second set of optical measurements were performed. See Figure 2 for a plot of all optical data. Finally, on 2004 March 15, while the source was in quiescence, we obtained an R-band image of the XTE J1814338 eld with the European Southern Observatory ( ESO) MultiMode Instrument on the 3.5 m New Technology Telescope at the La Silla Observatory in Chile (eld of view of 6A2 ; 6A2 with a scale of 0B167 pixel1; the seeing was approximately 1B0). The counterpart was not detected in this observation with a 3 limiting magnitude of R 23:3. 3. X-RAY SPECTROSCOPY We extracted separate X-ray spectra for the MEG and HEG data, co-added the 1 orders, and constructed the corresponding response les (auxiliary response les [ARFs] and redistribution matrix les [ RMFs]). We used the contamarf tool9 to correct the ARF for a decrease in low-energy sensitivity due to contamination on the ACIS CCDs (see, e.g., Marshall et al. 2003). The count rate in the rst-order MEG ( HEG) spectrum was 2.9 (1.5) counts s1. To improve statistics, we grouped the spectra such that there was a minimum of 100 counts per bin and assumed Poisson errors. Background spectra were created using the script tg_bkg and subtracted prior to tting. All tting was performed using the XSPEC version 11.2 spectral analysis package. Since the zeroth-order image is severely piled up, all spectral analysis was done on the dispersed rst-order spectra, which are not affected by pile-up. We t the MEG and HEG spectra simultaneously over the total energy range 0.510 keV, including a normalization factor, which was allowed to vary between the two instruments. We found the spectra to be best t by an absorbed power law plus blackbody model, where the equivalent

Information about contamarf can be found at http://space.mit.edu /CXC / analysis /ACIS_Contam /ACIS_Contam.html.
Fig. 2.Optical data for XTE J1814338. Measurements taken on 2003 June 6 and 7 are plotted as squares, and those from the nights of 2003 June 21 and 24 are shown as triangles (errors are on the order of the symbol size). The observed uxes were dereddened assuming a Galactic extinction of AV 0:71. The emission predicted by an X-rayheated accretion disk model is plotted as a dashed line. For this model, we used a source distance of 8 kpc with Galactic extinction and minimized 2 by setting cos i 0:6. Note that since these parameters are degenerate, the inclination angle tted here is not well constrained. The I-band ux (9000 8) lies well above the model prediction.
hydrogen column density ( NH ) is xed to the Galactic value of 1:63 ; cm 2 ( Dickey & Lockman 1990). We note, however, that the ux in the blackbody component is only about 10% of that in the power-law component. We also t an absorbed power law with NH allowed to vary, but the value did not deviate signicantly from the Galactic value, justifying ts where this parameter is frozen. The results of spectral tting are summarized in Table 2. Given that the tted absorption is consistent with a column equal to the integrated Galactic value, it is likely that XTE J1814338 lies at least 500 pc out of the Galactic plane. At the Galactic latitude of b 7N6, this implies a distance of k3.8 kpc. No signicant spectral features ( lines or edges) were observed. To quantify this, we searched the spectral residuals for Gaussian features of FWHM equal to 800 km s1, to match the velocity seen in the H emission (Steeghs 2003). The 3 upper limits are approximately 0.at and 0.at 2.5 8. To date, high-resolution X-ray spectroscopy has been obtained for four MSPs: XTE J1751305 (Miller et al. 2003), XTE J(Juett et al. 2003), XTE J1807294 (Campana et al. 2003), and XTE J1814338. None have shown signicant intrinsic spectral features, and the continua are generally well t by an absorbed power law plus blackbody. 4. DISCUSSION We have identied the optical counterpart of XTE J1814 334. The combined optical and X-ray observations allow us

No. 2, 2005

POSITION AND COUNTERPART OF XTE J1814338
TABLE 2 Spectral Fit Parametersa Blackbody Power Law kTin ( keV ).. 0.95 0.13
NH (1021 cm2) 1.67 0.17... 1.63 (xed)... 1.63 (xed)...

a b c d

1.36 0.03 1.35 0.02 1.41 0.06
A1b 3.71 0.15 3.68 0.05 3.32 0.11

Fluxc 3.1 3.1 2.6

Rkmd.. 1.6 0.3

Fluxc.. 0.3

2 (dof ) 1.41 (405) 1.41 (406) 1.29 (404)
All errors quoted are the 90% condence range. The amplitude of the power law is the ux at 1 keV in units of 102 photons keV1 cm2 s1. Fluxes are for the energy range 0.510 keV and are in units of 1010 ergs cm2 s1. Blackbody radius assuming a distance to XTE J1814338 of 8 kpc.
to place several constraints on the system parameters. The massradius relation for a low-mass Roche lobelling companion in a = 4.27 hr binary is Rc 0:28(Mc /0:01 M )R (see, e.g., Frank et al. 2002). Given the measured neutron star orbital parameters (Markwardt et al. 2003b), the lack of an X-ray eclipse thus restricts the binary inclination (dened as the angle between the line of sight and the orbital angular momentum vector) to i < 77 (cos i > 0:22) for a Roche lobelling companion. In Figure 3, we compare the companions mass-radius relation with the theoretical relation for low-mass hydrogen main-sequence stars (Tout et al. 1996). An ordinary hydrogen-rich companion is consistent with the required relation for a mass Mc % 0:54 M and i % 21 (cos i % 0:93), although this value has a small a priori probability of 7% for an isotropic sample of binary inclinations. We note, however, that bloating of a hydrogen-rich companion owing to X-ray heating could allow a somewhat less massive companion to ll its Roche lobe (Tout et al. 1989), permitting a slightly larger inclination angle with a higher a priori probability. A hydrogen main-sequence companion is also consistent with the observed H and He lines in the optical spectra (Steeghs 2003), whereas a white dwarf companion can be excluded as too small to ll its Roche lobe for any plausible donor mass.
Fig. 3.Mass-radius relationship for a Roche lobelling companion of XTE J1814338 (solid curve) and low-mass main-sequence stars (dashed curve), with corresponding inclination angles indicated. The dashed curve is based on the analytic mass-radius function presented in Tout et al. (1996). The intersection of the curves suggests a companion mass of %0.5 M.
We can check the consistency of our above methods using the brightness of the optical counterpart. Shahbaz & Kuulkers (1998) found that the magnitude of the optical outburst and the orbital period of soft X-ray transients are strongly correlated. Using the relation they derive gives V 9:5 1:1 for XTE J1814338, and, since the counterpart in outburst is V % 18:4, a quiescent magnitude of Vq % 27:9. Assuming AV ! 0:7 due to Galactic dust along the line of sight (Schlegel et al. 1998), and a minimum distance of 6.4 kpc (corresponding to the lower limit of the distance estimate derived from the radius expansion burst; see Strohmayer et al. 2003), we estimate a limit on the absolute magnitude of a main-sequence companion of MV < 13:2. We used the empirical mass-luminosity relation presented in Delfosse et al. (2000) to determine a minimum companion mass of %0.2 M (corresponding to i < 60 and cos i k 0:50). This companion mass is signicantly lower than the value derived from the perioddensity relationship and would imply that the companion star is signicantly bloated due to X-ray heating. However, we note that increasing the assumed distance to XTE J1814338 will increase the minimum companion mass and decrease the maximum inclination angle. It is also instructive to consider what we can infer about the system, given the R > 23:3 mag limit of the companion in quiescence. We again assume AV 0:7 and infer a value of AR 0:5 (given AV /EB V 3:1; see, e.g., Cardelli et al. 1989). We use the upper limit on the distance derived from the radius expansion burst of 9.6 kpc to give an absolute magnitude of MR 7:9. This corresponds to a main-sequence spectral type of later than M1 V and a mass of M P 0:5 M (corresponding to i > 22 and cos i P 0:93). Using the harmonic properties of the burst oscillations, Bhattacharyya et al. (2005) were able to derive 90% condence intervals of 26 < i < 50 (0:90 > cos i > 0:64). This range is in agreement with our determination of the inclination angle from the magnitude limit of the optical counterpart (i > 22 ) and lack of X-ray eclipse (i < 77 ). However, their lower limit of i > 26 is a bit higher than the value we calculate given the mass-radius relation for an ordinary main-sequence companion of 21 , again suggesting that the companion may be bloated as a result of X-ray heating. We would expect the optical emission from XTE J1814338 to originate from the combination of an X-rayheated accretion disk and companion. A simple X-rayheated accretion disk model ( Vrtilek et al. 1990; Chakrabarty 1998; Wang et al. 2001) is able to account for the observed BVR magnitudes for a wide range of plausible parameters, although the lack of data at bluer wavelengths precludes a well-constrained t. However, this model is unable to account for the I-band data, which are systematically brighter than predicted. ( For a plot of the optical

KRAUSS ET AL. tively). An alternative possibility is that the I-band excess could originate from synchrotron emission related to the outburst (see, e.g., Fender 2001). Future observations of XTE J1814338 in outburst spanning a large spectral range, including measurements in the visible as well as at IR and radio wavelengths, would further our understanding of the geometry and emission mechanisms of XTE J1814338. We thank Harvey Tananbaum for providing us with Directors Discretionary time on the Chandra X-Ray Observatory. We thank the mission planning team for coordinating and executing a rapid response to our target of opportunity request, and Joy Nichols for processing and checking the data extremely quickly. We thank Cole Miller for useful discussions, Jake Hartman for his assistance and helpful comments, and the anonymous referee for useful comments. D. S. acknowledges support through a Smithsonian Astrophysical Observatory Clay Fellowship. P. G. J. is supported by NASA through Chandra Postdoctoral Fellowship grant PF3-40027 awarded by the Chandra X-Ray Center, which is operated by the Smithsonian Astrophysical Observatory for NASA, under contract NAS8-39073. This work was supported in part by NASA, under contract NAS 8-01129.
data with a representative X-rayheated accretion disk model, see Fig. 2.) Although we do not have infrared data for XTE J1814338, the sharp increase in ux in the I-band data suggests an IR excess with respect to the X-rayheated disk model, similar to what was seen in the accretion-powered millisecond pulsar SAX J1808.43658 (Wang et al. 2001). First, let us consider the possibility that this emission arises from the companion: if we take its R-band magnitude to be R 23:3 (the limiting magnitude for its nondetection in quiescence) and assume it to be an M dwarf (in agreement with previous mass estimates), this corresponds to an I-band magnitude of I $ 21:8 (again assuming AV 0:71 to obtain AI 0:34; see, e.g., Cardelli et al. 1989). Although the companion could brighten due to X-ray heating during the outburst, in order for it to account for the excess ux, it would have to brighten to I $18, corresponding to a more than 30-fold increase in ux from quiescence. We consider this to be highly unlikely. Furthermore, if the secondary were substantially heated, its surface brightness would be highly anisotropic, and we would expect to see I-band variability at the timescale of the orbital period. The 4.3 hr orbit is well sampled on 2003 June 21, but we note that there is actually less variability at longer wavelengths (the percent rms variability in B, V, R, and I is 11%, 9%, 6%, and 4%, respec-

REFERENCES Bhattacharya, D., & van den Heuvel, E. P. J. 1991, Phys. Rep., 203, 1 Markwardt, C. B., Smith, E., & Swank, J. H. 2003a, IAU Circ., 8080, 2 Bhattacharyya, S., Strohmayer, T. E., Miller, M. C., & Markwardt, C. B. 2005, Markwardt, C. B., Strohmayer, T. E., & Swank, J. H. 2003b, Astron. Telegram, ApJ, 619, 483 164, 1 Campana, S., Ravasio, G. L., Israel, G. L., Mangano, V., & Belloni, T. 2003, Markwardt, C. B., & Swank, J. H. 2003, IAU Circ., 8144, 1 ApJ, 594, L39 Markwardt, C. B., Swank, J. H., Strohmayer, T. E., int Zand, J. J. M., & Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245 Marshall, F. E. 2002, ApJ, 575, L21 Chakrabarty, D. 1998, ApJ, 492, 342 Marshall, H. L., Tennant, A., Grant, C. E., Hitchcock, A. P., ODell, S., & Chakrabarty, D., & Morgan, E. H. 1998, Nature, 394, 346 Plucinsky, P. P. 2003, in X-Ray and Gamma-Ray Instrumentation for AsChakrabarty, D., Morgan, E. H., Muno, M. P., Galloway, D. K., Wijnands, R., tronomy XIII, ed. K. A. Flanagan & O. H. Siegmund ( Bellingham: SPIE ), van der Klis, M., & Markwardt, C. B. 2003, Nature, 424, Cutri, R. M., et al. 2001, Explanatory Supplement to the 2MASS Second Miller, J. M., et al. 2003, ApJ, 583, L99 Incremental Data Release (Pasadena: IPAC), http://www.ipac.caltech.edu/ Monet, D. G., et al. 2003, AJ, 125, 984 2mass/releases/second/doc/expslup.html Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525 Davis, J. E. 2001, ApJ, 562, 575 Shahbaz, T., & Kuulkers, E. 1998, MNRAS, 295, L1 Delfosse, X., Forveille, T., Segransan, D., Beuzit, J.-L., Udry, S., Perrier, C., & Steeghs, D. 2003, IAU Circ., 8155, 2 Mayor, M. 2000, A&A, 364, 217 Strohmayer, T. E., & Bildsten, L. 2005, in Compact Stellar X-Ray Sources, ed. Dickey, J. M., & Lockman, F. J. 1990, ARA&A, 28, 215 W. H. G. Lewin & M. van der Klis (Cambridge: Cambridge Univ. Press), in Fender, R. 2001, Ap&SS Suppl., 276, 69 press (astro-ph/0301544) Frank, J., King, A., & Raine, D. 2002, Accretion Power in Astrophysics (3rd Strohmayer, T. E., & Markwardt, C. B. 2002, ApJ, 577, 337 ed., Cambridge: Cambridge Univ. Press) Strohmayer, T. E., Markwardt, C. B., Swank, J. H., & int Zand, J. 2003, ApJ, Galloway, D. K., Chakrabarty, D., Morgan, E. H., & Remillard, R. A. 2002, 596, L67 ApJ, 576, L137 Tout, C. A., Eggleton, P. P., Fabian, A. C., & Pringle, J. E. 1989, MNRAS, 238, Galloway, D. K., Markwardt, C. B., Morgan, E. H., Chakrabarty, D., & 427 Strohmayer, T. E. 2005, ApJ, 622, L45 Tout, C. A., Pols, O. R., Eggleton, P. P., & Han, Z. 1996, MNRAS, 281, 257 Juett, A. M., Galloway, D. K., & Chakrabarty, D. 2003, ApJ, 587, 754 Vrtilek, S. D., Raymond, J. C., Garcia, M. R., Verbunt, F., & Hasinger, G. 1990, Krauss, M. I., Dullighan, A., Chakrabarty, D., van Kerkwijk, M. H., & A&A, 235, 162 Markwardt, C. B. 2003, IAU Circ., 8154, 3 Wang, Z., et al. 2001, ApJ, 563, L61 Landolt, A. U. 1992, AJ, 104, 340 Wijnands, R., & van der Klis, M. 1998, Nature, 394, 344