1: \documentclass[apj]{emulateapj}

2: \submitted{Accepted for publication in the Astrophysical Journal October 6, 2008}

3: \shortauthors{RANGWALA ET AL.} \shorttitle{FABRY-PEROT SPECTROSCOPY OF GALACTIC BAR: KINEMATICS}

4: %\documentclass[11pt,preprint2]{aastex}

5: \begin{document}

6: %\bibliographystyle{prsty} % Choose Phys. Rev. style for bibliography

7: 

8: %new commands and definitions

9: 

10: \newcommand{\ca}         {\mbox{\rm \ion{Ca}{2}} $\lambda 8542$}

11: \newcommand{\kms}        {km s$^{-1}$}

12: 

13: 

14: \title{Fabry-Perot Absorption Line Spectroscopy of the Galactic Bar. I. Kinematics}

15: \author{Naseem Rangwala\altaffilmark{1}, T. B. Williams\altaffilmark{1,2} and K. Z. Stanek\altaffilmark{2,3}}

16: \altaffiltext{1}{Department of Physics and Astronomy, Rutgers University, 136 Frelinghuysen Road,

17: Piscataway, NJ 08854} \altaffiltext{2}{Visiting astronomer, Cerro Tololo Inter-American

18: Observatory, National Optical Astronomy Observatory, which are operated by the Association of

19: Universities for Research in Astronomy, under contract with the National Science Foundation.}

20: \altaffiltext{3}{Department of Astronomy, Ohio State University, Columbus, OH}

21: 

22: \begin{abstract}

23: We use Fabry-P\'{e}rot absorption line imaging spectroscopy to measure radial velocities using the

24: \ca \ line in 3360 stars towards three lines of sight in the Milky Way's bar: Baade's Window and

25: offset position at $(l,b) \simeq (\pm 5.0, -3.5)\degr$. This sample includes 2488 bar red clump

26: giants, 339 bar M/K-giants, and 318 disk main sequence stars. We measure the first four moments of

27: the stellar velocity distribution of the red clump giants, and find it to be symmetric and

28: flat-topped. We also measure the line-of-sight average velocity and dispersion of the red clump

29: giants as a function of distance in the bar. We detect stellar streams at the near and far side of

30: the bar with velocity difference $\gtrsim$ 30 \kms at $l = \pm 5$\degr, but we do not detect two

31: separate streams in Baade's Window. Our M-giants kinematics agree well with previous studies, but

32: have dispersions systematically lower than those of the red clump giants by $\sim 10$ \kms. For

33: the disk main sequence stars we measure a velocity dispersion of $\sim 45$ \kms \ for all three

34: lines-of-sight, placing a majority of them in the thin disk within 3.5 kpc of the Sun, associated

35: with the Sagittarius spiral arm. We measure the equivalent widths of the \ca \ line that can be

36: used to infer metallicities. We find indications of a metallicity gradient with Galactic

37: longitude, with greater metallicity in Baade's Window. We find the bulge to be metal-rich,

38: consistent with some previous studies.

39: \end{abstract}

40: 

41: \keywords{Galaxy: bulge, Galaxy: kinematics and dynamics, stars: kinematics, techniques: radial

42: velocities, techniques: spectroscopic}

43: 

44: \section{Introduction}

45: More than half of the disk galaxies in the universe have central bars \citep{freeman96}. N-body

46: simulations show that bars form naturally in disk galaxies, and play a crucial role in their

47: formation and evolution \citep{sellwood93}. There is now abundant evidence that our own Galaxy is

48: barred \citep[see review by][]{gerhard99}. One of the earliest indications came from the large

49: non-circular motions of gas in the inner galaxy measured by \citet{devac64} and subsequently by

50: \citet{liszt80}. But it was not until the 1990s that the presence of a bar in the Milky Way (MW)

51: was firmly established by the observations of a distinct peanut-shaped bulge in the COBE NIR light

52: distribution \citep{blitz91,cobe98}, a non-axisymmetric signature in the gas kinematics

53: \citep{binney91, weiner99}, OGLE photometry showing a magnitude offset of the bulge red clump

54: giants at positive and negative longitudes \citep{stanek97}, and a large optical depth of the

55: bulge to microlensing \citep{alcock00}. All these observations agree that the bar is inclined at

56: some orientation angle with respect to the Sun-Galactic center line with its the near end in the

57: first Galactic quadrant ($0\degr \leq l \leq 90\degr$).

58: 

59: Nonaxisymmetric models of the inner Galaxy have been produced in the last decade.

60: \citet{bissantz02} and \citet{dwek95} used COBE data to model the density distribution in the

61: Galactic bar. \citet{weiner99} measured the properties of the bar using gas kinematics.

62: \citet{rattenbury07b} used red clump giants to trace the mass distribution of the bar.

63: \citet{zhao96} and \citet{bissantz04} have taken initial steps towards building fully

64: self-consistent stellar dynamical models based on the available density and kinematic data.

65: Existing stellar kinematic observation are insufficient to fully constrain the models, and there

66: is significant variation in the estimates of the bar parameters such as orientation ($15\degr -

67: 45\degr$)\footnote{Although most previous investigations find bar orientation angles between 20 --

68: 30\degr, others \citep{weiner99, benjamin05,cl08} have shown that the orientation angle could be

69: as high as 35\degr -- 45\degr.} , co-rotation radius (3.5 -- 5 kpc) and pattern speed ($35 - 60$

70: \kms \ kpc$^{-1}$). There is not yet a unified quantitative dynamical model of the Galactic bar

71: that explains all the observations in a consistent manner and the properties of the Galactic bar

72: are still poorly determined.

73: \begin{figure}[t]

74: % Requires \usepackage{graphicx}

75: \includegraphics[scale=0.37]{f1.ps}\\

76: \caption{The position of our fields against the COBE NIR background. From left to right are the

77: MM7B, BW and MM5B fields. MM7B and MM5B each have four overlapping sub-fields as shown in the

78: box.} \label{cobe}

79: \end{figure}

80: 

81: The signature of the Galactic bar should also be seen in the kinematic distortions of the stellar

82: velocity field, since the orbits are no longer circular, but rather are elongated streams along

83: the bar. The stellar kinematic evidence of bulge triaxiality is still ambiguous. There has been

84: significant work done in the past on the line-of-sight velocities of M and K giants \citep{rich90,

85: sharples, terndrup, minniti92, habing} and proper motions \citep{span, sumi04, rattenbury07a,

86: clarkson08} in the bulge. But most of these radial velocity observations were limited to the low

87: extinction Baade's Window ($l \sim 0$\degr). \citet{hafner} show that measurements in Baade's

88: Window (BW) alone are insensitive to the effects of triaxiality, and a stronger kinematic

89: signature of the bar will be apparent away from $l = 0$\degr. The ideal observation would be to

90: measure stellar velocity distributions with high S/N for several lines-of-sight (LOS) in the bar.

91: A sample size of $\sim 1000$ stars per LOS would enable the measurements of the higher moments of

92: the velocity distribution and of the average radial velocity shift between the stellar streams in

93: the bar \citep{mao02,deguchi01}. In addition, combining accurate proper motions with these LOS

94: velocities would provide even stronger constraints on the current dynamical models of the Galactic

95: bar.

96: 

97: Further, the ratio of organized stellar streaming motion around the bar to random motion

98: constrains the bar's angular rotation speed (also called pattern speed) \citep{sanders82}. Since

99: bars can interact strongly with dark matter halos, transferring angular momentum into the halo and

100: slowing the bar rotation dramatically \citep{victor00}, constraining the current pattern speed of

101: the MW bar will provide a bound on the disk to halo mass ratio for the inner Galaxy.

102: 

103: The most recent bulge radial velocity survey by \citet{howard08} \citep[also see][]{rich07}

104: measured radial velocities at many positions along the major and minor axes of the bar, obtaining

105: the rotation and velocity dispersion curves. These observations have addressed the crucial problem

106: of measuring kinematics of the bar away from BW. However, their sample size along each LOS is

107: limited to about 100 stars, insufficient to measure the higher moments of the individual velocity

108: distributions or the kinematics of the stellar streams. Our goal in the current investigation is

109: to obtain a large radial velocity sample along several LOS, in order to measure these important

110: quantities.

111: 

112: The basic procedure in Fabry-P\'{e}rot (FP) imaging spectroscopy is to obtain a series of

113: narrow-band images, tuning the interferometer over a range of wavelengths covering a spectral

114: feature of interest. The data cube thus produced is analyzed to extract a short portion of the

115: spectrum of each object in the field of view. This spectrum is fit to determine the strength,

116: width and central wavelength of the spectral feature, and the level of the adjacent continuum. If

117: there are many objects of interest within the field of view, the technique can be extremely

118: efficient. FP spectroscopy has traditionally been used to measure emission features in extended

119: objects, and has been highly successful for studying the velocity fields in disk galaxies

120: \citep[e.g.][]{palunas, zanmar08} star forming regions \citep{hartigan00}, planetary nebula

121: \citep{sluis06}, and supernova remnants \citep{ghavamian03}. It is less common to employ FP

122: techniques to measure absorption spectra, but they have been used to determine the kinematics of

123: stars in a large number of globular clusters \citep{gebhardt94}, and the pattern speed of the bar

124: in the early type galaxy NGC 7079 \citep{victor04}.

125: \begin{figure}[t]

126: \centering

127:  \includegraphics[scale=.67]{f2.eps}

128:  \caption{A Fabry-P\'{e}rot image of the Baade's Window stellar field.

129:  The FOV is $4^{\prime}$ and the central wavelength is 8543\,\AA. The increasing

130:   stellar density towards the southwest edge of the field is caused by the bulge globular cluster NGC 6522.

131:  \label{stellarfield}}

132: \end{figure}

133: 

134: In the present paper we use FP absorption line spectroscopy to obtain radial velocities using the

135: \ca \ line for 2488 Red Clump Giants (RCGs), 339 red giants (M/K type) and 318 disk main sequence

136: stars, towards three LOS in the MW bar: BW and two symmetric offset positions at $(l,b) \simeq

137: (\pm 5.0, -3.5)\degr$ (see Figure \ref{cobe}). This sample is an order of magnitude larger than

138: any previous sample for a given LOS. We also measure equivalent widths of the \ca \ absorption

139: lines that can be used to infer metallicities \citep[e.g.][]{az88} for the late type stars. This

140: will be the subject of a subsequent paper (Paper II) and is discussed only briefly here. In Paper

141: III we will use our measurements for new, more detailed dynamical models of the MW bar.

142: \begin{table*}

143:  \begin{center}

144:  \caption{Sub-fields observed in the MW bar. Right ascension ($\alpha$) and declination ($\delta$) are

145:  from epoch J2000 and $N_{stars}$ indicates the total number of stars (including foreground) obtained from each sub-field.}

146:  \begin{tabular}{llccccc} \tableline \tableline

147:  Field & Date Obs. & $\alpha$ & $\delta$ & $l(\degr)$ & $b(\degr) $

148:   & $N_{stars}$ \\

149:  \tableline

150: MM5B-a & 1996 & 17:47:24.05 & -35:00:21.07 & -5.00 & -3.47 & 472\\

151: MM5B-b & 1997 & 17:47:40.03 & -35:00:34.0 & -4.97 & -3.52 & 378\\

152: MM5B-c & 1999  & 17:47:20.71 & -35:56:52.84 & -4.95 & -3.43 & 608\\

153: MM5B-d & 1997/99 & 17:47:39.80 & -35:56:47.85 & -4.92 & -3.49 & 525\\

154: BWC-a & 1999 &  18:03:39.70 & -29:59:51.2 & 1.06 & -3.93 & 718\\

155: BW1-a & 1999  & 18:02:29.41 & -29:49:27.4 & 1.09 & -3.62 & 359\\

156: MM7B-a & 1996  & 18:11:29.61 & -25:54:06.6 & 5.50 & -3.46 & 353\\

157: MM7B-b & 1997  & 18:11:44.23 & -25:54:26.0 & 5.52 & -3.51 & 322\\

158: MM7B-c & 1997/99 & 18:11:30.32 & -25:50:38.5 & 5.55 & -3.44 & 288\\

159: MM7B-d & 1999 & 18:11:44.72 & -25:50:33.4 & 5.58 & -3.48 & 396\\

160: \tableline

161:  \end{tabular}

162:  \label{obs}

163:  \end{center}

164: \end{table*}

165: 

166: \section{Observations}

167: Observations were made between 1996 and 1999 using the Rutgers Fabry-P\'{e}rot (RFP) system at the

168: f/13.5 Cassegrain focus of the Cerro Tololo Inter-American Observatory (CTIO) 1.5m telescope. The

169: field of view of the RFP was circular with a diameter of $4^{\prime}$. The detector was a Tek 1024

170: x 1024 pixel CCD, giving an image scale of $0.36\arcsec$ per pixel. Typical seeing was in the

171: range of $1\arcsec - 1.5 \arcsec$. Each narrow-band exposure was of 900~s duration. A typical

172: image is shown in Figure \ref{stellarfield}. We observed three LOS: the MM5B field at l = -5\degr,

173: the MM7B field at l = +5\degr, and BW. There are four overlapping subfields in each of the MM5B

174: and MM7B fields, as shown in Figure \ref{cobe}, and two sub-fields in BW. Table \ref{obs} lists

175: the location of each sub-field and the total number of stars obtained per field. The choice and

176: nomenclature of the fields was motivated by the OGLE survey \citep{udalski}, which provides us

177: with excellent I and V band stellar photometry. In total we obtained $\sim 4400$ stellar spectra.

178: 

179: We chose to observe the \ca \ spectral line. The calcium triplet (8498.02~\AA, 8542.09~\AA\, and

180: 8662.14~\AA) is one of the strongest near-infrared spectral features in late-type stars like the

181: RCGs. The triplet arises from the transition between the two levels of $3^2\textrm{D}$ and the two

182: levels of $4^2\textrm{P}$ in \ion{Ca}{2}. The temperature of the bulk of the ISM is not high

183: enough to excite \ion{Ca}{2} to the $3^2\textrm{D}$ state, and therefore no contamination from the

184: foreground ISM is expected. The 8542 \AA\ line is the strongest of the triplet and falls in a

185: wavelength region where the night sky is relatively free of strong terrestrial emission lines.

186: 

187: We used a medium-resolution etalon with a spectral response function that is well fit by a Voigt

188: function with a FWHM of 4~\AA, equivalent to approximately 140 \kms\ at 8500~\AA. We typically

189: scanned a range of $8530 - 8555$ \AA\ with wavelength steps of 1~\AA. The FP images are not

190: monochromatic, because the rays from off-axis points pass through the collimated beam at an angle,

191: making the passband shift to the blue towards the edge of the field. The scanning range was

192: sufficient to compensate for the 4\AA\ center-to-edge gradient by including extra images at the

193: red end of the scan. The resulting spectra allowed measurement of the central wavelength of the

194: line with high precision (5 -- 10 \kms).

195: 

196: This project was plagued by bad weather and technical difficulties with the instrument and

197: telescope, requiring a significantly longer period of time to acquire the data than expected.

198: During two of the nights in the 1997 run the instrument shutter malfunctioned, affecting the data

199: for sub-fields MM5B-b and MM7B-b, which upon reduction showed large deviations in kinematics

200: compared to the neighboring fields that overlap with them. For this reason we excluded these two

201: fields from our analysis. Because of time constraints only 18 images were obtained for the BW1-a

202: field with heavy obscuration from clouds for half of them, and hence we also excluded this field.

203: These exclusions brought the total sample size down to 3360 stars from seven sub-fields, of which

204: 3144 are used in our kinematic analysis.

205: 

206: We calibrated the wavelength response of the system with several lines around 8542\,\AA\ from a

207: neon spectral lamp. The complete calibration sequence was done once per observing run, and was

208: stable over the duration of the run (and indeed from year to year). The precision of the

209: wavelength calibration was typically better than $0.05$~\AA, corresponding to 1.8 \kms at

210: 8500~\AA. The zero-point of the wavelength scale was affected by changes in the environmental

211: conditions such as temperature and atmospheric pressure, and was monitored by taking a single neon

212: calibration image once each hour during the observations. The wavelength zero-point for each

213: observation image was determined by interpolation from these hourly calibrations. On images that

214: have night sky lines within their spectral range, we can also use these sky lines to check the

215: zero point. There were two weak OH lines (at 8548.71~\AA\ \& 8538.67~\AA; \citet{osterbrock}) that

216: fall within our scanning range. Although these lines had very low S/N, the zero-point we measured

217: from them for a few frames was consistent with the interpolated calibration ring measurement.

218: \section{Data Reduction}

219: \begin{figure*}

220: % Requires \usepackage{graphicx}

221:  \includegraphics[scale=0.9]{f3.ps}\\

222:  \caption{A sample of the \ca\ spectra for

223:  almost the entire brightness range of our sample from I-band magnitude of 11.0 to 16.5.

224:  Each data point on the absorption line comes from a single image in the data cube.

225:  The solid line is a Voigt fit to the spectrum.}

226:  \label{spec}

227: \end{figure*}

228: \subsection{Fabry-P\'{e}rot Photometry}

229: The images in each data cube are overscan corrected, trimmed and bias subtracted using

230: IRAF\footnote{IRAF is distributed by NOAO, which is operated by AURA, Inc., under a cooperative

231: agreement with the NSF}. To perform photometry on the individual images we use the DAOPHOT

232: stand-alone package by \citet{stetson87} that was specially developed to perform crowded field

233: stellar photometry. The task FIND gives the initial centroids of all the stars in the field. After

234: performing the initial aperture photometry using the PHOT task, PSF-based photometry is performed

235: using a quadratically variable PSF, which we found to be especially useful at the edges of the

236: field. We used about 50 -- 100 stars per image to determine the PSF. The ALLSTAR task then

237: simultaneously fits all of the stars in the image using the measured PSF to give a star subtracted

238: image, and fluxes and positions for all stars. Because DAOPHOT performs local sky subtraction, the

239: fluxes are not affected by the presence of the weak OH sky lines. DAOPHOT includes the ALLFRAME

240: \citep{stetson94} program that is particularly useful for reducing a data cube. It simultaneously

241: performs the PSF fitting for all stars in all the images of the cube to derive positions and

242: magnitudes in a self-consistent manner, making use of the geometric transformations between the

243: individual images and photometry information for all images. Because our fields are much less

244: crowded than, for example, a globular cluster, DAOPHOT is able to resolve the stars easily. The

245: final output of DAOPHOT are the centroids, fluxes (in arbitrary units) and flux uncertainty for

246: each star.

247: \subsection{Flux Normalization}

248: In FP absorption line work, it is critically important to account for changes in flux due to

249: variations in atmospheric transmission. This can be significant if the images of a spectral scan

250: were taken over several nights or even in different years, or in non-photometric weather

251: conditions. To measure the image-to-image normalization, we select about 30 high S/N stars

252: uniformly covering the FOV, so that there is a range in velocity and wavelength coverage. We

253: assume that all of these stars have a single dominant absorption line that is well represented by

254: a Voigt function. We fit a Voigt profile to the spectrum of each star, and at each point calculate

255: the ratio of the fit to the observed flux. For each image in the scan, we then calculate the mean

256: and standard deviation of these ratios, using all of the normalization stars for that image. This

257: ratio is the normalization factor for that image. If an image was taken in average transparency

258: conditions, then its normalization factor should be 1.0, within the photometric uncertainty; if

259: the image had lower transparency, the normalization will be greater than unity. We correct the

260: fluxes for each image in the scan using the individual normalization factors. In order to

261: determine the final normalization factors, we re-fit the profile to the normalized data, and

262: iterate the entire process until it converges.

263: 

264: This procedure introduces an additional uncertainty in the flux measurement, which we call the

265: `normalization uncertainty'. We combine this in quadrature with the DAOPHOT flux uncertainties.

266: For the highest S/N\footnote{We define the S/N for an individual star as the mean ratio of the

267: measured flux to the estimated flux uncertainty, averaged over all the points measured in the

268: star's spectrum.} ($\sim 80$) stars the flux normalization error dominates, while for the majority

269: of stars the DAOPHOT flux error is more important. Since the different normalization stars have a

270: range of velocities, and also because of the center-to-edge wavelength gradient in a FP image, an

271: individual image samples the stars at several different points on the line profile; this avoids

272: ``building in" a particular profile shape through the normalization factors. Under photometric

273: conditions, the normalization factors were typically $1.00 \pm 0.02$, with uncertainties of $0.002

274: - 0.003$. In bad conditions, the worst normalizations were $1.5 - 2.0$, with uncertainties of

275: $\sim 0.01$. We often re-observed at some of the wavelengths taken under poor conditions, and

276: found excellent agreement between stellar fluxes after normalization.

277: \subsection{Spectroscopy}

278: The \ca \ absorption line profiles with the combined uncertainties are fit with a Voigt function

279: using the Levenberg-Marquardt method of least-squares fitting. The fit gives the central

280: wavelength, continuum level, line strength, and the Gaussian and Lorentzian widths along with

281: their measurement uncertainties. We find that the Gaussian contribution to the shape of the

282: absorption line is negligible for the majority of stars and the line could be fit equally well by

283: fixing the Gaussian width at $10^{-3}$ \AA, using essentially a Lorentzian profile. This is

284: because the instrumental profile itself is essentially Lorentzian with a negligible Gaussian core,

285: and the \ca \ line in these stars is strongly saturated with damping wings from collisional

286: broadening. Figure \ref{spec} shows a sub-sample of absorption lines covering almost the entire

287: brightness range for our sample from I-band magnitude of 10.0 to 16.5. The Voigt fit is

288: represented by the solid line. For the brightest stars the velocity uncertainties can be as low as

289: 2 \kms\, and even for the very faint stars we can measure the absorption line adequately well with

290: velocity uncertainties of 20 -- 30 \kms.

291: \section{Radial Velocities}

292: \begin{figure*}

293: % Requires \usepackage{graphicx}

294: \centering

295: \includegraphics[angle=90, scale=0.7]{f4.ps}\\

296: \caption{The distribution of velocity errors for the MM5B-c field (left panel) and velocity

297: uncertainties as a function of signal-to-noise for the same field (right panel).} \label{verr}

298: \end{figure*}

299: Table \ref{data} lists the heliocentric radial velocities and their uncertainties that we measured

300: for our complete sample of 3360 stars with 1605 and 1037 stars at $l = \pm 5$\degr, respectively,

301: and 718 in BW. The median velocity error for our sample is $\sim 12$ \kms. The uncertainties are

302: higher for fainter stars and also in cases where a star has a close or unresolved neighbor. In

303: Figure \ref{verr} we show the distribution of velocity errors and the error as a function of S/N

304: for the MM5B-c subfield. The major source of error in the radial velocity measurements comes from

305: the uncertainties in the flux measurements, which include contributions from DAOPHOT and

306: normalization errors. In the case of very high S/N stars in the best seeing sub-fields like MM5B-c

307: and BWC-a (with lower normalization errors) the velocity errors can be as small as 2 -- 5 \kms.

308: 

309: To check if the velocity errors obtained from our Voigt fitter are reasonable, we carried out

310: Monte Carlo simulations of the data. For each star in our sample, we generated 1000 random

311: realizations of its spectrum by perturbing the fluxes of each point of the Voigt profile fit by

312: the corresponding $1-\sigma$ individual flux uncertainty. We then fit a Voigt function to each of

313: these generated spectra and find that the mean ratio of the error estimate from a single fit to

314: the standard deviation of the velocities from the Monte Carlo fits is about 0.97 for all the

315: fields, indicating that the single fit errors from our Voigt fitter code are reliable.

316: 

317: For the BW LOS, there are nine bulge M/K-giant stars in our work that are also present in the

318: radial velocity samples of \citet{sharples} and \citet{terndrup}. The velocity comparison is shown

319: in Figure \ref{common}. Our error bars on the horizontal axis are equal to or smaller than the

320: size of the data points. Our velocities agree very well with the previous measurements, with a

321: reduced chi-square of 1.02.

322: 

323: We also have two measurements (repeated over 1 -- 3 years) of the radial velocities for some stars

324: in the region where the fields at a given LOS overlap with each other. The comparison between the

325: velocities of stars in these overlapping regions is shown in Figure \ref{comm} where the vertical

326: axis shows the velocity difference divided by the $1\sigma$ error bars ($\Delta$). The dashed,

327: dotted and dot-dashed lines show the $1\sigma$, $2\sigma$ and $3\sigma$ limits respectively. The

328: stars at the very edge of the field in the overlapping regions may not show up in every image

329: taken in the scan of that field, so they are not included in this comparison. There is also a 10\%

330: chance for a foreground disk main sequence star to be a spectroscopic binary which may result in

331: velocity variations of $\lesssim 30$ \kms\,\citep{DM91}. We find three such examples of foreground

332: stars with large velocity differences in the overlapping regions, and these three stars are

333: omitted from Figure \ref{comm}. The comparison for the MM5B overlapping regions has a reduced

334: chi-square of $\lesssim 1$. There is one additional star (not foreground) in each of the two

335: overlapping regions of the MM7B fields that has a deviation of more than $3 \sigma$, inflating the

336: reduced chi-square for these fields. We don't have a clear explanation for the deviations in these

337: two cases except that they could be spectroscopic binaries in the bulge. The reduced chi-square

338: for all the stars in the MM7B fields is 1.6; excluding the two outliers reduces chi-square to

339: 1.06.

340: 

341: For most of the stars in the overlapping region we list the weighted mean velocity and errors in

342: Table \ref{data}. For the faintest stars from the overlapping regions, we use the velocity from

343: only the field with the best image quality, where the fit to the absorption line was significantly

344: better than the other field. The comparison with previous work, and in the overlapping regions,

345: indicates that our velocity measurements are repeatable and reliable. The velocity uncertainty of

346: less than 30 \kms\ is small compared to the velocity dispersion of the stars in the bar, and will

347: have no effect on our conclusions.

348: \begin{figure}[t]

349: \includegraphics[scale=0.60]{f5.ps}\\

350: \caption{Radial velocities of 9 stars in common between this work and previous measurements by

351: \citet{sharples} (squares) and \citet{terndrup} (circles). \label{common}}

352: \end{figure}

353: 

354: \section{Photometry and Astrometry}

355: We use photometric information to distinguish between the foreground disk stars, bulge red giants

356: and bulge RCGs. Since our observed fields are coincident with the OGLE survey, we have precision

357: (1 -- 2\%) V and I band photometry and

358: astrometry\footnote{http://ogledb.astrouw.edu.pl/$\sim$ogle/photdb/} \citep{database}.

359: Unfortunately, the fields MM7B-d and a quarter of MM7B-c fell outside the OGLE field. Photometric

360: observations for these fields were obtained with the 1-m telescope at the South African

361: Astronomical Observatory (SAAO). Reduction of these data using the IRAF photometry package gives V

362: and I band magnitudes with an accuracy of 5\% -- 10\%.

363: 

364: There are some stars in our sample that have no OGLE or SAAO photometry. For these stars we use

365: the FP measurement of the continuum level of the \ca\ absorption line to estimate I-band

366: magnitudes. To check the accuracy of these FP I-band magnitudes we use all the stars in the MM7Bc

367: field where we have both OGLE and SAAO photometry. We determine the zero point of the FP I-band

368: magnitudes from the OGLE photometry for these stars. We compare them to the OGLE and the SAAO

369: magnitudes to find that they are surprisingly consistent with differences $\lesssim 10\%$. These

370: FP I-band magnitudes are used to exclude foreground stars as discussed in section 6. Comparison

371: with the much deeper OGLE survey shows that our sample of RCGs in the bar is essentially complete.

372: Table \ref{data} lists the astrometry and photometry. The photometry source (OGLE, SAAO or FP) is

373: indicated in the last column. The astrometry for all fields comes from OGLE except for the field

374: MM7B-d where the source is the Digitized Sky Survey.

375: \section{Selections from the Color-Magnitude Diagram}

376: The color-magnitude diagram (CMD) is shown in Figure \ref{mm5bselectfun} for the MM5B field with

377: 1605 stars. We chose to show the CMD for this particular LOS since it has the largest number of

378: stars and also has OGLE photometry for all them. The CMDs for the other two LOS look almost

379: identical. \citet{pac94} discuss various parts and features of the BW CMD using the photometry of

380: about $3 \times 10^{5}$ stars from the OGLE project. We discuss in brief these features and their

381: selection limits in the $(I_0, (V - I)_0)$ plane to be used in our analysis.

382: \begin{figure*}

383:   %\includegraphics[scale=0.65,angle=90,viewport=28 100 400 655]{f6.ps}\\

384:  \includegraphics[scale=0.65,angle=90]{f6.ps}\\

385:   \caption{Repeated radial velocity measurements of stars in the overlapping regions of

386:   the sub-fields of MM5B and MM7B.}\label{comm}

387: %\plotfiddle{common.eps}{-1.0in}{90.}{400.}{400.}{50}{200}

388: \end{figure*}

389: 

390: The first feature is a distinct red clump, marked by the red points. These RCGs are very useful

391: candidates for studying the kinematics of the bar. They are basically metal-rich counterparts of

392: the horizontal branch stars burning helium in their cores and hydrogen in their envelopes. Their

393: intrinsic luminosity distribution is very narrow and has a weak dependence on metallicity and age,

394: making them excellent distance indicators \citep{pac98}. They are bright and numerous in the inner

395: galaxy and are located in the region of CMD that is least contaminated by disk stars

396: \citep{stanek94}.

397: 

398: Our selection limits for RCGs towards all three LOS are $(I_0, (V-I)_0) = (13.0 - 15.5, 0.5 -

399: 1.4)$, corrected for the extinction ($A_I$) and reddening ($E_{V-I}$) values from \citet{sumi}.

400: This I-band limit corresponds to a distance range of about 5.5 -- 13 kpc for intrinsic red clump

401: magnitude $M_I = -0.28 \pm 0.20$ \citep{pac98}. \citet{stanek97} used a slightly different

402: selection criteria for OGLE RCGs where they define extinction independent I-band magnitudes. The

403: effect of these two different selection methods on the overall kinematics of the RCGs is

404: negligible, with changes in the mean radial velocity and dispersion of about 0.75 \kms\, and 1.15

405: \kms\, respectively. The latter selection also gives 2\% fewer stars at $l = \pm 5$\degr and 10\%

406: fewer stars in BW. We have also investigated the effect of varying the brighter end of the I-band

407: selection limit from 13.0 -- 13.6 magnitudes, which produced negligible changes to the mean

408: velocity and dispersion ($\sim 0.5$ \kms). The effects of different selection criteria for RCGs

409: will be more significant when modeling the data (subject of paper III), where it will be

410: investigated in greater detail.

411: 

412: About 210 stars at $(l,b) = (\pm 5.0,-3.5)$\degr and 39 in BW have only I-band photometry. Without

413: color information it is not possible to distinguish them from the foreground disk stars (blue

414: points in Figure \ref{mm5bselectfun}). If we apply only I-band cuts then the maximum difference in

415: the kinematics between including or excluding these stars without color information is about 1.3

416: \kms\, in mean velocity and 0.6 \kms\, in dispersion. The 2MASS point source catalogue has J, H

417: and K band photometry for our lines of sight. We found J-K colors for 60\% of these stars with no

418: V-I color information. The J-K colors confirmed that after applying the I-band or distance cut

419: about 96\% of these stars were RCGs. By analyzing the field MM7B-c, where 99\% of the sample has

420: complete V and I band photometry, we find that the contamination due to lack of color information

421: is $\sim 14\%$, after applying the I-band or distance cut. This amounts to contamination by only

422: 11 stars at $l = \pm 5$\degr and 2 stars in BW. These numbers are negligible compared to the

423: overall sample size and will have little effect on the kinematics. Thus we include these stars

424: (with just the I-band limits) in our final sample that will be used to measure bar kinematics,

425: which now has a total of 1193 RCGs at $(l,b) = (-5.0, -3.5)$\degr, 738 at $(l,b) = (5.5,

426: -3.5)$\degr and 557 in BW.

427: 

428: \begin{figure}[t]

429:  %\epsscale{2.5}\plottwo{cm.ps}{idist.ps}

430:  \includegraphics[scale=0.85, angle=90]{f7.eps}

431:  \caption{Extinction-corrected color-Magnitude diagram for the MM5B LOS shows a distinct red clump (red dots).

432:  The disk main sequence and the M/K -- giants are shown in blue and green dots respectively.

433:  We measure the radial velocity of every star in the diagram.}

434:  \label{mm5bselectfun}

435: \end{figure}

436: The region marked by the green points at roughly $(I_0, (V-I)_0) = (11.8 - 13.25, 0.5 - 3.0)$

437: contains the bulge M/K-giants \citep{sharples,terndrup}. This is the same M-giant population used

438: in the recent radial velocity survey by \citet{rich07}. In section 7.2 we use them to make a

439: direct comparison with that study. This sample, measured simultaneously from the same data cube,

440: allows us to compare the kinematics of two different stellar populations in the bar/bulge region.

441: 

442: The feature marked by the blue points at $(I_0, (V-I)_0) = (10.0 - 15.7, -0.3 - 0.5)$ is formed by

443: the disk main sequence (DMS) stars. The stars in this feature are thought to be associated with

444: the Sagittarius spiral arm located at $\sim 2$ kpc from the Sun \citep{pac94, ng96}. We have

445: radial velocities of about 350 stars in this region over three LOS. This feature and its possible

446: association with the Sagittarius arm is discussed in section 7.3.

447: \begin{figure*}

448:  % Requires \usepackage{graphicx}

449:  \includegraphics[scale=0.7,angle=90]{f8.ps}\\

450:  \caption{Stellar radial velocity distribution for three LOS in the bar.

451:  Right and Left panel: the dotted line is the Gauss-Hermite fit with $h_3$ and $h_4$ moments shown

452:  by dashed and dot-dashed lines respectively. Middle panel: the dotted line is double

453:  Gaussian with a narrow component for the globular cluster NGC 6522 and the solid line is the broad

454:  component (only) for the Galactic bar/bulge.}

455:  \label{vhist}

456: \end{figure*}

457: 

458: Finally, the feature at $(I_0,(V-I)_0) = (10.0 - 13.0, 0.5 - 0.9)$\footnote{Note that there is an

459: overlap in the selection limits between other features and the M-giant sample. We chose the

460: selection of the M-giants to be as close as possible to that used in previous investigations, in

461: order to make kinematic comparisons.} on the CMD marked by the magenta points consists of the disk

462: red clump giants spread over a distance of about 1.0 -- 4.0 kpc. We had just enough stars in our

463: sample along this LOS to detect this feature and measure its kinematics. The FP imaging

464: spectroscopy has allowed us to measure simultaneously the kinematics of all these features,

465: providing a powerful and reliable way to look for differences between the various stellar

466: populations in a consistent manner.

467: \section{Kinematics}

468: \subsection{Kinematics of Red Clump Giants in the Bar: Measurement of $h_3$ and $h_4$}

469: The large size of our sample allows us to measure, for the first time, the higher moments of the

470: velocity distribution along our three LOS. We fit a Gauss-Hermite series to the velocity

471: distribution \citep{gerhard93,gh93} of RCGs to extract the four velocity moments -- mean line of

472: sight velocity ($<V_{\textrm{los}}>$), line-of-sight velocity dispersion

473: ($\sigma_{\textrm{los}}$), asymmetry of the distribution ($h_3$), and flatness of the distribution

474: ($h_4$)  -- using a maximum likelihood estimator (MLE). An advantage of using the MLE is that it

475: corrects for the individual velocity errors that can broaden the width of the distribution. We use

476: the MLE code developed by Glenn Van de Ven (private communication; see \citet{glenn06} for more

477: details) to fit a Gauss-Hermite series to our velocity distributions and to determine the velocity

478: moments. The code gives a robust measurement of the uncertainties of each moment in two ways. The

479: bootstrap method randomly draws from the observed velocity distribution with replacement

480: \citep[see Sect. 15.6 of][]{press92}; this gives the maximum possible errors of the parameters.

481: The Gaussian randomization (GR) method randomly draws velocities from an intrinsic Gaussian

482: distribution with a given mean and dispersion that is broadened by velocity errors randomly drawn

483: from the observed velocity error distribution \citep[see Appendix B of ][]{glenn06}. GR gives

484: reliable estimates if the individual errors are not underestimated, which we believe is the case

485: with our analysis. Finally, the code uses the biweight statistic to measure just the first two

486: velocity moments ($<V_{\textrm{los}}>$ and $\sigma_{\textrm{los}}$) and their uncertainties; these

487: are the appropriate measures to compare with previous investigations that had too few stars to

488: reliably determine the higher moments.

489: 

490: MM5B-a and MM5B-c have identical kinematics and can be combined into a single field MM5B-(a+c).

491: MM5B-d is at the same LOS and overlaps with the MM5B-c sub-field but has a mean velocity that is

492: offset by 13 \kms\ (i.e. about a $2\sigma$ deviation). The dispersions in all three sub-fields

493: however agree within the $1\sigma$ errors. For this reason we measure the velocity moments at $l =

494: -5.0$\degr\ using both MM5B-(a+c) as well as the three sub-fields combined MM5B-(a+c+d). There are

495: no significant differences in the kinematics of sub-fields MM7B-a, MM7B-c and MM7B-d, so they are

496: combined into a single field MM7B-(a+c+d). The velocity distributions for MM5B-(a+c) and

497: MM7B-(a+c+d) are shown in the left and right panel of Figure \ref{vhist}. The dotted line is the

498: Gauss-Hermite fit with the $h_3$ and $h_4$ moments represented by the dashed and dot-dashed lines

499: respectively. Table \ref{kine} lists the velocity moments along with their uncertainties. Our

500: large sample size allows us to measure $<V_{\textrm{los}}>$ and $\sigma_{\textrm{los}}$ with high

501: accuracy. As expected, in Galactocentric coordinates $<V_{\textrm{los}}>$ has equal amplitude of

502: about $\pm 63$ \kms\ at $l = \pm 5$\degr, respectively. The $h_4$ moment is significantly non-zero

503: while the $h_3$ is consistent with zero. Our high S/N velocity distributions show no sign of cold

504: kinematic features that could be related to disrupted satellites, as suggested by \citet{rich07}.

505: \begin{figure*}

506:  % Requires \usepackage{graphicx}

507:  \includegraphics[scale=0.8,angle=90]{f9.ps}\\

508:  \caption{Comparison of the kinematics (in galactocentric frame) of RCGs (red triangles) and M-giants (green diamonds) from our sample to

509:  \citet{rich07} data (open circles) and \citet{zhao96} model (blue dashed line)}

510:  \label{comp}

511: \end{figure*}

512: 

513: The analysis of the BW velocity distribution (middle panel) is slightly complicated by the

514: presence of stars from the bulge globular cluster NGC 6522, the outskirts of which extend into the

515: southwest edge of our field (see Figure \ref{stellarfield}). The spike at the center of the BW

516: velocity distribution (shown in Figure \ref{vhist} middle panel) is caused by the stars that are

517: part of this cluster. To extract kinematics we fit a double Gaussian (broad for the bulge and

518: narrow for NGC 6522) to this distribution. The narrow Gaussian has a mean velocity of $-14.67 \pm

519: 3.77$ \kms \ and dispersion of $7.9 \pm 2.82$ \kms \ ; consistent with previous measurements for

520: NGC 6522 of $-10.3 \pm 1.6$ \kms \ and $7.3^{+3.5}_{-2.0}$ \kms \, respectively \citep{dubath97}.

521: The dotted line in Figure \ref{vhist} shows the double gaussian fit to the distribution and the

522: solid line is the broad component representing the bulge velocity distribution. We also fit a

523: Gauss-Hermite series to this distribution and list the values in Table \ref{kine}, although this

524: measurement is most probably affected by the presence of the spike at the center. Since this is

525: the most heavily studied LOS, there are several measurements of the kinematics in BW for

526: comparison. Our values of the mean line-of-sight velocity and dispersion are in excellent

527: agreement with all previous measurements of the kinematics in BW such as \citet{mould83}

528: ($<V_{\textrm{los}}> = -10 \pm 19$ \kms, $\sigma_{\textrm{los}} = 113 \pm 11$ \kms),

529: \citet{sharples} ($<V_{\textrm{los}}> = 4 \pm 8$ \kms, $\sigma_{\textrm{los}} = 113^{+6}_{-5}$

530: \kms), \citet{ rich90} ($<V_{\textrm{los}}> = -13 \pm 12$ \kms, $\sigma_{\textrm{los}} = 111 \pm

531: 8$ \kms), \citet{terndrup} ($<V_{\textrm{los}}> = -8 \pm 6$ \kms, $\sigma_{\textrm{los}} = 110 \pm

532: 10$ \kms) and \citet{rich07} ($<V_{\textrm{los}}> = -1.1\pm 12.9$ \kms, $\sigma_{\textrm{los}} =

533: 110.2 \pm 9.1$ \kms).

534: 

535: \subsection{Kinematics of M-giants}

536: Our measurements of the kinematics of M-giants (in the heliocentric coordinate system) are listed

537: in Table \ref{kine2}. We do not tabulate the measurements for BW since the globular cluster

538: M-giants dominate the already small sample, and are difficult to separate from the bulge

539: population, making the measurement unreliable. In Figure \ref{comp} we compare the kinematics of

540: the bulge RCGs and M-giants from our sample with that of \citet{rich07} M-giants and

541: \citet{zhao96} (Zhao hereafter) model. We find that the radial velocities of the RCGs and M-giants

542: stars agree well with each other and with the \citet{rich07} measurements. However the LOS

543: dispersion for our M-giant sample (green diamonds) is systematically lower than the RCGs (red

544: triangles), and is consistent with the velocity dispersion of \citet{rich07} M-giants (black

545: circles). Note that the higher dispersion of the RCGs in our sample is not because they are

546: fainter and thus have somewhat larger velocity uncertainties. The MLE method we use removes the

547: effects of the uncertainties from the estimate of the dispersion. In any event, to produce the

548: difference between the M-giant and the RCG dispersion would require addition in quadrature of 40

549: -- 50 \kms\, compared to the uncertainties of 10 -- 25 \kms.

550: 

551: The blue dashed line in Figure \ref{comp} is the prediction from Zhao's model. The bar in this

552: model rotates faster than the observed rotation curve at greater longitudes. According to Zhao and

553: \citet{rich07} this rapid rotation is due to a lack of retrograde orbits in the model. But adding

554: retrograde orbits will increase the $\sigma_{\textrm{los}}$ of the model, which may then agree

555: with the dispersion of RCGs rather than that of the M-giants.

556: 

557: \begin{table}[t]

558: \caption{Kinematics of bar M-giants and disk main sequence stars}

559: \begin{tabular}{cccc}

560:  \tableline\tableline

561:  LOS & $<V_{\mathrm{los}}>$ & $\sigma_{\textrm{los}}$ & N\\

562:   & ($\mathrm{km\, s^{-1}}$) & ($\mathrm{km\, s^{-1}})$ & \\

563:  \tableline

564:  \multicolumn{4}{c}{M-giants}\\

565:  \tableline

566:  %\cutinhead{Kinematics of M-giants}

567:  (-5.0, -3.5) & $-37.2 \pm 6.9$ & $89.9 \pm 4.9$ & 171\\

568:  (5.5, -3.5) & $29.9 \pm 9.6$ & $95.2 \pm 6.8$ & 98\\

569:  \tableline

570:  %\cutinhead{Kinematics of DMS}

571:  \multicolumn{4}{c}{DMS}\\

572:  \tableline

573:  (-5.0, -3.5) & $-17.8\pm 4.0$ & $45.9 \pm 2.8$ & 134\\

574:  (1.06, -3.9) & $ -3.1\pm 4.8$ & $41.3 \pm 3.4$ & 75\\

575:  (5.5, -3.5) & $0.9 \pm 4.3$ & $45.7\pm 3.1$ & 109\\

576:  \tableline

577: \end{tabular}\label{kine2}

578: \end{table}

579: 

580: Is there really a difference in dispersion between the M-giants and RCGs? One possibility is that

581: the M-giant sample may have more foreground contamination, which could cause the observed

582: decrement in the dispersion. However, our M-giant selection criteria are virtually the same as

583: those of \citet{rich07}, who argue that contamination is negligible. If the difference is real,

584: then perhaps there is an age difference between the RCGs and the M-giants, with RCG's being older.

585: Another possibility is that the M-giants represent multiple dynamical populations. But at this

586: point we have no means to distinguish between these or other hypotheses. We plan to obtain more

587: lines-of-sight in the bar using the FP system \citep{rangwala08} on the 10-m Southern African

588: Large Telescope (SALT) \citep{buckley03} to explore this discrepancy further.

589: \subsection{Kinematics of Disk Stars}

590: Our measurements of the kinematics of the DMS stars (blue dots in figure \ref{mm5bselectfun}) are

591: listed in Table \ref{kine2}. Earlier studies of the Baade's window CMD \citep{terndrup, pac94,

592: ng96} found an excess of stars within 2.5 kpc of the Sun. They associated this increased stellar

593: density (by a factor $\sim 2$) of DMS stars with the location of the Sagittarius spiral arm at a

594: distance of $\sim 2$ kpc. The simulations performed by \citet{ng96} showed this density enhanced

595: feature to consist of some very young stars (0.1 -- 2.0 Gyr) that are identified with the recent

596: spiral arm population superimposed on a larger population of young and intermediate age (2.0 --

597: 7.0 Gyr) stars identified with some past spiral arm population. We find the

598: $\sigma_{\textrm{los}}$ to be the same ($\sim$ 45 \kms) towards all three LOS. According to the

599: relationship between velocity dispersion and distance from the Galactic center \citep{freeman89},

600: our dispersion corresponds to a position within 3.5 kpc of the Sun. This suggests that the

601: majority of the stars in this feature may belong to the Sagittarius spiral arm population.

602: \begin{figure}[t]

603: \includegraphics[scale=0.7]{f10.eps}

604: \caption{I-band apparent magnitude distribution of the red clump giants in MM5B field. The two

605: bands show the sample limits for stars on the near and far sides of the bar.\label{idist}}

606: \end{figure}

607: 

608: The dispersion of the disk red clump feature (Magenta points on the CMD) is about 53 \kms,

609: consistent with the disk kinematics.

610: \section{Radial Streaming Motion}

611: Previous observations have established that the near end of the Galactic bar is in the first

612: Galactic quadrant, i.e., at positive longitudes. Since the stars forming the bar stream in the

613: same sense as the Galactic rotation, for this bar orientation the stars on the near and far side

614: of the bar will have radial velocity components towards and away from us, respectively. The radial

615: velocity difference between these approaching and receding stellar streams depends on the

616: orientation of the bar, among other things. Since RCGs are good distance indicators with a small

617: dispersion in intrinsic luminosity, they can be localized in space and used to measure precisely

618: the velocity shift between the two streams (\citet{mao02}; MP02 hereafter). This measurement will

619: provide an additional strong constraint on the dynamical models of the Galactic bar, especially

620: its orientation angle.

621: \begin{figure*}

622:   % Requires \usepackage{graphicx}

623:   \includegraphics[scale=0.7, angle=90]{f11.ps}\\

624:   \caption{The radial velocity distribution for the stars in the near

625:   and far side samples, shown by the solid and dotted histograms respectively.}\label{stream}

626: \end{figure*}

627: 

628: MP02 discuss the optimal selection of the bright and faint samples to isolate these stellar

629: streams. We adapt their criteria for the characteristics of our data set, selecting RCGs 0.2 to

630: 0.6 magnitudes brighter and fainter than the peak in the observed luminosity distribution for each

631: field to define the bright (front side) and faint (back side), respectively. Figure \ref{idist}

632: illustrates the two samples for the l = -5.0\degr\ field, and the details of these bright and

633: faint samples for the three LOS are listed in Table \ref{streamdata}. Figure \ref{stream} shows

634: the velocity distribution of the near and far side samples. We detect a clear velocity shift of

635: $\sim 35$ \kms\ in each of $l \simeq \pm 5$\degr\ LOS. The velocity shift in BW is much smaller

636: and is consistent with zero. This may be because in BW we are seeing an older spheroidal bulge

637: population superimposed on the bar population, masking this signal. The higher velocity dispersion

638: in BW also indicates that there are higher random motions compared to the streaming motions. Our

639: observed shifts of 35 \kms\ are much greater than any possible projection effects arising from the

640: Galactic rotation and the location of the near and far samples (1 -- 2 \kms). This suggests that

641: there are strong non-circular streaming motions associated with the bar. MP02 predicted the

642: difference of the LOS component of the streaming motions between the bright and faint sample to be

643: $\sim 33$ \kms\, based on the E2 model of \citet{dwek95}. This predicted shift is consistent with

644: our measurements in the $l = \pm 5$\degr\ fields but not in BW. MP02 also discussed the effects of

645: red giant contamination on these shifts and concluded that their presence in the RCG sample could

646: significantly dilute the magnitude of the velocity shifts, so that the actual shifts could be

647: larger than those that we measure.

648: 

649: \citet{deguchi01} \citep[see also][]{deguchi} used SiO masers in the bar to measure the velocity

650: shifts. They find an average radial velocity difference of $-21 \pm 27$ \kms\ for sources with

651: distances less than 7 kpc compared to those at distances greater than 7 kpc, essentially

652: independent of longitude. While their results agree with our measurements, they suffer from a

653: large (50\%) uncertainty in distance and large statistical errors in kinematics (due to a smaller

654: sample size).

655: 

656: As pointed out by \citet{deguchi01} most of the dynamical models of the inner galaxy do not

657: predict or even discuss the change in average radial velocity with distance. The dynamical model

658: of \citet{hafner} was one of the first to discuss this in some detail. In their Figures 9 and 10

659: they plot line-of-sight and proper motion kinematics as a function of distance for four windows,

660: one of which is BW. We have a large enough sample of RCGs to plot LOS average velocity and

661: dispersion as a function of distance in the inner galaxy, which we show in Figure \ref{vimag}.

662: Although we cannot make an exact comparison with Hafner's predictions for BW since they use a

663: generic selection function for K giants, our data are nonetheless in excellent agreement with

664: their model (open diamonds in Figure \ref{vimag}).

665: 

666: \begin{figure*}

667:   % Requires \usepackage{graphicx}

668:   %\includegraphics[scale=0.7, angle=90,viewport=20 100 340 685]{f12.ps}\\

669:   \includegraphics[scale=0.7,angle=90]{f12.ps}\\

670:   \caption{Line-of-sight radial velocity and dispersion as a function of I-band

671: magnitude (or distance) of the bulge red clump giants for three LOS in the Galactic bar. The

672: vertical dashed line shows the position of the peak magnitude about which the bright and faint

673: samples were measured. Note that these are apparent magnitude not corrected for extinction. The

674: open diamonds show \citet{hafner}'s dynamical model.}\label{vimag}

675: \end{figure*}

676: 

677: \section{Equivalent Widths}

678: Table \ref{data} lists the equivalent widths and their uncertainties for our complete sample of

679: 3360 stars. We measured the equivalent widths of \ca \ lines in our sample by integrating the

680: Voigt fit over all wavelengths. Our measurements of mean equivalent width and dispersion for the

681: bulge RCGs are listed in Table \ref{ew}. The mean equivalent width ($<\mathrm{EW}>$) and its

682: dispersion decrease on either side of BW similar to the LOS velocity dispersion. Because the \ca \

683: line is a good proxy for metallicity \citep{az88}, the gradient in the mean equivalent width may

684: imply a gradient in metallicity \citep{frogel88, minniti95} along the Galactic bar. We also find

685: that the mean EW of the DMS stars is about 25\% lower than the bulge RCGs, which is consistent

686: with a bulge that is more metal rich than the solar neighborhood. This has been the consensus of

687: many previous studies in BW, for example \citet{rich88}. The factors affecting the measurement of

688: FP equivalent widths and inferring [Fe/H] using the calcium infrared triplet method is addressed

689: in detail in Paper II. For recent work on the metallicity of the Galactic bulge see

690: \citet{fulbright06, minniti08}.

691: \section{Discussion and Conclusions}

692: \begin{table}[t]

693: \caption{Mean equivalent width and dispersion for the \ca\ line in the bar RCGs}\label{ew}

694: \begin{tabular}{cccc}

695:  \tableline\tableline

696:  LOS & $<\mathrm{EW}>$ & $\sigma_{\mathrm{EW}}$ & N\\

697:   & \AA & \AA & \\

698:  \tableline

699: (-5.0,-3.5) & $3.68 \pm 0.04$ & $0.46\pm 0.04$ & 804\\

700: (1.1,-3.9) & $3.86 \pm 0.05$ & $0.68\pm 0.04$ & 557\\

701: (5.5,-3.5) & $3.28 \pm 0.04$ & $0.48\pm 0.04$ & 738\\

702: \tableline

703: \end{tabular}

704: \end{table}

705: 

706: In this paper we have presented measurements of the \ca\ absorption line using Fabry-P\'{e}rot

707: imaging spectroscopy. This work shows the strength and reliability of this technique in measuring

708: radial velocities for a large sample of stars with I-band magnitudes ranging from 10.0 -- 16.5.

709: Past investigations have mostly used red giants to study the bar's stellar kinematics. Our work is

710: the first to measure the kinematics of red clump giants, which are more numerous in the inner

711: galaxy than the M and K giants. The main focus of this paper is to present the data and

712: preliminary interpretations, but full interpretation will only be possible with extensive modeling

713: that we will present in Paper III.

714: 

715: We obtained radial velocities for three LOS in the MW bar that include two offset positions at $l

716: \simeq \pm 5$\degr and the central field of BW. The large sample size enabled high precision

717: measurements of the first four moments of the velocity distributions for the first time. The

718: symmetric moment $h_4$ is significantly non-zero, and its negative value indicates that the

719: distributions are flat-topped rather than peaked. This seems consistent with a model of the bar

720: with stars in elongated orbits forming two streams at different mean radial velocities, broadening

721: and flattening the total distribution. Because our sample of bar RCGs is essentially complete, the

722: continuity equation suggests that the number of stars in the two streams should be the same. Table

723: \ref{streamdata} shows that this is the case within the Poisson noise. We then expect no asymmetry

724: in the velocity distribution function, and our measures of $h_3$ are indeed consistent with zero.

725: 

726: When we divide our sample into bright and faint samples we detect a clear kinematic signature of

727: streaming motions about the bar. The stellar streams have velocity difference of about 35~\kms\ at

728: $l = \pm 5$\degr. There is no indication of these streams in BW, perhaps because they are masked

729: by a spheroidal bulge component. The larger velocity dispersion in BW is consistent with this

730: interpretation. Our large sample size allows us to take the first step towards measuring the LOS

731: kinematics as a function of apparent magnitude (or distance) using the RCGs. As \citet{hafner}

732: points out, these measurements can provide powerful constraints on dynamical models of the inner

733: Galaxy. With larger data sets it will be possible to detect the transition from disk to bar, and

734: measure the bar's thickness.

735: 

736: The technique of FP imaging spectroscopy has allowed us to measure the radial velocities samples

737: that are an order of magnitude larger (at individual LOS) than previous studies. This large sample

738: allows us to determine the kinematics of various stellar populations: bar RCGs, bar M-giants and

739: the disk main sequence stars. Our measurements show that the RCG and M-giant populations have the

740: same $<V_{\textrm{los}}>$ but their dispersions differ by about 10 \kms. The kinematics of the DMS

741: population suggests that most of these stars belong to the Sagittarius spiral arm.

742: 

743: In addition to the kinematic measurement, we have used the FP spectra to measure the \ca\ line

744: strengths. We can use these results to measure the metallicities of the individual stars.

745: Metallicity distributions for several LOS in the bar/bulge can provide strong constraints on the

746: chemical evolution models that can determine the star formation rate, initial mass function, and

747: formation time scale \citep[e.g.][]{matteucci99,zoccali07} for the inner Galaxy. Our measurements

748: indicate a gradient in mean and dispersion of equivalent width as a function of Galactic

749: longitude, and a clear difference in the mean line strengths between the disk and bulge/bar. The

750: metallicity measurements will be the subject of Paper II.

751: 

752: Motivated by the success of this project, we plan to use the FP system on the 10-m class SALT to

753: obtain a much more extensive determination of the kinematics and metallicity of stars in the inner

754: Galaxy. SALT's much greater aperture and larger FOV will enable us to measure $\gtrsim 2000$ stars

755: along a single LOS in an hour of observing time. We will investigate at least 10 LOS along the

756: major axis of the bar, obtaining 15000 -- 20000 stellar spectra. This data set will provide the

757: basis for models of the structure and chemical evolution of the Galactic bar and bulge. We can

758: also investigate the long bar suggested by recent Spitzer observations \citep{benjamin08} and

759: search for hypervelocity stars. This planned increase by another order of magnitude in the size of

760: the data sample will revolutionize the studies of the inner Galaxy.

761: 

762: \acknowledgements Bohdan Paczynski was an early and enthusiastic collaborator on this project, and

763: provided insight and inspiration that contributed greatly to its success. Jerry Sellwood has

764: provided the theoretical foundation for the program and cheerfully tutored our understanding of

765: bar dynamics. Tad Pryor was our DAOPHOT expert and statistical guru.  Rutgers undergraduates

766: Robert Polocz and Brian Schnitzer assisted with data reduction. The director and staff of CTIO

767: provided their usual outstanding level of support for the observations, especially when amazingly

768: bad weather and recalcitrant equipment threatened to scuttle the project. We thank the director of

769: SAAO for allocating observing time and Dr. John Menzies for obtaining the additional photometry

770: observations. Victor Debattista has provided invaluable help in developing dynamical models using

771: these data. This research has made use of the SIMBAD database, operated at CDS, Strasbourg,

772: France. This project has been supported by the National Science Foundation through grant

773: AST0507323.

774: 

775: %\bibliographystyle{apj}

776: 

777: %\bibliography{refs}

778: 

779: 

780: \begin{thebibliography}{}

781: %\expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

782: 

783: \bibitem[{{Alcock} {et~al.}(2000){Alcock}, {Allsman}, {Alves}, {Axelrod},

784:   {Becker}, {Bennett}, {Cook}, {Drake}, {Freeman}, {Geha}, {Griest}, {Lehner},

785:   {Marshall}, {Minniti}, {Nelson}, {Peterson}, {Popowski}, {Pratt}, {Quinn},

786:   {Stubbs}, {Sutherland}, {Tomaney}, {Vandehei}, \& {Welch}}]{alcock00}

787: {Alcock}, C., {Allsman}, R.~A., {Alves}, D.~R., {Axelrod}, T.~S., {Becker},

788:   A.~C., {Bennett}, D.~P., {Cook}, K.~H., {Drake}, A.~J., {Freeman}, K.~C.,

789:   {Geha}, M., {Griest}, K., {Lehner}, M.~J., {Marshall}, S.~L., {Minniti}, D.,

790:   {Nelson}, C.~A., {Peterson}, B.~A., {Popowski}, P., {Pratt}, M.~R., {Quinn},

791:   P.~J., {Stubbs}, C.~W., {Sutherland}, W., {Tomaney}, A.~B., {Vandehei}, T.,

792:   \& {Welch}, D.~L. 2000, \apj, 541, 734

793: 

794: \bibitem[{{Armandroff} \& {Zinn}(1988)}]{az88}

795: {Armandroff}, T.~E., \& {Zinn}, R. 1988, \aj, 96, 92

796: 

797: \bibitem[{{Benjamin}(2008)}]{benjamin08}

798: {Benjamin}, R.~A. 2008, in Bulletin of the American Astronomical Society,

799:   Vol.~40, Bulletin of the American Astronomical Society, 266--+

800: 

801: \bibitem[{{Benjamin} {et~al.}(2005){Benjamin}, {Churchwell}, {Babler},

802:   {Indebetouw}, {Meade}, {Whitney}, {Watson}, {Wolfire}, {Wolff}, {Ignace},

803:   {Bania}, {Bracker}, {Clemens}, {Chomiuk}, {Cohen}, {Dickey}, {Jackson},

804:   {Kobulnicky}, {Mercer}, {Mathis}, {Stolovy}, \& {Uzpen}}]{benjamin05}

805: {Benjamin}, R.~A., {Churchwell}, E., {Babler}, B.~L., {Indebetouw}, R.,

806:   {Meade}, M.~R., {Whitney}, B.~A., {Watson}, C., {Wolfire}, M.~G., {Wolff},

807:   M.~J., {Ignace}, R., {Bania}, T.~M., {Bracker}, S., {Clemens}, D.~P.,

808:   {Chomiuk}, L., {Cohen}, M., {Dickey}, J.~M., {Jackson}, J.~M., {Kobulnicky},

809:   H.~A., {Mercer}, E.~P., {Mathis}, J.~S., {Stolovy}, S.~R., \& {Uzpen}, B.

810:   2005, \apjl, 630, L149

811: 

812: \bibitem[{{Binney} {et~al.}(1991){Binney}, {Gerhard}, {Stark}, {Bally}, \&

813:   {Uchida}}]{binney91}

814: {Binney}, J., {Gerhard}, O.~E., {Stark}, A.~A., {Bally}, J., \& {Uchida}, K.~I.

815:   1991, \mnras, 252, 210

816: 

817: \bibitem[{{Bissantz} {et~al.}(2004){Bissantz}, {Debattista}, \&

818:   {Gerhard}}]{bissantz04}

819: {Bissantz}, N., {Debattista}, V.~P., \& {Gerhard}, O. 2004, \apjl, 601, L155

820: 

821: \bibitem[{{Bissantz} \& {Gerhard}(2002)}]{bissantz02}

822: {Bissantz}, N., \& {Gerhard}, O. 2002, \mnras, 330, 591

823: 

824: \bibitem[{{Blitz} \& {Spergel}(1991)}]{blitz91}

825: {Blitz}, L., \& {Spergel}, D.~N. 1991, \apj, 379, 631

826: 

827: \bibitem[{{Buckley} {et~al.}(2003){Buckley}, {Hearnshaw}, {Nordsieck}, \&

828:   {O'Donoghue}}]{buckley03}

829: {Buckley}, D.~A.~H., {Hearnshaw}, J.~B., {Nordsieck}, K.~H., \& {O'Donoghue},

830:   D. 2003, in Presented at the Society of Photo-Optical Instrumentation

831:   Engineers (SPIE) Conference, Vol. 4834, Discoveries and Research Prospects

832:   from 6- to 10-Meter-Class Telescopes II. Edited by Guhathakurta, Puragra.

833:   Proceedings of the SPIE, Volume 4834, pp. 264-275 (2003)., ed.

834:   P.~{Guhathakurta}, 264--275

835: 

836: \bibitem[{{Cabrera-Lavers} {et~al.}(2008){Cabrera-Lavers},

837:   {Gonzalez-Fernandez}, {Garzon}, {Hammersley}, \& {Lopez-Corredoira}}]{cl08}

838: {Cabrera-Lavers}, A., {Gonzalez-Fernandez}, C., {Garzon}, F., {Hammersley},

839:   P.~L., \& {Lopez-Corredoira}, M. 2008, ArXiv e-prints, 809

840: 

841: \bibitem[{{Clarkson} {et~al.}(2008){Clarkson}, {Sahu}, {Anderson}, {Smith},

842:   {Brown}, {Rich}, {Casertano}, {Bond}, {Livio}, {Minniti}, {Panagia},

843:   {Renzini}, {Valenti}, \& {Zoccali}}]{clarkson08}

844: {Clarkson}, W., {Sahu}, K., {Anderson}, J., {Smith}, T.~E., {Brown}, T.~M.,

845:   {Rich}, R.~M., {Casertano}, S., {Bond}, H.~E., {Livio}, M., {Minniti}, D.,

846:   {Panagia}, N., {Renzini}, A., {Valenti}, J., \& {Zoccali}, M. 2008, \apj,

847:   684, 1110

848: 

849: \bibitem[{{de Vaucouleurs}(1964)}]{devac64}

850: {de Vaucouleurs}, G. 1964, in IAU Symposium, Vol.~20, The Galaxy and the

851:   Magellanic Clouds, ed. F.~J. {Kerr}, 195--+

852: 

853: \bibitem[{{Debattista} \& {Sellwood}(2000)}]{victor00}

854: {Debattista}, V.~P., \& {Sellwood}, J.~A. 2000, \apj, 543, 704

855: 

856: \bibitem[{{Debattista} \& {Williams}(2004)}]{victor04}

857: {Debattista}, V.~P., \& {Williams}, T.~B. 2004, \apj, 605, 714

858: 

859: \bibitem[{{Deguchi}(2001)}]{deguchi}

860: {Deguchi}, S. 2001, in Astronomical Society of the Pacific Conference Series,

861:   Vol. 228, Dynamics of Star Clusters and the Milky Way, ed. S.~{Deiters},

862:   B.~{Fuchs}, A.~{Just}, R.~{Spurzem}, \& R.~{Wielen}, 404--+

863: 

864: \bibitem[{{Deguchi} {et~al.}(2001){Deguchi}, {Fujii}, {Matsumoto}, {Nakashima},

865:   \& {Wood}}]{deguchi01}

866: {Deguchi}, S., {Fujii}, T., {Matsumoto}, S., {Nakashima}, J.-I., \& {Wood},

867:   P.~R. 2001, \pasj, 53, 293

868: 

869: \bibitem[{{Dubath} {et~al.}(1997){Dubath}, {Meylan}, \& {Mayor}}]{dubath97}

870: {Dubath}, P., {Meylan}, G., \& {Mayor}, M. 1997, \aap, 324, 505

871: 

872: \bibitem[{{Duquennoy} \& {Mayor}(1991)}]{DM91}

873: {Duquennoy}, A., \& {Mayor}, M. 1991, \aap, 248, 485

874: 

875: \bibitem[{{Dwek} {et~al.}(1995){Dwek}, {Arendt}, {Hauser}, {Kelsall}, {Lisse},

876:   {Moseley}, {Silverberg}, {Sodroski}, \& {Weiland}}]{dwek95}

877: {Dwek}, E., {Arendt}, R.~G., {Hauser}, M.~G., {Kelsall}, T., {Lisse}, C.~M.,

878:   {Moseley}, S.~H., {Silverberg}, R.~F., {Sodroski}, T.~J., \& {Weiland}, J.~L.

879:   1995, \apj, 445, 716

880: 

881: \bibitem[{{Freeman}(1996)}]{freeman96}

882: {Freeman}, K.~C. 1996, in Astronomical Society of the Pacific Conference

883:   Series, Vol.~91, IAU Colloq. 157: Barred Galaxies, ed. R.~{Buta}, D.~A.

884:   {Crocker}, \& B.~G. {Elmegreen}, 1--+

885: 

886: \bibitem[{{Freudenreich}(1998)}]{cobe98}

887: {Freudenreich}, H.~T. 1998, \apj, 492, 495

888: 

889: \bibitem[{{Frogel}(1988)}]{frogel88}

890: {Frogel}, J.~A. 1988, \araa, 26, 51

891: 

892: \bibitem[{{Fulbright} {et~al.}(2006){Fulbright}, {McWilliam}, \&

893:   {Rich}}]{fulbright06}

894: {Fulbright}, J.~P., {McWilliam}, A., \& {Rich}, R.~M. 2006, \apj, 636, 821

895: 

896: \bibitem[{{Gebhardt} {et~al.}(1994){Gebhardt}, {Pryor}, {Williams}, \&

897:   {Hesser}}]{gebhardt94}

898: {Gebhardt}, K., {Pryor}, C., {Williams}, T.~B., \& {Hesser}, J.~E. 1994, \aj,

899:   107, 2067

900: 

901: \bibitem[{{Gerhard}(1993)}]{gerhard93}

902: {Gerhard}, O.~E. 1993, \mnras, 265, 213

903: 

904: \bibitem[{{Gerhard}(1999)}]{gerhard99}

905: {Gerhard}, O.~E. 1999, in Astronomical Society of the Pacific Conference

906:   Series, Vol. 182, Galaxy Dynamics - A Rutgers Symposium, ed. D.~R. {Merritt},

907:   M.~{Valluri}, \& J.~A. {Sellwood}, 307--+

908: 

909: \bibitem[{{Ghavamian} {et~al.}(2003){Ghavamian}, {Rakowski}, {Hughes}, \&

910:   {Williams}}]{ghavamian03}

911: {Ghavamian}, P., {Rakowski}, C.~E., {Hughes}, J.~P., \& {Williams}, T.~B. 2003,

912:   \apj, 590, 833

913: 

914: \bibitem[{{Habing} {et~al.}(2006){Habing}, {Sevenster}, {Messineo}, {van de

915:   Ven}, \& {Kuijken}}]{habing}

916: {Habing}, H.~J., {Sevenster}, M.~N., {Messineo}, M., {van de Ven}, G., \&

917:   {Kuijken}, K. 2006, \aap, 458, 151

918: 

919: \bibitem[{{H{\"a}fner} {et~al.}(2000){H{\"a}fner}, {Evans}, {Dehnen}, \&

920:   {Binney}}]{hafner}

921: {H{\"a}fner}, R., {Evans}, N.~W., {Dehnen}, W., \& {Binney}, J. 2000, \mnras,

922:   314, 433

923: 

924: \bibitem[{{Hartigan} {et~al.}(2000){Hartigan}, {Morse}, {Palunas}, {Bally}, \&

925:   {Devine}}]{hartigan00}

926: {Hartigan}, P., {Morse}, J., {Palunas}, P., {Bally}, J., \& {Devine}, D. 2000,

927:   \aj, 119, 1872

928: 

929: \bibitem[{{Howard} {et~al.}(2008){Howard}, {Rich}, {Reitzel}, {Koch}, {De

930:   Propris}, \& {Zhao}}]{howard08}

931: {Howard}, C.~D., {Rich}, R.~M., {Reitzel}, D.~B., {Koch}, A., {De Propris}, R.,

932:   \& {Zhao}, H. 2008, ArXiv e-prints, 807

933: 

934: \bibitem[{{Lewis} \& {Freeman}(1989)}]{freeman89}

935: {Lewis}, J.~R., \& {Freeman}, K.~C. 1989, \aj, 97, 139

936: 

937: \bibitem[{{Liszt} \& {Burton}(1980)}]{liszt80}

938: {Liszt}, H.~S., \& {Burton}, W.~B. 1980, \apj, 236, 779

939: 

940: \bibitem[{{Mao} \& {Paczy{\'n}ski}(2002)}]{mao02}

941: {Mao}, S., \& {Paczy{\'n}ski}, B. 2002, \mnras, 337, 895

942: 

943: \bibitem[{{Matteucci} \& {Romano}(1999)}]{matteucci99}

944: {Matteucci}, F., \& {Romano}, D. 1999, \apss, 265, 311

945: 

946: \bibitem[{{Minniti} {et~al.}(1995){Minniti}, {Olszewski}, {Liebert}, {White},

947:   {Hill}, \& {Irwin}}]{minniti95}

948: {Minniti}, D., {Olszewski}, E.~W., {Liebert}, J., {White}, S.~D.~M., {Hill},

949:   J.~M., \& {Irwin}, M.~J. 1995, \mnras, 277, 1293

950: 

951: \bibitem[{{Minniti} {et~al.}(1992){Minniti}, {White}, {Olszewski}, \&

952:   {Hill}}]{minniti92}

953: {Minniti}, D., {White}, S.~D.~M., {Olszewski}, E.~W., \& {Hill}, J.~M. 1992,

954:   \apjl, 393, L47

955: 

956: \bibitem[{{Minniti} \& {Zoccali}(2008)}]{minniti08}

957: {Minniti}, D., \& {Zoccali}, M. 2008, 245, 323

958: 

959: \bibitem[{{Mould}(1983)}]{mould83}

960: {Mould}, J.~R. 1983, \apj, 266, 255

961: 

962: \bibitem[{{Ng} {et~al.}(1996){Ng}, {Bertelli}, {Chiosi}, \& {Bressan}}]{ng96}

963: {Ng}, Y.~K., {Bertelli}, G., {Chiosi}, C., \& {Bressan}, A. 1996, \aap, 310,

964:   771

965: 

966: \bibitem[{{Osterbrock} \& {Martel}(1992)}]{osterbrock}

967: {Osterbrock}, D.~E., \& {Martel}, A. 1992, \pasp, 104, 76

968: 

969: \bibitem[{{Paczynski} \& {Stanek}(1998)}]{pac98}

970: {Paczynski}, B., \& {Stanek}, K.~Z. 1998, \apjl, 494, L219+

971: 

972: \bibitem[{{Paczynski} {et~al.}(1994){Paczynski}, {Stanek}, {Udalski},

973:   {Szymanski}, {Kaluzny}, {Kubiak}, \& {Mateo}}]{pac94}

974: {Paczynski}, B., {Stanek}, K.~Z., {Udalski}, A., {Szymanski}, M., {Kaluzny},

975:   J., {Kubiak}, M., \& {Mateo}, M. 1994, \aj, 107, 2060

976: 

977: \bibitem[{{Palunas} \& {Williams}(2000)}]{palunas}

978: {Palunas}, P., \& {Williams}, T.~B. 2000, \aj, 120, 2884

979: 

980: \bibitem[{{Press} {et~al.}(1992){Press}, {Teukolsky}, {Vetterling}, \&

981:   {Flannery}}]{press92}

982: {Press}, W.~H., {Teukolsky}, S.~A., {Vetterling}, W.~T., \& {Flannery}, B.~P.

983:   1992, {Numerical recipes in FORTRAN. The art of scientific computing}

984:   (Cambridge: University Press, |c1992, 2nd ed.)

985: 

986: \bibitem[{{Rangwala} {et~al.}(2008){Rangwala}, {Williams}, {Pietraszewski}, \&

987:   {Joseph}}]{rangwala08}

988: {Rangwala}, N., {Williams}, T.~B., {Pietraszewski}, C., \& {Joseph}, C.~L.

989:   2008, \aj, 135, 1825

990: 

991: \bibitem[{{Rattenbury} {et~al.}(2007{\natexlab{a}}){Rattenbury}, {Mao},

992:   {Debattista}, {Sumi}, {Gerhard}, \& {de Lorenzi}}]{rattenbury07a}

993: {Rattenbury}, N.~J., {Mao}, S., {Debattista}, V.~P., {Sumi}, T., {Gerhard}, O.,

994:   \& {de Lorenzi}, F. 2007{\natexlab{a}}, \mnras, 378, 1165

995: 

996: \bibitem[{{Rattenbury} {et~al.}(2007{\natexlab{b}}){Rattenbury}, {Mao}, {Sumi},

997:   \& {Smith}}]{rattenbury07b}

998: {Rattenbury}, N.~J., {Mao}, S., {Sumi}, T., \& {Smith}, M.~C.

999:   2007{\natexlab{b}}, \mnras, 378, 1064

1000: 

1001: \bibitem[{{Rich}(1988)}]{rich88}

1002: {Rich}, R.~M. 1988, \aj, 95, 828

1003: 

1004: \bibitem[{{Rich}(1990)}]{rich90}

1005: ---. 1990, \apj, 362, 604

1006: 

1007: \bibitem[{{Rich} {et~al.}(2007){Rich}, {Reitzel}, {Howard}, \& {Zhao}}]{rich07}

1008: {Rich}, R.~M., {Reitzel}, D.~B., {Howard}, C.~D., \& {Zhao}, H. 2007, \apjl,

1009:   658, L29

1010: 

1011: \bibitem[{{Sellwood} \& {Wilkinson}(1993)}]{sellwood93}

1012: {Sellwood}, J.~A., \& {Wilkinson}, A. 1993, Reports of Progress in Physics, 56,

1013:   173

1014: 

1015: \bibitem[{{Sharples} {et~al.}(1990){Sharples}, {Walker}, \&

1016:   {Cropper}}]{sharples}

1017: {Sharples}, R., {Walker}, A., \& {Cropper}, M. 1990, \mnras, 246, 54

1018: 

1019: \bibitem[{{Sluis} \& {Williams}(2006)}]{sluis06}

1020: {Sluis}, A.~P.~N., \& {Williams}, T.~B. 2006, \aj, 131, 2089

1021: 

1022: \bibitem[{{Spaenhauer} {et~al.}(1992){Spaenhauer}, {Jones}, \&

1023:   {Whitford}}]{span}

1024: {Spaenhauer}, A., {Jones}, B.~F., \& {Whitford}, A.~E. 1992, \aj, 103, 297

1025: 

1026: \bibitem[{{Stanek} {et~al.}(1994){Stanek}, {Mateo}, {Udalski}, {Szymanski},

1027:   {Kaluzny}, \& {Kubiak}}]{stanek94}

1028: {Stanek}, K.~Z., {Mateo}, M., {Udalski}, A., {Szymanski}, M., {Kaluzny}, J., \&

1029:   {Kubiak}, M. 1994, \apjl, 429, L73

1030: 

1031: \bibitem[{{Stanek} {et~al.}(1997){Stanek}, {Udalski}, {Szymanski}, {Kaluzny},

1032:   {Kubiak}, {Mateo}, \& {Krzeminski}}]{stanek97}

1033: {Stanek}, K.~Z., {Udalski}, A., {Szymanski}, M., {Kaluzny}, J., {Kubiak}, M.,

1034:   {Mateo}, M., \& {Krzeminski}, W. 1997, \apj, 477, 163

1035: 

1036: \bibitem[{{Stetson}(1987)}]{stetson87}

1037: {Stetson}, P.~B. 1987, \pasp, 99, 191

1038: 

1039: \bibitem[{{Stetson}(1994)}]{stetson94}

1040: ---. 1994, \pasp, 106, 250

1041: 

1042: \bibitem[{{Sumi}(2004)}]{sumi}

1043: {Sumi}, T. 2004, \mnras, 349, 193

1044: 

1045: \bibitem[{{Sumi} {et~al.}(2004){Sumi}, {Wu}, {Udalski}, {Szyma{\'n}ski},

1046:   {Kubiak}, {Pietrzy{\'n}ski}, {Soszy{\'n}ski}, {Wo{\'z}niak},

1047:   {{\.Z}ebru{\'n}}, {Szewczyk}, \& {Wyrzykowski}}]{sumi04}

1048: {Sumi}, T., {Wu}, X., {Udalski}, A., {Szyma{\'n}ski}, M., {Kubiak}, M.,

1049:   {Pietrzy{\'n}ski}, G., {Soszy{\'n}ski}, I., {Wo{\'z}niak}, P.,

1050:   {{\.Z}ebru{\'n}}, K., {Szewczyk}, O., \& {Wyrzykowski}, {\L}. 2004, \mnras,

1051:   348, 1439

1052: 

1053: \bibitem[{{Szymanski}(2005)}]{database}

1054: {Szymanski}, M.~K. 2005, Acta Astronomica, 55, 43

1055: 

1056: \bibitem[{{Terndrup} {et~al.}(1995){Terndrup}, {Sadler}, \& {Rich}}]{terndrup}

1057: {Terndrup}, D.~M., {Sadler}, E.~M., \& {Rich}, R.~M. 1995, \aj, 110, 1774

1058: 

1059: \bibitem[{{Udalski} {et~al.}(1992){Udalski}, {Szymanski}, {Kaluzny}, {Kubiak},

1060:   \& {Mateo}}]{udalski}

1061: {Udalski}, A., {Szymanski}, M., {Kaluzny}, J., {Kubiak}, M., \& {Mateo}, M.

1062:   1992, Acta Astronomica, 42, 253

1063: 

1064: \bibitem[{{van Albada} \& {Sanders}(1982)}]{sanders82}

1065: {van Albada}, T.~S., \& {Sanders}, R.~H. 1982, \mnras, 201, 303

1066: 

1067: \bibitem[{{van de Ven} {et~al.}(2006){van de Ven}, {van den Bosch}, {Verolme},

1068:   \& {de Zeeuw}}]{glenn06}

1069: {van de Ven}, G., {van den Bosch}, R.~C.~E., {Verolme}, E.~K., \& {de Zeeuw},

1070:   P.~T. 2006, \aap, 445, 513

1071: 

1072: \bibitem[{{van der Marel} \& {Franx}(1993)}]{gh93}

1073: {van der Marel}, R.~P., \& {Franx}, M. 1993, \apj, 407, 525

1074: 

1075: \bibitem[{{Weiner} \& {Sellwood}(1999)}]{weiner99}

1076: {Weiner}, B.~J., \& {Sellwood}, J.~A. 1999, \apj, 524, 112

1077: 

1078: \bibitem[{{Z{\'a}nmar S{\'a}nchez} {et~al.}(2008){Z{\'a}nmar S{\'a}nchez},

1079:   {Sellwood}, {Weiner}, \& {Williams}}]{zanmar08}

1080: {Z{\'a}nmar S{\'a}nchez}, R., {Sellwood}, J.~A., {Weiner}, B.~J., \&

1081:   {Williams}, T.~B. 2008, \apj, 674, 797

1082: 

1083: \bibitem[{{Zhao}(1996)}]{zhao96}

1084: {Zhao}, H.~S. 1996, \mnras, 283, 149

1085: 

1086: \bibitem[{{Zoccali} {et~al.}(2007){Zoccali}, {Lecureur}, {Barbuy}, {Hill},

1087:   {Renzini}, {Minniti}, {Momany}, {G{\'o}mez}, \& {Ortolani}}]{zoccali07}

1088: {Zoccali}, M., {Lecureur}, A., {Barbuy}, B., {Hill}, V., {Renzini}, A.,

1089:   {Minniti}, D., {Momany}, Y., {G{\'o}mez}, A., \& {Ortolani}, S. 2007, 241, 73

1090: 

1091: \end{thebibliography}

1092: 

1093: 

1094: \clearpage

1095: 

1096: \clearpage

1097: \begin{deluxetable}{ccccrcccccc}

1098: \tabletypesize{\scriptsize} \tablecaption{Radial Velocities and Equivalent Widths}

1099: \tablehead{\colhead{ID}& \colhead{OGLEID} & \colhead{R.A.} & \colhead{Dec.} & \colhead{$V$} &

1100: \colhead{$V_{\textrm{err}}$} & \colhead{EW} & \colhead{EW$_{\textrm{err}}$} & \colhead{I} & \colhead{V-I} & \colhead{Source}\\

1101: \colhead{} & \colhead{}& \colhead{j2000} & \colhead{j2000} & \colhead{km $\textrm{s}^{-1}$} &

1102: \colhead{km $\textrm{s}^{-1}$} & \colhead{\AA} & \colhead{\AA} & \colhead{} & \colhead{} &

1103: \colhead{}}

1104: 

1105: \startdata

1106: BWCa-1  &        256217  &     18:03:35.16  &     -29:58:12.3  &     -4.49  &    1.60  &      4.64  &      0.17  &    11.560  &     2.117  &  o  \\

1107: BWCa-2  &        256200  &     18:03:31.91  &     -30:00:28.3  &    -27.71  &    2.55  &      2.12  &      0.12  &    11.523  &     3.960  &  o  \\

1108: BWCa-3  &        256193  &     18:03:31.67  &     -30:00:43.4  &    -44.86  &    1.91  &      4.55  &      0.19  &    11.730  &     2.522  &  o  \\

1109: BWCa-4  &        402256  &     18:03:43.52  &     -30:01:46.1  &    -17.77  &    1.70  &      4.34  &      0.15  &    11.733  &     1.358  &  o  \\

1110: BWCa-5  &        256201  &     18:03:30.78  &     -30:00:27.9  &     92.33  &    2.41  &      3.79  &      0.18  &    12.241  &     2.702  &  o  \\

1111: BWCa-6  &        256199  &     18:03:32.87  &     -30:00:31.4  &      9.84  &    3.60  &      2.42  &      0.20  &    12.098  &     0.710  &  o  \\

1112: BWCa-8  &        402258  &     18:03:45.27  &     -30:01:32.5  &    -74.90  &    3.48  &      2.94  &      0.22  &    12.214  &     3.876  &  o  \\

1113: BWCa-9  &          --  &     18:03:36.34  &     -30:01:50.2  &    -20.10  &    3.03  &      3.80  &      0.23  &    12.020  &  --  &  f  \\

1114: BWCa-11  &        244433  &     18:03:34.54  &     -30:01:37.6  &     32.20  &    3.41  &      1.20  &      0.11  &    12.229  &     4.566  &  o  \\

1115: BWCa-12  &        412692  &     18:03:38.98  &     -29:58:25.9  &    -38.04  &    2.72  &      5.65  &      0.31  &    12.215  &     2.595  &  o  \\

1116: BWCa-13  &        423223  &     18:03:41.96  &     -29:57:49.1  &    -70.86  &    2.80  &      4.07  &      0.25  &    12.289  &     2.120  &  o  \\

1117: BWCa-15  &        412682  &     18:03:38.18  &     -30:00:51.2  &    -16.60  &    2.91  &      4.07  &      0.23  &    12.467  &     1.912  &  o  \\

1118: BWCa-16  &        412681  &     18:03:38.24  &     -30:00:59.7  &     67.05  &    2.41  &      4.64  &      0.23  &    12.476  &     2.346  &  o  \\

1119: BWCa-19  &        412690  &     18:03:36.96  &     -29:58:47.1  &    -58.58  &    2.45  &      6.04  &      0.31  &    12.483  &     2.094  &  o  \\

1120: BWCa-20  &        256196  &     18:03:33.66  &     -30:00:36.1  &    -24.73  &    2.47  &      3.46  &      0.20  &    12.579  &     1.894  &  o  \\

1121: BWCa-21  &        412685  &     18:03:42.64  &     -30:00:06.5  &    131.70  &    3.01  &      4.69  &      0.32  &    12.493  &     2.927  &  o  \\

1122: BWCa-22  &        402257  &     18:03:37.71  &     -30:01:34.0  &    -15.85  &    2.90  &      3.79  &      0.22  &    12.637  &     2.040  &  o  \\

1123: BWCa-23  &        412683  &     18:03:47.96  &     -30:00:38.1  &    -17.18  &    3.30  &      4.40  &      0.32  &    12.627  &     2.948  &  o  \\

1124: BWCa-27  &        412719  &     18:03:39.69  &     -29:58:47.5  &   -108.48  &    3.33  &      6.48  &      0.35  &    12.787  &     2.576  &  o  \\

1125: BWCa-28  &        412708  &     18:03:49.05  &     -29:59:35.0  &     -7.34  &    2.95  &      5.23  &      0.35  &    12.905  &     2.475  &  o  \\

1126: BWCa-29  &        256222  &     18:03:34.70  &     -30:01:15.7  &    -18.07  &    3.08  &      3.48  &      0.23  &    12.923  &     1.883  &  o  \\

1127: BWCa-30  &        412702  &     18:03:37.64  &     -30:00:26.0  &    -45.72  &    6.02  &      3.66  &      0.41  &    12.877  &     0.925  &  o  \\

1128: BWCa-32  &        256236  &     18:03:34.96  &     -29:59:48.6  &    -76.35  &    5.27  &      4.78  &      0.45  &    12.982  &     3.493  &  o  \\

1129: BWCa-33  &        412727  &     18:03:45.61  &     -29:58:12.0  &   -132.32  &    3.95  &      4.81  &      0.40  &    13.174  &     2.579  &  o  \\

1130: BWCa-34  &        412711  &     18:03:41.51  &     -29:59:20.3  &   -134.86  &    3.59  &      4.75  &      0.26  &    13.081  &     2.852  &  o  \\

1131: BWCa-35  &        256235  &     18:03:30.56  &     -29:59:51.0  &     44.72  &    3.22  &      4.07  &      0.29  &    13.173  &     2.307  &  o  \\

1132: BWCa-36  &        256238  &     18:03:32.78  &     -29:59:39.1  &    143.70  &    2.71  &      4.15  &      0.27  &    13.132  &     2.736  &  o  \\

1133: BWCa-38  &        412728  &     18:03:42.44  &     -29:58:10.6  &   -168.91  &    3.82  &      5.29  &      0.34  &    13.205  &     2.625  &  o  \\

1134: BWCa-39  &        412712  &     18:03:47.60  &     -29:59:15.4  &    -65.47  &    3.57  &      5.50  &      0.38  &    13.223  &     2.865  &  o  \\

1135: BWCa-40  &        412718  &     18:03:38.12  &     -29:58:58.3  &      3.29  &    3.78  &      5.35  &      0.39  &    13.150  &     2.474  &  o  \\

1136: BWCa-42  &        412732  &     18:03:37.09  &     -29:57:56.3  &      4.77  &    3.54  &      2.35  &      0.22  &    13.133  &     1.571  &  o  \\

1137: BWCa-43  &        412724  &     18:03:39.04  &     -29:58:28.0  &   -153.07  &    4.33  &      5.49  &      0.39  &    13.173  &     1.917  &  o  \\

1138: BWCa-44  &        412704  &     18:03:38.05  &     -30:00:10.2  &    -61.60  &    4.52  &      3.80  &      0.27  &    13.141  &     1.419  &  o  \\

1139: BWCa-45  &        412716  &     18:03:47.48  &     -29:59:06.4  &    110.99  &    2.89  &      4.14  &      0.28  &    13.309  &     2.530  &  o  \\

1140: BWCa-46  &        412699  &     18:03:46.37  &     -30:00:47.9  &   -130.64  &    3.93  &      4.99  &      0.36  &    13.333  &     2.210  &  o  \\

1141: BWCa-47  &        412720  &     18:03:45.08  &     -29:58:46.3  &     89.38  &    3.35  &      4.34  &      0.31  &    13.284  &     2.242  &  o  \\

1142: BWCa-48  &        402288  &     18:03:40.62  &     -30:01:46.0  &    -10.80  &    4.33  &      3.03  &      0.30  &    13.377  &     1.751  &  o  \\

1143: BWCa-49  &        412729  &     18:03:39.95  &     -29:58:07.8  &   -102.93  &    4.34  &      4.77  &      0.37  &    13.295  &     1.956  &  o  \\

1144: BWCa-50  &        256226  &     18:03:36.45  &     -30:00:39.8  &    -13.72  &    3.74  &      5.04  &      0.39  &    13.311  &     2.713  &  o  \\

1145: %BWCa-51  &        412705  &     18:03:38.93  &     -29:59:52.8  &    -42.56  &    3.89  &      4.48  &      0.28  &    13.268  &     2.114  &  o  \\

1146: %BWCa-52  &        244493  &     18:03:33.92  &     -30:01:34.3  &     98.50  &    4.68  &      4.86  &      0.47  &    13.509  &     2.395  &  o  \\

1147: %BWCa-53  &        256239  &     18:03:33.89  &     -29:59:32.6  &    -23.32  &    9.72  &      3.36  &      0.78  &    13.248  &     0.639  &  o  \\

1148: %BWCa-54  &        256231  &     18:03:36.02  &     -30:00:21.4  &    -27.40  &    4.12  &      4.72  &      0.37  &    13.325  &     2.050  &  o  \\

1149: %BWCa-55  &        412710  &     18:03:44.47  &     -29:59:28.6  &    -30.94  &    3.61  &      5.69  &      0.42  &    13.373  &     2.592  &  o  \\

1150: %BWCa-56  &        412700  &     18:03:48.57  &     -30:00:33.3  &   -139.76  &    3.60  &      4.91  &      0.34  &    13.482  &     2.237  &  o  \\

1151: %BWCa-57  &        402290  &     18:03:39.22  &     -30:01:40.1  &    -13.48  &    4.86  &      2.84  &      0.30  &    13.513  &     1.471  &  o  \\

1152: %BWCa-59  &        412706  &     18:03:39.31  &     -29:59:49.3  &      7.99  &    3.89  &      4.95  &      0.42  &    13.407  &     2.087  &  o  \\

1153: %BWCa-60  &        244605  &     18:03:35.04  &     -30:01:35.1  &    -13.31  &    4.93  &      3.40  &      0.36  &    13.640  &     1.803  &  o  \\

1154: %BWCa-61  &        256228  &     18:03:32.48  &     -30:00:26.6  &    -79.88  &    5.07  &      3.60  &      0.42  &    13.565  &     1.648  &  o  \\

1155: %BWCa-62  &        412717  &     18:03:42.74  &     -29:58:59.2  &    122.58  &    3.82  &      4.73  &      0.40  &    13.495  &     2.205  &  o  \\

1156: \enddata

1157:  \tablecomments{Table \ref{data} is published in its entirety in the electronic edition of the

1158: {\it Astrophysical Journal}.  A portion is shown here for guidance regarding its form and

1159: content.\label{data}} \tablenotetext{a}{Source of Photometry: `o' for OGLE, `s' for SAAO and `f'

1160: for fabry-p\'{e}rot}

1161: \end{deluxetable}

1162: 

1163: 

1164: 

1165: \begin{deluxetable}{lcccc}

1166: \tabletypesize{\scriptsize} \tablecaption{Heliocentric kinematics of the red clump giants for

1167: three LOS in the Milky Way bar.\label{kine}}

1168:  \tablehead{

1169:  \colhead{Method} & \colhead{$<V_{\textrm{los}}>$} & \colhead{$\sigma_{\textrm{los}}$} & \colhead{$h_3$} & \colhead{$h_4$}}

1170: % \colhead{} & \colhead{kms} & \colhead{kms} & \colhead{kms} & \colhead{kms}}

1171:  \startdata

1172:  \cutinhead{MM5B-(a+c); $(l,b) = (-5.0, -3.5)$; $N_{\textrm{stars}} = 804$}

1173:  Likelihood & $-54.15 \pm 3.66\, (\pm 6.14)$ & $101.72 \pm 0.66\, (\pm 0.80)$ & $-0.030 \pm 0.017\, (\pm 0.034)$

1174:   & $-0.032 \pm 0.008\, (\pm 0.017)$\\

1175:  %Likelihood (Bootstrap)& $-54.15 \pm 6.14$ & $101.72 \pm 0.80$ & $-0.030 \pm 0.034$ & $-0.032 \pm 0.017$\\

1176:  Biweight  & $-52.56 \pm 0.60\, (\pm 3.85)$ & $102.30 \pm 1.25\, (\pm 2.50)$ & \nodata & \nodata\\

1177:  %Biweight (Bootstrap) & $-52.56 \pm 3.85$ & $102.30 \pm 2.50$ & \nodata & \nodata\\

1178:  \cutinhead{MM5B-(a+c+d); $(l,b) = (-5.0, -3.5)$; $N_{\textrm{stars}} = 1193$}

1179:  Likelihood  & $-56.46 \pm 1.84\, (\pm 4.48)$ & $102.04 \pm 0.09\, (\pm 0.28)$ & $-0.018 \pm 0.007\, (\pm 0.018)$

1180:  & $-0.031 \pm 0.005\, (\pm 0.012)$ \\

1181:  %Likelihood (Bootstrap) & $-56.46 \pm 4.48$ & $102.04 \pm 0.28$ & $-0.018 \pm 0.018$ & $-0.031 \pm 0.012$ \\

1182:  Biweight  & $-58.57 \pm 0.48 \, (\pm 3.35)$ & $101.46 \pm 1.21\, (\pm 2.03)$ & \nodata & \nodata \\

1183:  %Biweight (Bootstrap) & $-58.57 \pm 3.35$ & $101.46 \pm 2.03$ & \nodata & \nodata \\

1184:  \cutinhead{MM7B-(a+c+d); $(l,b) = (5.5, -3.5)$; $N_{\textrm{stars}} = 738$}

1185:  Likelihood & $35.73 \pm 7.20\,(\pm 9.44)$ & $102.40 \pm 0.15\,(\pm 0.31)$ & $-0.016 \pm 0.031\,(\pm 0.043)$

1186:  & $-0.047 \pm 0.022\, (\pm 0.026)$\\

1187:  %Likelihood (Bootstrap) & $35.73 \pm 9.44$ & $102.40 \pm 0.31$ & $-0.016 \pm 0.043$ & $-0.047 \pm 0.026$\\

1188:  Biweight & $30.73 \pm 0.58\, (\pm 3.89)$ & $102.38 \pm 1.21\, (\pm 2.39)$ & \nodata & \nodata \\

1189:  %Biweight (Bootstrap) & $30.73 \pm 3.89$ & $102.38 \pm 2.39$ & \nodata & \nodata \\

1190:  \cutinhead{BWC-a; $(l,b) = (1.06, -3.9)$; $N_{\textrm{stars}} = 557$}

1191:  Likelihood  & $4.32 \pm 0.26\,(\pm 5.21)$ & $110.29 \pm 0.36\,(\pm 0.73)$ & $-0.031 \pm 0.011\, (\pm 0.019)$

1192:  & $-0.035 \pm 0.009\, (\pm 0.017)$\\

1193:  BW (B)  & $-1.01 \pm 5.42$ & $112.62 \pm 3.41$ & \nodata & \nodata \\

1194:  NGC 6522 (N) & $-14.67 \pm 3.77$ & $7.9 \pm 2.82$ & & \\

1195:  %Likelihood (Bootstrap) & $4.32 \pm 5.21$ & $110.29 \pm 0.73$ & $-0.031 \pm 0.019$ & $-0.035 \pm 0.017$\\

1196:  \enddata

1197:  \tablecomments{There are two types of errors for the velocity moments: Gaussian

1198:   randomization and bootstrap (in parenthesis).\\

1199:  For BW kinematics the B and N stand for the broad and narrow Gaussian components.}

1200: \end{deluxetable}

1201: 

1202: 

1203: 

1204: \begin{deluxetable}{ccrcccr}

1205:  %\small

1206:  %\tablewidth{5.2in}

1207:  \tablecaption{Radial streaming motions.\label{streamdata}}

1208:  \tablehead{\colhead{LOS} & \colhead{I$_{\mathrm{peak}}$} & \colhead{$\langle V_{\textrm{bright}}\rangle$} & \colhead{$N_{\textrm{bright}}$}

1209:  & \colhead{$\langle V_{\textrm{faint}}\rangle$} & \colhead{$N_{\textrm{faint}}$} & \colhead{$\Delta V$}\\

1210:  \colhead{(l,b)} & \colhead{} & \colhead{(\kms)} & \colhead{} & \colhead{(\kms)} & \colhead{}  & \colhead{(\kms)}}

1211:  \startdata

1212:  %$(+5.5, -3.5)$ & 6.7 -- 8.0 & $13.57 \pm 7.14$ & 186 & 9.7 -- 11.6 & $52.90 \pm 8.0$ & 140 & $-39.69 \pm 11.0$ \\

1213:  %$(+1.06, -3.9)$ & 6.2 -- 7.8 & $-11.61 \pm 7.14$ & 103 & 9.4 -- 11.9 & $-5.71 \pm 8.0$ & 133 & $-5.91 \pm 11.0$ \\

1214:  %$(-5.5, -3.7)$ & 7.5 -- 9.0 & $ -73.48 \pm 7.14$ & 174 & 10.8 -- 12.9 $41.81 \pm 8.0$ & 167 & $-31.67 \pm 11.0$ \\

1215:  $(+5.5, -3.5)$ & 15.4 & $13.57 \pm 7.5$ & 186 & $52.90 \pm 8.3$ & 140 &  $-39.69 \pm 11.2 $  \\

1216:  $(+1.06, -3.9)$ & 15.2 &$-7.05 \pm 11.3$ & 94 &  $-2.64 \pm 10.7$ & 109 & $-4.41 \pm 15.5$ \\

1217:  $(-5.0, -3.5)$ & 15.6 & $ -73.48 \pm 8.1$ & 174 & $41.81 \pm 8.1$ & 167 & $-31.67 \pm 11.5$ \\

1218:  \enddata

1219: \end{deluxetable}

1220: 

1221: 

1222: 

1223: 

1224: 

1225: \end{document}

1226: