0806.3004/ms.tex
1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
2: % 
3: %  FY CMa  
4: % 
5: % 
6: %  Revision history: 
7: % 
8: %  August 13, 2007:  First draft by Gies
9: %  January 29, 2008: Revised figures, text 
10: %  February 27, 2008: Text revisions & text/figure addition by Peters
11: %  March 19, 2008: Additional text revisions by Peters
12: %  June 5, 2008: Text revisions by Peters
13: %  June 10, 2008: Text revisions by Peters
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
15: % 
16: %\documentclass{aastex} 
17: %\usepackage{emulateapj5} 
18: \documentclass[12pt,preprint]{aastex} 
19: %\voffset 0.7truein 
20: % 
21: % 
22: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
23:  
24: \shorttitle{Hot Subdwarf Companion to the Be Star FY~CMa} 
25: \shortauthors{Peters et al.} 
26:  
27: \begin{document} 
28:  
29: \received{2008 March 21} 
30: \accepted{} 
31:  
32: \title{Detection of a Hot Subdwarf Companion to the Be Star FY~Canis Majoris}  
33:  
34: \author{Geraldine J. Peters\altaffilmark{1,2}} 
35: \affil{Space Sciences Center, University of Southern California, Los Angeles, CA 90089-1341;  
36: gjpeters@mucen.usc.edu} 
37:  
38: \author{Douglas R. Gies\altaffilmark{1}, Erika D. Grundstrom\altaffilmark{1,3}} 
39: \affil{Center for High Angular Resolution Astronomy and  
40:  Department of Physics and Astronomy,\\ 
41:  Georgia State University, P. O. Box 4106, Atlanta, GA 30302-4106;  
42:  gies@chara.gsu.edu, erika.grundstrom@vanderbilt.edu} 
43:  
44: \author{M. Virginia McSwain\altaffilmark{1}} 
45: \affil{Department of Physics, Lehigh University, Bethlehem, PA 18105; 
46:  mcswain@lehigh.edu
47: } 
48:  
49: \altaffiltext{1}{Visiting Astronomer, Kitt Peak National Observatory, 
50: National Optical Astronomy Observatory, operated by the Association 
51: of Universities for Research in Astronomy, Inc., under contract with 
52: the National Science Foundation.} 
53: \altaffiltext{2}{Guest Observer with the International Ultraviolet Explorer
54: Satellite.} 
55: \altaffiltext{3}{Current address: Department of Physics and Astronomy, 
56: Vanderbilt University, Nashville, TN 37206.} 
57: 
58:  
59: %\slugcomment{Second Submission (6/11/2008)} 
60: %\paperid{} 
61:  
62: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
63:  
64: \begin{abstract} 
65: The rapid rotation of Be stars may be caused in some cases 
66: by past mass and angular momentum accretion in an interacting 
67: binary in which the mass donor is currently viewed as a small, hot
68: subdwarf stripped of its outer envelope. Here we report on the 
69: spectroscopic detection of such a subdwarf in the Be binary system 
70: FY~Canis Majoris from the analysis of data acquired by the {\it IUE} 
71: spacecraft and KPNO Coud\'{e} Feed Telescope over the course of 
72: 16 and 21 years, respectively. 
73: We present a double-lined spectroscopic orbit for the binary
74: based upon radial velocities from the {\it IUE} spectra 
75: and use the orbital solutions with a Doppler tomography 
76: algorithm to reconstruct the components' UV spectra. 
77: The subdwarf is hot ($T_{\rm eff} = 45\pm5$~kK) and has a mass of
78: about $1.3 M_\odot$ and a radius of about $0.6~R_\odot$. 
79: It contributes about $4\%$ as much flux as the Be star does in the 
80: FUV.  We also present observations of the H$\alpha$ and \ion{He}{1} 
81: $\lambda 6678$ emission features that are formed in the 
82: circumstellar disk of the Be star.  Orbital flux and velocity 
83: variations in the \ion{He}{1} $\lambda 6678$ 
84: profile indicate that much of the emission forms along the 
85: disk rim facing the hot subdwarf where the disk is 
86: probably heated by the incident radiation from the subdwarf. 
87: A study of
88: the FUV infall shell lines discovered in the 1980s confirms 
89: their episodic presence but reveals that they tend to be found 
90: around both quadrature phases, unlike the pattern in Algol 
91: binaries. Phase-dependent variations in the UV \ion{N}{5} doublet 
92: suggest the presence of a N-enhanced wind from the subdwarf and a 
93: possible shock-interaction region between the stars where 
94: the subdwarf's wind collides with the disk of the Be star. 
95: 
96: 
97:   
98: \end{abstract} 
99:  
100: \keywords{stars: emission-line, Be  
101: --- stars: individual (HR~2855, HD~58978, FY~CMa)  
102: --- stars: binaries: spectroscopic  
103: --- stars: evolution  
104: --- stars: subdwarfs} 
105:  
106: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
107:  
108: \setcounter{footnote}{3} 
109:  
110: \section{Introduction}                              % Section 1 
111:  
112: The component stars of massive interacting binaries experience  
113: profound transformations over their lifetimes.  Their evolutionary  
114: stages are defined by how mass and angular momentum are shared and  
115: lost during the mass transfer stages \citep*{wel01}.  The originally 
116: more massive star grows faster and upon reaching a size comparable to  
117: its Roche lobe begins mass and angular momentum transfer to the  
118: mass gainer companion.  The orbit shrinks until the mass ratio reverses, 
119: but mass transfer will continue with an expanding orbit as long as  
120: the donor grows in proportion.  The result will be a donor that is  
121: stripped of most of its outer envelope and a gainer that now appears  
122: rejuvenated by mass transfer \citep{dra07}.  Mass transfer may  
123: also cause the gainer to spin up to a rotational speed close to the 
124: critical value where centripetal and gravitational accelerations  
125: are equal at the equator \citep*{pet05}.  \citet{pol91} and 
126: \citet{van97} suggest that these mass gainer stars may include 
127: some of the rapidly rotating, classical Be stars,  
128: B-type stars that display emission lines formed in outflowing,  
129: circumstellar disks \citep{por03}. 
130:  
131: There is now a substantial body of evidence that the supports  
132: the idea that some Be stars were spun up by mass transfer in interacting  
133: binaries.  Most of the known massive X-ray binaries consist of a Be star  
134: and neutron star companion, the remains of the former donor star \citep{coe00}. 
135: Donor stars that end up with a mass lower than the Chandrasekhar limit 
136: will probably avoid a supernova explosion, but will appear first as a  
137: hot subdwarf star and then eventually as a massive white dwarf.  Although  
138: there are no known examples of Be plus white dwarf binaries, there are  
139: a few identified cases of Be plus hot subdwarf binaries.  The first  
140: discovery of the spectrum of a subdwarf companion was made from the 
141: analysis of UV spectra from the   
142: {\it International Ultraviolet Explorer (IUE)} spacecraft \citep{tha95}  
143: and the {\it Hubble Space Telescope}  
144: \citep{gie98} of the well-known Be star $\phi$~Persei.  The hot companion in  
145: \object{$\phi$~Per} is an object of solar mass and radius but with an effective  
146: temperature of $53$~kK.  However, the hot subdwarf is much fainter  
147: than the Be star, and its presence is perhaps best observed through its heating  
148: effects on the disk gas nearest to the subdwarf \citep*{ste00,hum01}.  
149: \citet{mai05} show that the binary 59~Cygni shows many of the same  
150: characteristics found in $\phi$~Per and is probably the second example 
151: of a Be plus hot subdwarf system.  A third candidate, 
152: \object{HR~2142} \citep{pet83}, may also  
153: belong this category \citep{wat91} but at the present the evidence is 
154: not compelling.  The fact that the nature of the companions in $\phi$~Per 
155: and \object{59~Cyg} were only discovered after a century of spectroscopic 
156: observation, however, highlights the difficulties 
157: of their detection \citep{gie00}.  
158:  
159: Here we report on the identification of a hot subdwarf companion in the Be 
160: binary \object{FY~CMa} (HR 2855, HD 58978).  Balmer emission lines in the 
161: Be star primary in this system (B0.5~IVnpe,  
162: \citealt*{hil69}; B0.5~IVe, \citealt*{sle82}) were first discovered 
163: by \citet{pic05}.
164: Although at first glance it appears to be a classical Be star,
165: it has historically displayed much more spectroscopic variability
166: in its optical and UV spectrum than is typical for Be stars
167: \citep{sle82,pet82,pet88,gra88,cao01}.  Some of the first  
168: {\it IUE} observations of FY~CMa showed evidence of sharp, red-shifted, 
169: ``shell'' spectral lines that varied significantly on a timescale of days  
170: \citep{gra88}.  \citet{pet88} argued that the time scales of  
171: variability indicated a binary origin. Spectroscopic variability with a  
172: period of 37.26~d was subsequently found by \citet{riv04} who showed that 
173: the strength and profile of the \ion{N}{5} $\lambda 1240$  
174: wind line fluctuates with this period.  \citet{riv04} also  
175: presented observations of the H$\alpha$ and \ion{He}{1} $\lambda6678$  
176: emission lines that suggested that they also varied in radial velocity  
177: and strength in much the same way as observed in $\phi$~Per and 59~Cyg,
178: and suggested that FY~CMa might be generically similar to these systems.  
179: However, \citet{riv04} could not confirm binarity as the orbital 
180: elements and the nature of the companion remained unknown. 
181:  
182: Below we present an investigation of FY~CMa that is based upon 
183: a re-examination of the {\it IUE} spectra and  
184: an analysis of a large collection of optical spectra obtained between  
185: 1985 and 2006 (\S2).  We first discuss the use of cross-correlation  
186: measurements of the {\it IUE} spectra to obtain radial velocities for  
187: both components and to derive the orbital elements (\S3).  We then  
188: use the orbital velocity curves to reconstruct the individual UV spectra  
189: of the components using a Doppler tomography algorithm, and we compare 
190: the reconstructed spectra to models to estimate basic stellar properties (\S4). 
191: We then describe the orbital and temporal variations we find in the  
192: H$\alpha$ and \ion{He}{1} $\lambda6678$ emission lines (\S5).  We argue 
193: in \S6 that the emission line variations are probably due to heating  
194: of the part of the disk rim that faces the hot subdwarf companion.
195: We also discuss the long-term and phase-dependent variability in the 
196: strength and velocity of the \ion{Fe}{3} shell-type infall lines that were 
197: discovered in the mid-1980s and suggest two possible mechanisms for their 
198: formation. Finally, we show that phase-dependent profile, strength, and 
199: velocity variability in the \ion{N}{5} wind lines suggest that
200: the subdwarf's wind is N-enhanced and that there is a significant 
201: interaction between the subdwarf's wind and the disk of the Be 
202: star.   
203:  
204: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
205:  
206: \section{Observations}                              % Section 2 
207:  
208: Ninety-seven high resolution FUV spectra were obtained with {\it IUE} 
209: over the period from 1979 to 1995.  
210: The spectra were secured with the Short Wavelength Prime (SWP) camera 
211: in echelle mode, and the extracted and flux calibrated  
212: spectra from the final archive were downloaded from the Multimission Archive at 
213: the Space Telescope Science Institute\footnote{http://archive.stsci.edu/iue/}. 
214: We binned each spectrum onto a uniform $\log\lambda$ wavelength grid  
215: (bin size equivalent to 10 km~s$^{-1}$),  
216: aligned them with a global average using cross--correlation 
217: shifts from the vicinity of strong interstellar lines (after which the 
218: interstellar lines are removed by interpolation), and then rectified the 
219: fluxes by a spline fit to a series of pseudo-continuum zones.   
220: We discuss the radial velocity measurements and the appearance of the  
221: UV spectra in the next two sections.  
222:  
223: We also obtained 158 red spectra of FY~CMa with the Kitt Peak  
224: National Observatory 0.9~m Coud\'{e} Feed Telescope between 1985 and  
225: 2006.  The various observing runs are summarized in Table~1 that 
226: lists the heliocentric Julian dates of observation, 
227: Besselian year range, wavelength range, 
228: spectral resolving power, number of individual spectra 
229: obtained, and a summary of the instrumental configuration. 
230: The spectra all record the H$\alpha$ $\lambda 6563$ emission line  
231: and all but six also record the \ion{He}{1} $\lambda 6678$ feature.  
232: All the spectra were made with the long collimator, camera 5,  
233: and either grating A (632 grooves mm$^{-1}$) or  
234: grating B (312 grooves mm$^{-1}$) in second order.   
235: The CCD detector of choice changed over the years  
236: as better instruments became available, and these are  
237: listed in the final column using their standard KPNO  
238: designations.  Most of the spectra are well exposed 
239: and have a S/N of 200 or better in the continuum. 
240:  
241: \placetable{tab1}      % Table 1 - Journal of runs 
242:  
243: The spectra were extracted and calibrated 
244: using standard routines in IRAF\footnote{IRAF is distributed by the 
245: National Optical Astronomy Observatory, which is operated by 
246: the Association of Universities for Research in Astronomy, Inc., 
247: under cooperative agreement with the National Science Foundation.}. 
248: All the spectra were rectified to a unit continuum by fitting 
249: line-free regions with a low-order polynomial.  During most of the runs  
250: from 2000 to 2006, we also observed a number of rapidly rotating  
251: A-type stars (primarily $\zeta$~Aql) for the removal of  
252: atmospheric telluric lines from the target spectra.  We  
253: created a library of atmospheric spectra for each run by removing 
254: the broad stellar features, and then we divided each target 
255: spectrum by the modified atmospheric spectrum that most closely 
256: matched the target spectrum in a selected region dominated by 
257: atmospheric absorptions.  The spectra from runs before 2000 were  
258: transformed to the wavelength grid for the fourth to last run listed  
259: in Table~1 and then were corrected for telluric absorption using the atmospheric 
260: spectra from that run (sometimes with less than optimal results because 
261: of differences in spectral resolution).  The spectra from each run were then 
262: transformed to a common heliocentric wavelength grid.  
263: Measurements of these red spectra are presented in \S5. 
264:  
265: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
266:  
267: \section{Radial Velocities and Orbital Elements}    % Section 3 
268:  
269: The first detection of the UV spectrum of the hot subdwarf in  
270: $\phi$~Per was based upon {\it IUE} spectra that recorded many  
271: lines of high excitation transitions especially in the 1310 -- 1385 \AA 
272: ~range \citep{tha95}.  Thus, we began our search for evidence of  
273: a hot companion by calculating cross-correlation functions (ccfs) of  
274: each spectrum with a template spectrum within this restricted  
275: wavelength range (where the lines of the primary are relatively  
276: broad and weak).  We formed the template spectrum from the synthetic  
277: spectral atlas associated with the non-LTE, line-blanketed models of  
278: \citet{lan03} that were rebinned to the {\it IUE} wavelength grid. 
279: We selected a solar abundance model with $T_{\rm eff}=50$~kK and  
280: $\log g = 4.5$ for the template (close to the optimal values; \S4). 
281: The resulting ccfs immediately revealed a narrow but  
282: weak signal that displayed a coherent Doppler shift  
283: when plotted against the orbital period from \citet{riv04} (see Fig.~1).  
284: We then made Gaussian fits of these ccfs to determine the radial  
285: velocity shifts for 50 of the 97 spectra where the ccf signal was  
286: clearly visible.   We used the method of \citet{zuc03} to estimate 
287: errors in the ccf velocities, and these range from 7 to 20 km~s$^{-1}$
288: depending on the S/N ratio of the spectrum.
289: 
290: \placefigure{fig1}     % Figure 1 - Subdwarf ccf plot  
291:  
292: We also measured ccf radial velocities for the Be star primary  
293: using two spectral regions where its photospheric lines are  
294: particularly strong (1423 -- 1523 and 1570 -- 1800 \AA ). 
295: In this case we used a template spectrum derived from the  
296: models of \citet{lan03} for $T_{\rm eff}=30$~kK and $\log g = 4.0$, 
297: first guesses based upon calibrations for the star's spectral  
298: classification.  The resulting ccfs are broad because of the  
299: star's large projected rotational velocity, and instead of  
300: fitting these with a Gaussian, we determined the radial velocity  
301: shifts relative to the average ccf. According to the method
302: of \citet{zuc03}, the errors in the ccf velocities for the Be star
303: are about 3~km~s$^{-1}$. We then added the Gaussian-fitted 
304: radial velocity of the average ccf ($V_r = 34.3\pm 0.3$ km~s$^{-1}$) 
305: to each of these relative velocities to end up with the primary  
306: star velocities listed in Table~2.  This table lists the heliocentric
307: Julian and UT dates of mid-exposure, the {\it IUE} image number 
308: for the spectrum, the orbital phase (see below), and
309: the radial velocities $V_1$ and $V_2$ for the Be star (subscript 1) and
310: the hot subdwarf (subscript 2), respectively.  Also given with the same
311: subscripts are the observed minus calculated residuals $(O-C)$ from the
312: orbital fits (see Table~3), and we see that in general the
313: deviations from the fit are comparable to the error estimates
314: and to the scatter between measurements of pairs of spectra obtained
315: close in time.   
316:  
317: \placetable{tab2}      % Table 2 - IUE ccf velocities 
318:  
319: We made a period search of the two sets of velocities using 
320: the discrete Fourier transform methods of \citet*{rob87}, and this  
321: analysis confirmed the period found by \citet{riv04}. 
322: We then used the non-linear, least-squares program of  
323: \citet{mor74} to determine independent orbital elements for  
324: the primary and secondary.  These produced consistent results  
325: for the orbital period and epoch, and we adopted the period and  
326: epoch from the larger amplitude secondary that had significantly 
327: lower errors.  We found that elliptical solutions did not yield  
328: any significant improvement in the residuals to the fits, so we 
329: assumed that the orbit is circular.  
330: Our final orbital elements are presented in column 3 of Table~3  
331: and the radial velocity curves are plotted in Figure~2.   
332: Table~3 lists period, $P$, epoch of the superior  
333: conjunction of the Be star, $T_{SC}$ (defining orbital phase 0.0),
334: semiamplitudes, $K$, and systemic velocities, $\gamma$, for
335: both components, mass ratio, $M_2/M_1$, mass--inclination functions,  
336: $M\sin ^3 i$, projected semimajor axis, $a\sin i$,
337: and r.m.s.\ residuals for each fit (where subscripts 1 and 2 
338: refer to the orbits of the Be and subdwarf star, respectively)  
339:  
340: \placetable{tab3}      % Table 3 - Orbital elements 
341:  
342: \placefigure{fig2}     % Figure 2 - Velocity curves 
343:  
344: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
345:  
346: \section{Tomographic Reconstruction of UV Spectra}  % Section 4 
347:  
348: We used the Doppler tomography algorithm described by  
349: \citet{bag94} to reconstruct the individual primary and  
350: secondary spectra from the composite {\it IUE} spectra.  
351: We took the radial velocity shifts for each component  
352: from the orbital solutions in Table~3, and then the reconstruction  
353: was run for 50 iterations with a gain of 0.8 (the results  
354: are insensitive to these values).  The flux ratio required for  
355: the reconstruction was determined from the secondary's line  
356: depths after making a comparison of the secondary's spectral properties 
357: with those in model spectra.  
358:  
359: We began our comparison of the reconstructed secondary spectrum  
360: and model spectra from \citet{lan03} by considering the  
361: appropriate line broadening.  We selected several strong and  
362: minimally blended metallic lines to compare with a grid of  
363: synthetic profiles formed by convolution with a rotational broadening  
364: function for a linear limb darkening law \citep{wad85} and  
365: a range in assumed projected rotational velocity $V\sin i$.  
366: The resulting mean of $V\sin i = 41 \pm 5$ km~s$^{-1}$ indicates  
367: that the secondary's lines are narrow and only barely resolved in the  
368: {\it IUE} spectra.  We then experimented with a range of model  
369: effective temperatures and gravities to try to reproduce  
370: the relative patterns of line strength (especially in the  
371: region surrounding the weak \ion{Si}{3} $\lambda 1299$ feature). 
372: We also compared the reconstructed and model profiles of  
373: the Stark broadened \ion{He}{2} $\lambda 1640$ feature to  
374: help estimate the appropriate value of gravity $\log g$.  
375: The parameters for the best matching solar abundance model  
376: spectrum are listed in column 3 of Table~4.   
377:  
378: \placetable{tab4}      % Table 4 - Physical properties 
379:  
380: We then estimated the mean UV flux ratio to be  
381: $F_2/F_1 = 0.04 \pm 0.01$ based upon a comparison of the  
382: line strengths in the best match model and in the reconstructed  
383: secondary spectrum for a range in test values of flux ratio.  
384: We show in Figures 3 and 4 the good agreement between the  
385: reconstructed secondary and model spectra for two regions  
386: of the UV spectrum that contain a number of prominent  
387: \ion{Fe}{5} and \ion{O}{5} photospheric lines.  
388:  
389: \placefigure{fig3}     % Figure 3 - Sec. reconstruction 
390:  
391: \placefigure{fig4}     % Figure 4 - Sec. reconstruction 
392:  
393: We also explored the range in synthetic spectral matches  
394: that were consistent with the appearance of the reconstructed 
395: Be star spectrum.  Once again we compared the profile widths 
396: of several conspicuous absorption lines with models for  
397: a range in projected rotational velocity, and we found  
398: a mean value of $V\sin i = 340 \pm 40$ km~s$^{-1}$.  
399: This estimate is similar to the result from \citet{fre05},
400: 375 km~s$^{-1}$, but it is larger than those reported  
401: in other optical studies (280 km~s$^{-1}$, \citealt{sle82}; 
402: 155 km~s$^{-1}$, \citealt*{abt02}).  We suspect that 
403: the optical lines may at times appear narrower  
404: because of unrecognized shell components formed in  
405: the Be star's disk.  We estimated the best match effective 
406: temperature based on the metallic line strengths  
407: (avoiding the use of narrow features like \ion{Si}{3} $\lambda 1299$ 
408: that are probably partially formed in the disk; \S6), 
409: and the best match model has a temperature consistent with 
410: the star's spectral classification (Table~4).  We simply 
411: assumed the gravity for the Be star based on its luminosity class 
412: since there are no easily made gravity estimates from the 
413: UV photospheric lines.  We note that both the temperature
414: and the gravity agree well with estimates made by \citet{pet76} 
415: based upon the continuum colors and the profiles of 
416: H$\gamma$ and H$\delta$.
417:  
418: FY~CMa is not a known eclipsing binary, so we cannot directly  
419: determine the system inclination and masses.  However, the appearance 
420: of shell lines in the UV and optical spectra suggests that the 
421: inclination is probably large, $i > 77^\circ$ \citep{han96}, if the  
422: disk resides in the orbital plane.  The typical mass for a B0.5~V star 
423: is $\approx 13.2 M_\odot$ \citep{har88}, and from the value of $M_1 \sin ^3 i$ 
424: in Table~3, this would imply that the inclination could be as small  
425: as $i=66^\circ$.  The range in probable masses is given in Table~4. 
426:  
427: We can estimate the radius of the Be star primary from its visual  
428: magnitude $V$ and its parallax $\pi$ as measured by {\it Hipparcos} \citep{esa97}.
429: Adopting $V=5.63$ \citep*{ste96}, $E(B-V)=0.14$ \citep{kai89}, 
430: $A_V=3.1 E(B-V)$, $\pi = 2.30\pm 0.70$ mas, and a bolometric correction  
431: of $BC=-2.65$ \citep{lan03}, then the radius from the luminosity  
432: and adopted temperature is $R_1 = 5.3 R_\odot$, where we  
433: have ignored the minor flux contributions from the disk and subdwarf.  
434: The radius of the subdwarf then follows from the observed and model  
435: UV flux ratios, $R_2 = 0.63 R_\odot$.  These mass and radius  
436: estimates yield gravities that are similar within errors to those  
437: from spectroscopic considerations.   
438:  
439: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
440:  
441: \section{Red Spectra Measurements}                  % Section 5 
442:  
443: The H$\alpha$ feature in our spectra always appeared as a 
444: strong emission line indicating that the Be disk was present  
445: throughout the period 1985 -- 2006.  We show in Figure~5 a montage  
446: of all the H$\alpha$ profiles arranged according to orbital phase.  
447: The profile is often double-peaked, usually in the sense that  
448: the red peak is stronger around phase $\phi=0.25$ while the  
449: blue peak is stronger near phase $\phi=0.75$.  Thus, the  
450: observations appear to confirm the suggestion by \citet{riv04}  
451: that the variations are orbitally modulated.  On the other hand,  
452: our spectra also show that there are temporal variations present 
453: unrelated to orbital phase that appear as alternating bright and 
454: dark horizontal segments in the gray-scale representation.  
455: We measured the H$\alpha$ equivalent widths by making a  
456: numerical integration of the line flux over the interval  
457: from 6528 to 6599 \AA ~(a region including weak emission  
458: components of \ion{C}{2} $\lambda 6578,6582$), and these 
459: are listed in Table~5. The typical equivalent width measurement 
460: error is 0.04 \AA . We find that there are significant variations 
461: in line strength over time scales of a week or so that appear to be 
462: unrelated to orbital phase and that probably reflect changes in the 
463: disk density distribution. 
464:  
465: \placefigure{fig5}     % Figure 5 - H-alpha gray-scale  
466:  
467: \placetable{tab5}      % Table 5 - Red spectra measurements 
468:  
469: The outer wings of H$\alpha$ show a hint of an ``S''-shaped  
470: velocity variation that is similar to the orbital velocity  
471: curve of the Be star.  Since the outer parts of the profile  
472: are formed in the disk gas closest to the star, the line wings may  
473: share in the orbital motion of the underlying star.  We measured  
474: the bisector position of the emission line wings near the  
475: 25\% maximum flux level using the method of \citet*{sha86},  
476: and these wing velocities are listed in column 4 of Table~5 
477: and are plotted together with the UV ccf velocities in Figure~2.
478: The typical error for the radial 
479: velocities from the H$\alpha$ wings is 1.0 km~s$^{-1}$.   
480: The H$\alpha$ wing velocities indeed follow the orbital  
481: motion of the Be star, and the orbital elements based  
482: upon them are in substantial agreement with those from  
483: the UV ccfs (see column 2 of Table 2).  The differences in  
484: systemic velocity $\gamma$ are not significant given the asymmetries 
485: and line blending present in the H$\alpha$ profile.  
486:  
487: We show the dramatic variations of the \ion{He}{1} $\lambda 6678$  
488: profile with orbital phase in Figure~6.  Here the emission displays  
489: a backwards ``S''-shaped velocity variation that is similar to
490: the secondary's orbital velocity curve.  The emission is  
491: strongest around orbital phase 0.0 when the hot subdwarf is in  
492: the foreground.  The emission sometimes assumes a pronounced  
493: double-peaked shape, especially when the emission is relatively  
494: strong.  In the same way as H$\alpha$, the \ion{He}{1} line also 
495: displays non-orbital related variations that give a banded  
496: appearance to the gray-scale representation of the variations.  
497: We also see the occasional development of a central  
498: shell absorption feature at phases prior to superior conjunction  
499: of the Be star. 
500:  
501: \placefigure{fig6}     % Figure 6 - He I 6678 gray-scale 
502:  
503: We made simple measurements of the peak velocity positions by  
504: locating the maximum flux positions in the \ion{He}{1} profile. 
505: If two peaks were present and clearly separated by a local  
506: minimum, then we measured both peaks and recorded their  
507: mean position $V_r$ and their separation $\triangle V_r$. 
508: If only one peak was visible, then only $V_r$ was measured.  
509: These measurements are given in the final two columns of  
510: Table~5 for all those cases where at least one strong peak  
511: was visible.  They are also plotted in Figure~7 together  
512: with the orbital velocity curve of the hot subdwarf.  
513: We see that the velocity amplitude of the \ion{He}{1} emission  
514: is actually larger than that of the subdwarf and that  
515: the episodes of greatest peak separation occur just prior  
516: to the orbital conjunctions.  We show in the next section 
517: that these characteristics are consistent with an emission  
518: origin in the outer part of the disk that faces the subdwarf.  
519:  
520: \placefigure{fig7}     % Figure 7 - He I 6678 peak velocities 
521:  
522: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
523:  
524: \section{Discussion}                                % Section 6 
525: 
526: Our spectroscopic results provide the ingredients to estimate
527: the basic elements of the system geometry.  Here we discuss 
528: the system dimensions and the evidence that the hot subdwarf 
529: is implicated in local heating of the Be star's disk in 
530: much the same way as was found for $\phi$~Per by \citet{ste00}
531: and \citet{hum01} and for 59~Cyg by \citet{mai05}. 
532: We present in Figure~8 a sketch of the inferred system geometry as 
533: viewed from above the orbital plane.  For the purpose of this
534: illustration, we assume an orbital inclination of $i=70^\circ$
535: and a semimajor axis of $112~R_\odot$ (where the acceptable 
536: range is probably $105 - 115 R_\odot$ for $i=90^\circ - 66^\circ$).
537: We also adopt a Be star radius of $R_1=5.3 R_\odot$ 
538: (where the probable range is $4.1 - 7.6 R_\odot$ for distances 
539: of $333 - 625$~pc). The Be star is represented by the solid disk
540: on the left while the hot subdwarf is shown by the small dot 
541: on the right.  The double-lobe surrounding them shows the 
542: much larger Roche surface boundary.   We can estimate the half light
543: radius of the H$\alpha$ emitting disk using the method described by 
544: \citet{gru06}.  This estimate is derived from a simple density power
545: law for the disk, the mean equivalent width ($W_\lambda = -13.6$ \AA ),
546: the adopted Be star temperature and disk inclination, and an assumed
547: disk outer boundary equal to the volume-equivalent, Roche radius.
548: These parameters lead to an estimated disk to star radius ratio of 
549: $R_d/R_1=9.4$, or a half-light emitting radius of $50 R_\odot$ for 
550: the adopted radius of the Be star.  
551: We caution that the simplicity of the models
552: may lead to errors in the half-light radius of $\pm 30\%$ and that the models 
553: assume an azimuthal symmetry for the emission flux, which is probably not the case 
554: for the disk emission of FY~CMa (see below).  Nevertheless, this simple estimate
555: of disk radius is indicated by the lighter gray shaded region on the left side of  
556: Figure~8.  \citet{oka01} argue that the disks of Be stars in binaries 
557: with nearly circular orbits are truncated at the 3:1 orbital resonance radius,
558: and this occurs at a distance from the Be star of $52 R_\odot$ that is only 
559: slightly larger than our disk half-light radius estimate. 
560:  
561: \placefigure{fig8}     % Figure 8 - Cartoon of system geometry 
562: 
563: We showed in \S5 that the \ion{He}{1} $\lambda 6678$ emission has 
564: an orbital modulation that leads to greatest emission strength 
565: when the hot subdwarf is in the foreground.  Furthermore we found that the 
566: semiamplitude of the orbital Doppler shift of the emission peaks 
567: was larger than that of the subdwarf (from Fig.~7, $K_{em}\approx 173$
568: km~s$^{-1}$ $> K_2 = 128$ km~s$^{-1}$).  These facts suggest that a 
569: significant fraction of the emission forms in the part of the disk near the 
570: subdwarf where the Keplerian velocity of the disk gas is larger than the 
571: orbital velocity of the subdwarf. Ignoring three-body effects, we can 
572: estimate the radius of the emission source by assuming both
573: the disk gas and the subdwarf obey Keplerian motion around the Be star, 
574: $R_{em}/a = (M_1/(M_1+M_2)) ((K_1+K_2)/(K_1+K_{em}))^2 = 0.51$ 
575: or $R_{em}=58 R_\odot$.
576: This radius is about $15\%$ larger than the half-light radius,
577: which suggests that the emission forms mainly along the outer rim 
578: of a disk that may be extended towards the hot subdwarf (see Fig.~8).  
579: We suspect that this outer disk region is heated by the combined 
580: effects of the incident subdwarf radiation and the impact of the
581: subdwarf's wind on the disk.  We imagine that the rest of the disk also 
582: contributes to the line emission flux, and, for example, those times 
583: when weak emission is observed (see Fig.~7) may correspond to episodes 
584: when the gas density declines in the outer rim facing the subdwarf 
585: and the flux from the rest of the disk dominates.
586: 
587: Since the heated disk gas nearest the hot subdwarf will 
588: have a higher Keplerian velocity, the center of the heated region 
589: will move ahead of the axis joining the stars by an amount that 
590: depends on the cooling and orbital timescales.  We can estimate 
591: the phase offset $\triangle \phi$ of this asymmetry in two
592: ways.  First, the central heated region should reach inferior
593: conjunction about midway between the observed velocity extrema 
594: at phases $\phi=0.77$ and 0.17 (see Fig.~7), i.e., at $\phi=0.96$.
595: Second, the velocity peak separation is presumably caused by 
596: the heated region forming in a sector of a ring.  We will observe
597: the largest velocity separations when the mid-point of the ring segment 
598: is oriented along our line of sight.  According to the peak separations 
599: plotted in Figure~7, greatest peak separations occur around 
600: $\phi=0.42$ and $\phi=0.92$.   This suggests that the apex of the
601: heated rim is offset by about 0.06 in phase ($22^\circ$) ahead of the 
602: subdwarf inferior junction at $\phi=0.0$.  From simple geometric 
603: arguments, we expect that the maximum half-peak separation 
604: $V_m = {\rm max}(\triangle V_r /2)$ is related to the emission 
605: semiamplitude by $V_m=K_{em} \sin \theta$, where $\theta$ is 
606: half the ring sector opening angle.  From the values of 
607: peak separation in Table~5, we estimate $V_m=66$ km~s$^{-1}$ and 
608: $\theta = 22^\circ$.  These values were used to plot a 
609: representative heated region in the darker gray shaded area 
610: of Figure~8.  The actual extent of the heated ring is probably 
611: larger than shown since the separated emission peaks actually
612: sample a range of ring azimuthal angles and not just the termination
613: angles. 
614:  
615: The subdwarf may also affect the kinematics of the
616: gas motion in the disk rim.  We show in Figure~9 the orbital 
617: phase variations in the appearance of the \ion{Si}{3} lines 
618: near \ion{Si}{3} $\lambda 1299$.  The cores of these
619: lines are shell-type features formed in the disk gas seen in projection 
620: against the photosphere of the Be star.  We see evidence that 
621: the shell feature broadens from $\phi=0.5$ to $\phi=0.0$ and 
622: apparently shifts from a redshifted to a blueshifted position
623: at phases surrounding conjunction at $\phi=0.0$. 
624: We imagine that the motion of the outer disk gas is influenced by 
625: subdwarf as the disk gas passes and overtakes the star.  
626: The  gravitational attraction would cause acceleration and outward motion 
627: as the disk gas approaches the subdwarf, but the same forces would cause 
628: deceleration and inward motion following closest passage.
629: Thus, we would observe infalling (redshifted) disk gas before $\phi=0.0$ 
630: and then outward moving (blueshifted) gas after $\phi=0.0$. 
631: On the other hand, the disk gas projected against the Be star has
632: more nearly Keplerian and tangential motion at $\phi=0.5$ leading to  
633: the appearance of narrow shell lines at that phase. 
634: This kind of perturbation in the Keplerian motions may also partially
635: explain the appearance of a redshifted shell feature in the 
636: \ion{He}{1} $\lambda6678$ profiles near $\phi=0.9$ (Fig.~6), 
637: an orientation where the disk gas projected against the Be star  
638: would be infalling and relatively dense. 
639: 
640: \placefigure{fig9}     % Figure 9 - Si III 1300 variations
641: 
642: 
643: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
644: 
645: In addition to the shell features discussed above that predominantly are 
646: formed in the disk, another circumstellar component is seen in some  
647: intermediately-ionized species (e.g. \ion{S}{3}, \ion{Fe}{3}). 
648: These variable features, which rarely show velocities 
649: less than the photospheric value and can display velocities as large as
650: 50~$\rm km~s^{-1}$ relative to the photosphere of the Be star, were discovered 
651: in {\it IUE} images obtained in the 1980s \citep*{pet88,gra88} and they clearly 
652: show the presence of significant infall of material toward the Be star. Often 
653: times these lines are asymmetric with enhanced absorption on the red side 
654: of the profile. \citet*{gra88} identified six infall episodes that are best 
655: seen in the \ion{Fe}{3} (multiplet~34) line at 1895.456~\AA, since the feature 
656: does not have to compete with a strong photospheric component and it is not 
657: blended with interstellar or other circumstellar lines. 
658: 
659: From the entire set of 97 {\it IUE} SWP high resolution (HIRES) 
660: images that span more than 15 years of observation and our 
661: knowledge of the orbit of the system, we now have more insight on the
662: nature of the infall lines. Their behavior in strength and velocity are shown in
663: Figures~10, 11, and 12. Representative examples of the three basic types of 
664: observed features, photospheric, photospheric with additional absorption from the 
665: Be star's disk, and infall are given in Figure~10. In order to eliminate 
666: phase-dependent effects, we have chosen spectra taken at approximately the same 
667: quadrature phase on the {\it trailing} hemisphere of the Be star but at different 
668: epochs. Note that the central core of the profile taken during a period of enhanced 
669: disk absorption shows the same velocity as the photospheric line, but that the 
670: feature formed in infalling material has a significant redshift.
671: In the upper two-part panel of Figure~11 we plot the residual intensity\footnote{The
672: observed flux divided by the flux in the local continuum outside of the spectral 
673: feature.} of the 
674: deepest part of the \ion{Fe}{3} line and its radial velocity versus observation date.
675: All data are included in the intensity plot, but in the velocity plot we use a 
676: different symbol for observations that show no, or minimal shell absorption. Points 
677: that fall above the {\it dashed} horizontal line in the upper panel visually appeared 
678: to be mostly photospheric. The double horizontal {\it dashed lines}  in the second 
679: panel delineate the domain of photospheric motion. Features that fall above the 
680: double line are clearly associated with infall phases. Seven or eight clear infall 
681: episodes are observed. The data suggest that the duration of the infall phases varies 
682: from days to a year, but significant data gaps from 1979-85 and 1990-94 render it 
683: impossible to determine a firm value or its degree of variability. An apparently 
684: extended infall phase of about 300 days was seen from HJD~2446919--2447215. In 
685: the lower two panels of Figure~11, the same data are plotted versus orbital phase.
686: From the third panel it can be seen that on the average the shell-type absorption is 
687: enhanced in the phase interval 0.3--0.7. But from the display in the fourth panel it 
688: is apparent that features seen around phase 0.5 tend to display the systemic velocity, 
689: implying they are formed in the disk of the Be star.
690: In the fourth panel core velocities from only the spectra 
691: showing enhanced shell absorption are compared with the radial velocity curve of  
692: the Be star. It is clear that infall features with velocities up 
693: to 50~$\rm km~s^{-1}$ relative to the photosphere are seen {\it at most phases}
694: with a minimal occurrence around phase 0.5. The infall features tend to cluster 
695: around the quadrature points at which their velocities are highest relative to the
696: Be star's photosphere. Many of the phase dependent strength and velocity variations 
697: discussed above can also be seen in the gray-scale plot presented in Figure~12. 
698: 
699: The episodic infall of material is fundamentally different from that observed in 
700: the Algol-type binaries \citep{pet01}, where mass motion toward a mass gainer 
701: is typically seen only between phases 0.75--0.95. In the case of FY~CMa, the infall
702: features predominantly cluster around the quadrature phases from 0.15--0.35 and 
703: 0.65--0.90. The velocities of the shell features around phase 0.5 tend to be either 
704: photospheric or negative relative to the Be star\footnote{Evidence for material
705: outflow is often seen in interacting binaries of the Algol type \citep{pet01} and 
706: similar hydrodynamics might be occuring in this Be + subdwarf system.}. \citet{pet88} 
707: discussed various scenarios to explain the observed simultaneous presence of infall 
708: and outflow features in this system including a sudden accretion event, 
709: the possible presence of magnetic loops, or flow from a polar jet. 
710: \citet{cao01} reported that infall components in the \ion{N}{5} doublet 
711: were contemporaneously observed with those in \ion{He}{1} $\lambda5876$ and also
712: suggested that magnetic loops might be involved, but the phase-dependence of the infall
713: features does not support this speculation. The infall is more likely a consequence of an 
714: interaction between the wind from the subdwarf and the periphery of the massive disk 
715: of the Be star or disk perturbations due to the tidal force produced by the subdwarf.
716: The motion of a shock-interaction region would result from variations in the location of 
717: the interface where there is a balance between the
718: dynamic pressure of the subdwarf's wind and the gas pressure at the periphery 
719: of the Be star's disk.  It is not clear at this time whether it is the subdwarf or
720: the Be star that might be driving the variability though. Epochs when no enhanced shell 
721: absorption is observed would correspond to minimal wind-disk interaction.
722: Alternatively the episodic infall features could simply be 
723: a result of gas dynamics in the disk due to a tidal effect from the subdwarf. 
724: Hydrodynamical simulations\footnote{cf. 
725: http://harmas.arc.hokkai-s-u.ac.jp/\%7Eokazaki/BeX/sim/index.html} of the 
726: gas motion in a perturbed disk in a Be/X-Ray binary
727: \citep{oka02} reveal turbulence in the outer disk and even the formation of 
728: spiral structure. Similar disk variability could be operating in FY~CMa.
729: The details are probably complex but an interplay of such gravitationally related
730: forces could create time variable and sometimes coherent gas flows.
731: For example, the tidal deformation of the disk leads to an axial elongation 
732: that moves ahead of the motion binary.  Disk gas returning from the
733: extended regions would be observed to be moving inwards when
734: viewed at the quadrature phases.
735: 
736: 
737: 
738: 
739: \placefigure{fig10}     % Figure 10 - Selected Fe III profiles
740: \placefigure{fig11}     % Figure 11 - Fe III core strength and velocity vs. HJD and phase 
741: \placefigure{fig12}     % Figure 12 - Fe III 1895 variations - gray-scale
742: 
743: 
744: 
745: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
746:  
747:  
748:  
749:  
750: Finally, we return to the cyclic variation in the \ion{N}{5} $\lambda1240$ 
751: wind line that first led \citet{riv04} to find the orbital period
752: of the binary.  The plot of these variations in Figure~13 shows that 
753: the greatest extent of blueshifted absorption occurs around $\phi=0.0$
754: when the subdwarf is in the foreground.  We note that such a strong \ion{N}{5} 
755: wind feature is quite unusual for Be stars \citep*{gra87}, and the orbital 
756: phasing suggests instead that the feature may form through absorption of 
757: the Be star's flux by the wind of the hot subdwarf when it is in the foreground. 
758: The profile would mainly be affected by blueshifted wind absorption since 
759: any redshifted absorption component would probably by canceled by redshifted emission
760: in the subdwarf's wind.  This kind of orbital wind variation is not seen in the 
761: other major wind line, the \ion{C}{4} $\lambda 1550$ doublet, and this may suggest  
762: that the subdwarf is N-enriched and C-depleted as expected for a star that has 
763: been stripped down to near the CNO-burning core.  One problem with this 
764: explanation is that the largest wind speeds observed, $\approx 400$ km~s$^{-1}$,
765: are much smaller than the wind terminal velocity of a few thousand km~s$^{-1}$
766: expected for a radiatively driven wind from the subdwarf \citep{lam99}. However, 
767: it may be that the wind optical depths are too small for absorption to occur at 
768: higher outflow velocities.  
769:  
770: \placefigure{fig13}     % Figure 13 - N V 1240 orbital variations 
771:  
772: In 1995 March a sequence of daily {\it IUE} SWP HIRES spectra of FY~CMa 
773: were obtained over the course of 16 days.  Fortuitously these spectra 
774: were centered in time around orbital phase 0.0 and extended from  
775: phases 0.784 to 0.213.  The temporal variations in the \ion{N}{5} $\lambda1240$ 
776: doublet are illustrated in a gray-scale diagram in Figure~14. More detail can be seen
777: from the plot of selected observations shown in Figure~15.  The \ion{N}{5} 
778: features initially showed very little structure until a core with the velocity of
779: the center-of-mass of the system appeared around phase 0.97.  The core, still
780: with no velocity shift, reached its greatest strength at conjunction. 
781: It broadened by phase 0.08, and had completely vanished by phase 0.21.
782: The core is probably formed in a shock-heated region between the stars 
783: where the N-enhanced wind from the subdwarf collides with the disk of 
784: the Be star. 
785: 
786: 
787: 
788: \placefigure{fig14}     % Figure 14 - N V doublet - selected profiles
789: \placefigure{fig15}     % Figure 15 - N V 1240 temporal variations in 1995 March
790: 
791: 
792: 
793: Our work demonstrates that FY~CMa represents the third known case of 
794: a Be star with a hot subdwarf companion, the remains of an intense 
795: binary interaction that stripped down the mass donor to a tiny fraction 
796: of its original mass and left the mass gainer spinning rapidly. 
797: The subdwarf secondary in FY~CMa has a mass that is close to the 
798: Chandrasekhar limit, and lacking a significant H-envelope, it may represent 
799: the progenitor of a supernova type SN~Ibc.  The system is now wide enough 
800: that it is unlikely that the Be star will be substantially spun down by tidal 
801: effects, so its rapid rotation will be important throughout its subsequent evolution. 
802: Such rapidly rotating massive stars may be related to the progenitors of   
803: collapsars and gamma-ray bursters \citep{can07}.  
804:  
805: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
806:  
807: \acknowledgments 
808:  
809: We thank Daryl Willmarth and the staff of KPNO for their assistance  
810: in making these observations possible and  
811: Diana Gudkova for her assistance in organizing the optical spectra.
812: The {\it IUE} data presented in this paper were obtained from  
813: the Multimission Archive at the Space Telescope Science  
814: Institute (MAST). STScI is operated by the Association of  
815: Universities for Research in Astronomy, Inc., under NASA  
816: contract NAS5-26555. Support for MAST for non-HST data is  
817: provided by the NASA Office of Space Science via grant  
818: NAG5-7584 and by other grants and contracts. 
819: Many researchers contributed to the 97 {\it IUE} SWP HIRES images
820: used in this study, including programs by K. S. Bjorkman,
821: C. A. Grady, H. Henrichs,  J. Nichols, G. J. Peters, and C.-C. Wu. 
822: We especially thank Carol Grady, who had the foresight to obtain the 
823: unique set of daily {\it IUE} spectra in 1995, more than a decade
824: before analysis software allowed us to suggest an interpretation for the
825: striking short-term variability seen in the \ion{N}{5} doublet and other
826: wind features. We also thank an anonymous referee whose suggestions led to 
827: several improvements in the paper. This work was supported in part by the 
828: National Science Foundation under grant AST-0606861. 
829: Institutional support has been provided from the USC 
830: Women in Science \& Engineering (WiSE) program and from the GSU College 
831: of Arts and Sciences and the Research Program Enhancement 
832: fund of the Board of Regents of the University System of Georgia, 
833: administered through the GSU Office of the Vice President 
834: for Research.  
835: 
836: {\it Facilities:} \facility{IUE }, \facility{KPNO:CFT } 
837:  
838: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
839:  
840: % References 
841:  
842: \clearpage 
843:  
844: \begin{thebibliography}{} 
845: \bibitem[Abt et al.(2002)Abt, Levato, \& Grosso]{abt02} 
846:          Abt, H. A., Levato, H., \& Grosso, M. 2002, \apj, 573, 359 
847: \bibitem[Bagnuolo et al.(1994)]{bag94} 
848:          Bagnuolo, W. G., Jr., Gies, D. R., 
849:          Hahula, M. E., Wiemker, R., \& 
850:          Wiggs, M. S. 1994, \apj, 423, 446 
851: \bibitem[Cantiello et al.(2007)]{can07} 
852:          Cantiello, M., Yoon, S.-C., Langer, N., \& Livio, M. 
853:          2007, \aap, 465, L29 
854: \bibitem[Cao(2001)]{cao01} 
855:          Cao, H.-L. 2001, Chinese J. Astron. Astrophys., 1, 514 
856: \bibitem[Coe(2000)]{coe00} 
857:          Coe, M. 2000, in The Be Phenomenon in Early-Type Stars,  
858:          IAU Coll. 175 (ASP Conf. Proc. 214), ed. M. A. Smith, 
859:          H. F. Henrichs, \& J. Fabregat (San Francisco: ASP), 656 
860: \bibitem[Dray \& Tout(2007)]{dra07} 
861:          Dray, L. M., \& Tout, C. A. 2007, \mnras, 376, 61
862: \bibitem[ESA(1997)]{esa97}
863:          ESA 1997, The Hipparcos Catalogue, ESA SP-1200 (Noordwijk: ESA)
864: \bibitem[Fr\'{e}mat et al.(2005)]{fre05}
865:          Fr\'{e}mat, Y., Zorec, J., Hubert, A.-M., \& Floquet, M. 
866:          2005, \aap, 440, 305  
867: \bibitem[Gies(2000)]{gie00} 
868:          Gies, D. R. 2000, in The Be Phenomenon in Early-Type Stars,  
869:          IAU Coll. 175 (ASP Conf. Proc. 214), ed. M. A. Smith, 
870:          H. F. Henrichs, \& J. Fabregat (San Francisco: ASP), 668 
871: \bibitem[Gies et al.(1998)]{gie98} 
872:          Gies, D. R., Bagnuolo, W. G., Jr., Ferrara, E. C.,  
873:          Kaye, A. B., Thaller, M. L., Penny, L. R., \& Peters, G. J. 
874:          1998, \apj, 493, 440 
875: \bibitem[Grady et al.(1987)Grady, Bjorkman, \& Snow]{gra87} 
876:          Grady, C. A., Bjorkman, K. S., \& Snow, T. P. 
877:          1987, \apj, 320, 376 
878: \bibitem[Grady et al.(1988)]{gra88} 
879:          Grady, C. A., Peters, G. J., Bjorkman, K. S., \&  
880:          Henrichs, H. F. 1988, in A Decade of UV Astronomy with the  
881:          IUE Satellite (ESA SP-281) (Paris: ESA), 1, 257 
882: \bibitem[Grundstrom \& Gies(2006)]{gru06} 
883:          Grundstrom, E. D., \& Gies, D. R. 2006, \apj, 651, L53  
884: \bibitem[Hanuschik(1996)]{han96} 
885:          Hanuschik, R. W. 1996, \aap, 308, 170 
886: \bibitem[Harmanec(1988)]{har88} 
887:          Harmanec, P. 1988, Bull. Astron. Inst. Cz., 39, no. 6, 329 
888: \bibitem[Hiltner et al.(1969)Hiltner, Garrison, \& Schild]{hil69} 
889:          Hiltner, W. A., Garrison, R. F., \& Schild, R. E. 1969, 
890:          \apj, 157, 313 
891: \bibitem[Hummel \& \v{S}tefl(2001)]{hum01} 
892:          Hummel, W., \& \v{S}tefl, S. 2001, \aap, 368, 471
893: \bibitem[Kaiser(1989)]{kai89}
894:          Kaiser, D. 1989, \aap, 222, 187 
895: \bibitem[Lamers \& Cassinelli(1999)]{lam99}
896:          Lamers, H. J. G. L. M., \& Cassinelli, J. P. 1999, 
897:          Introduction to Stellar Winds (Cambridge: Cambridge Univ. Press) 
898: \bibitem[Lanz \& Hubeny(2003)]{lan03} 
899:          Lanz,T., \& Hubeny, I. 2003, \apjs, 146, 417 
900: \bibitem[Maintz et al.(2005)]{mai05} 
901:          Maintz, M., Rivinius, T., Stahl, O., \v{S}tefl, S., \& 
902:          Appenzeller, I. 2005, Publ. Astron. Inst. Cz., 93, 21 
903: \bibitem[Morbey \& Brosterhus(1974)]{mor74} 
904:          Morbey, C. L., \& Brosterhus, E. B. 1974, \pasp, 86, 455 
905: \bibitem[Okazaki et al. (2002)]{oka02} 
906:          Okazaki, A. T., Bate, M. R., Ogilvie, G. I., \& Pringle, J. E. 2002, 
907:          \mnras, 337, 967
908: \bibitem[Okazaki \& Negueruela(2001)]{oka01} 
909:          Okazaki, A. T., \& Negueruela, I. 2001, \aap, 377, 161
910: \bibitem[Peters(1976)]{pet76} 
911:          Peters, G. J. 1976, in Be and Shell Stars, IAU Symp. 70, ed.
912:          A. Slettebak (Dordrecht: Reidel), 69 
913: \bibitem[Peters(1982)]{pet82} 
914:          Peters, G. J. 1982, in Be stars, IAU Symp. 98, ed.
915:          M. Jaschek \& H.-G. Groth (Dordrecht: Reidel), 411 
916: \bibitem[Peters(1983)]{pet83} 
917:          Peters, G. J. 1983, \pasp, 95, 311 
918: \bibitem[Peters(1988)]{pet88} 
919:          Peters, G. J. 1988, \apj, 331, L33 
920: \bibitem[Peters (2001)]{pet01} 
921:          Peters, G. J. 2001, in The Influence of Binaries on Stellar Population 
922:         Studies, ed. D. Vanbeveren (Dordrecht: Kluwer), 79
923:  \bibitem[Petrovic et al.(2005)Petrovic, Langer, \& van der Hucht]{pet05} 
924:          Petrovic, J., Langer, N., \& van der Hucht, K. A. 
925:          2005, \aap, 435, 1013 
926: \bibitem[Pickering \& Fleming(1905)]{pic05}
927:          Pickering, E. C., \& Fleming, W. P. 1905, \apj, 22, 87
928: \bibitem[Pols et al.(1991)]{pol91} 
929:          Pols, O. R., Cot\'{e}, J., Waters, L. B. F. M., \& 
930:          Heise, J. 1991, \aap, 241, 419 
931: \bibitem[Porter \& Rivinius(2003)]{por03} 
932:          Porter, J. M., \& Rivinius, T. 2003, \pasp, 115, 1153 
933: \bibitem[Rivinius et al.(2004)]{riv04} 
934:          Rivinius, T., \v{S}tefl, S., Maintz, M., Stahl, O., \& 
935:          Baade, D. 2004, \aap, 427, 307 
936: \bibitem[Roberts et al.(1987)Roberts, Leh\'{a}r, \& Dreher]{rob87} 
937:          Roberts, D. H., Leh\'{a}r, J., \& Dreher, J. W. 1987, \aj, 93, 968 
938: \bibitem[Shafter et al.(1986)Shafter, Szkody, \& Thorstensen]{sha86}
939:          Shafter, A. W., Szkody, P., \& Thorstensen, J. R. 1986, \apj, 308, 765
940: \bibitem[Slettebak(1982)Slettebak]{sle82} 
941:          Slettebak, A. 1982, \apjs, 50, 55 
942: \bibitem[\v{S}tefl et al.(2000)\v{S}tefl, Hummel, \& Rivinius]{ste00} 
943:          \v{S}tefl, S., Hummel, W., \& Rivinius, T. 2000, \aap, 358, 208 
944: \bibitem[Sterken et al.(1996)Sterken, Vogt, \& Mennickent]{ste96}
945:          Sterken, C., Vogt, N., \& Mennickent, R. E. 1996, \aap, 311, 579
946: \bibitem[Thaller et al.(1995)]{tha95} 
947:          Thaller, M. L., Bagnuolo, W. G., Jr., Gies, D. R., \& 
948:          Penny, L. R. 1995, \apj, 448, 878 
949: \bibitem[van Bever \& Vanbeveren(1997)]{van97} 
950:          van Bever, J., \& Vanbeveren, D. 1997, \aap, 322, 116
951: \bibitem[Wade \& Rucinski(1985)]{wad85}
952:          Wade, R. A., \& Rucinski, S. M. 1985, \aaps, 60, 471 
953: \bibitem[Waters et al.(1991) Waters, Cot\'{e}, \& Pols]{wat91} 
954:          Waters, L. B. F. M., Cot\'{e}, J., \& Pols, O. R. 
955:          1991, \aap, 250, 437 
956: \bibitem[Wellstein et al.(2001)Wellstein, Langer, \& Braun]{wel01} 
957:          Wellstein, S., Langer, N., \& Braun, H. 2001, \aap, 369, 939
958: \bibitem[Zucker(2003)]{zuc03}
959:          Zucker, S. 2003, \mnras, 342, 1291 
960: \end{thebibliography} 
961: 
962: 
963: \clearpage
964: 
965: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
966: 
967:  
968: % Tables  
969:  
970: % Table 1 
971: \begin{deluxetable}{ccccccl} 
972: \tabletypesize{\scriptsize} 
973: \tablewidth{0pc} 
974: \tablecaption{Journal of Spectroscopy \label{tab1}} 
975: \tablehead{ 
976: \colhead{Dates} & 
977: \colhead{Besselian} & 
978: \colhead{Range} & 
979: \colhead{Resolving Power} & 
980: \colhead{Number of} & 
981: \colhead{Telescope/} \\ 
982: \colhead{(HJD-2,400,000)} & 
983: \colhead{Year} & 
984: \colhead{(\AA)} & 
985: \colhead{($\lambda/\triangle\lambda$)} & 
986: \colhead{Spectra} & 
987: \colhead{Grating/Detector}} 
988: \startdata 
989: 44170 -- 49804   &  1979.8 -- 1995.2 &   1200 -- 1900    & 10000    &   97 & {\it IUE}/Echelle/SWP  \\
990: 46123            &  1985.2           &   6477 -- 6654    & 13000    &\phn1 &  KPNO CF/B/TI3  \\
991: 46153            &  1985.2           &   6514 -- 6691    & 13000    &\phn1 &  KPNO CF/B/TI3  \\
992: 46735            &  1986.8           &   6511 -- 6688    & 13000    &\phn1 &  KPNO CF/B/TI3  \\
993: 46903 -- 46906   &  1987.3 -- 1987.3 &   6522 -- 6698    & 13000    &\phn3 &  KPNO CF/B/TI3  \\
994: 46918 -- 46922   &  1987.3 -- 1987.3 &   6520 -- 6698    & 13000    &\phn6 &  KPNO CF/B/TI3  \\
995: 47033 -- 47034   &  1987.6 -- 1987.7 &   6518 -- 6695    & 13000    &\phn2 &  KPNO CF/B/TI3  \\
996: 47303 -- 47307   &  1988.4 -- 1988.4 &   6517 -- 6694    & 13000    &\phn9 &  KPNO CF/B/TI3  \\
997: 47470 -- 47473   &  1988.8 -- 1988.9 &   6495 -- 6717    & 11000    &   11 &  KPNO CF/B/RCA2  \\
998: 47561 -- 47562   &  1989.1 -- 1989.1 &   6514 -- 6690    & 14000    &\phn3 &  KPNO CF/B/TI3  \\
999: 47637 -- 47641   &  1989.3 -- 1989.3 &   6520 -- 6697    & 14000    &\phn4 &  KPNO CF/B/TI3  \\
1000: 47940 -- 47941   &  1990.1 -- 1990.1 &   6523 -- 6699    & 14000    &\phn2 &  KPNO CF/B/TI3  \\
1001: 47982 -- 47984   &  1990.2 -- 1990.3 &   6520 -- 6697    & 14000    &\phn2 &  KPNO CF/B/TI3  \\
1002: 48314 -- 48315   &  1991.2 -- 1991.2 &   6542 -- 6690    & 18000    &\phn3 &  KPNO CF/A/ST1K \\
1003: 48319 -- 48322   &  1991.2 -- 1991.2 &   6520 -- 6696    & 14000    &\phn4 &  KPNO CF/B/TI3  \\
1004: 48515 -- 48517   &  1991.7 -- 1991.7 &   6517 -- 6693    & 14000    &\phn5 &  KPNO CF/B/TI3  \\
1005: 49058 -- 49061   &  1993.2 -- 1993.2 &   6384 -- 6723    & 19000    &\phn3 &  KPNO CF/A/T2KB \\
1006: 49444 -- 49448   &  1994.2 -- 1994.3 &   6525 -- 6693    & 18000    &\phn4 &  KPNO CF/A/T1KA \\
1007: 51123 -- 51126   &  1998.8 -- 1998.9 &   6359 -- 6677    & 20000    &\phn4 &  KPNO CF/A/F3KB \\
1008: 51192 -- 51197   &  1999.0 -- 1999.0 &   6420 -- 6737    & 20000    &\phn5 &  KPNO CF/A/F3KB \\
1009: 51502 -- 51512   &  1999.9 -- 1999.9 &   6415 -- 6733    & 20000    &   16 &  KPNO CF/A/F3KB \\
1010: 51613 -- 51614   &  2000.2 -- 2000.2 &   6459 -- 6776    & 20000    &\phn8 &  KPNO CF/A/F3KB \\
1011: 51615 -- 51618   &  2000.2 -- 2000.2 &   6413 -- 6730    & 20000    &\phn9 &  KPNO CF/A/F3KB \\
1012: 51818 -- 51831   &  2000.7 -- 2000.8 &   6440 -- 7105    &\phn9500  &\phn9 &  KPNO CF/B/F3KB \\
1013: 51850 -- 51852   &  2000.8 -- 2000.8 &   6413 -- 6733    & 20000    &\phn7 &  KPNO CF/A/F3KB \\
1014: 51890 -- 51902   &  2000.9 -- 2001.0 &   6440 -- 7105    &\phn9500  &   12 &  KPNO CF/B/F3KB \\
1015: 51914 -- 51918   &  2001.0 -- 2001.0 &   6432 -- 6749    & 20000    &   18 &  KPNO CF/A/F3KB \\
1016: 53292 -- 53295   &  2004.8 -- 2004.8 &   6470 -- 7140    &\phn9500  &\phn4 &  KPNO CF/B/T2KB \\
1017: 54021 -- 54025   &  2006.8 -- 2006.8 &   6473 -- 7143    &\phn9500  &\phn2 &  KPNO CF/B/T2KB \\
1018: \enddata 
1019: \end{deluxetable} 
1020: \newpage 
1021: 
1022: 
1023: 
1024: % Table 2 
1025: \begin{deluxetable}{lccccccc} 
1026: \tabletypesize{\scriptsize} 
1027: \tablewidth{0pc} 
1028: \tablenum{2} 
1029: \tablecaption{{\it IUE} Radial Velocity Measurements\label{tab2}} 
1030: \tablehead{ 
1031: \colhead{Date}             & 
1032: \colhead{UT}               & 
1033: \colhead{SWP}              & 
1034: \colhead{Orbital}          & 
1035: \colhead{$V_1$}            & 
1036: \colhead{$(O-C)_1$}        & 
1037: \colhead{$V_2$}            & 
1038: \colhead{$(O-C)_2$}        \\  
1039: \colhead{(HJD--2,400,000)} & 
1040: \colhead{(yyyy-mm-dd)}    & 
1041: \colhead{Number}           & 
1042: \colhead{Phase}            & 
1043: \colhead{(km s$^{-1}$)}    & 
1044: \colhead{(km s$^{-1}$)}    & 
1045: \colhead{(km s$^{-1}$)}    & 
1046: \colhead{(km s$^{-1}$)}    } 
1047: \startdata 
1048:  44170.044 \dotfill & 1979-10-23 &  \phn6963 &  0.985 &  
1049: \phn\phs     $  51.5$ &\phs     $  16.6$ & 
1050: \phn\phn\phs $   1.6$ &         $ -17.9$ \\ 
1051:  44330.226 \dotfill & 1980-03-31 &  \phn8617 &  0.285 &  
1052: \phn\phs     $  23.9$ &\phn\phs $   4.4$ & 
1053: \phs         $ 159.8$ &\phn\phs $   3.4$ \\ 
1054:  44853.156 \dotfill & 1981-09-05 & 14910 &  0.321 &  
1055: \phn\phs     $  22.2$ &\phn\phs $   1.6$ & 
1056: \phs         $ 138.6$ &\phn     $  -8.4$ \\ 
1057:  44867.935 \dotfill & 1981-09-20 & 15053 &  0.717 &  
1058: \phn\phs     $  40.5$ &\phn     $  -7.3$ & 
1059:              $-123.5$ &         $ -29.2$ \\ 
1060:  44920.747 \dotfill & 1981-11-12 & 15478 &  0.135 &  
1061: \phn\phs     $  39.0$ &\phs     $  16.2$ & 
1062:              \nodata  &         \nodata  \\ 
1063: \enddata 
1064: \tablecomments{A machine readable version of the full Table 2 is available 
1065: in the electronic edition of the {\it Astrophysical Journal.}  A portion 
1066: is shown here for guidance regarding its form and content.}
1067: \end{deluxetable}
1068:  
1069: \newpage 
1070: 
1071: 
1072:  
1073: % Table 3 
1074: \begin{deluxetable}{lcc} 
1075: %\tabletypesize{\scriptsize} 
1076: \tablewidth{0pc} 
1077: \tablenum{3} 
1078: \tablecaption{Orbital Elements for FY CMa\label{tab3}} 
1079: \tablehead{ 
1080: \colhead{Element} & 
1081: \colhead{H$\alpha$ Wings} & 
1082: \colhead{{\it IUE} ccfs}} 
1083: \startdata 
1084: $P$~(days)                \dotfill & $37.253 \pm 0.007$ & $37.257 \pm 0.003$   \\ 
1085: $T_{SC}$ (HJD--2,448,000) \dotfill & $530.19 \pm 0.76$  & $529.64 \pm 0.15$    \\ 
1086: $K_1$ (km s$^{-1}$)       \dotfill & $17.4 \pm  0.9$    & $14.4 \pm  0.9$      \\ 
1087: $K_2$ (km s$^{-1}$)       \dotfill & \nodata            & $128.2 \pm 2.2$      \\ 
1088: $\gamma_1$ (km s$^{-1}$)  \dotfill & $47.1 \pm  0.6$    & $33.6 \pm 0.7$       \\ 
1089: $\gamma_2$ (km s$^{-1}$)  \dotfill & \nodata            & $31.2 \pm 1.7$       \\ 
1090: $M_2/M_1$                 \dotfill & \nodata            & $0.113 \pm 0.007$    \\ 
1091: $M_1\sin ^3 i$ ($M_\odot$)\dotfill & \nodata            & $10.1 \pm 0.4$       \\ 
1092: $M_2\sin ^3 i$ ($M_\odot$)\dotfill & \nodata            & $1.13 \pm 0.08$      \\ 
1093: $a\sin i$ ($R_\odot$)     \dotfill & \nodata            & $105.0 \pm 1.7$      \\ 
1094: rms$_1$ (km s$^{-1}$)     \dotfill & 7.6                & 6.5                  \\ 
1095: rms$_2$ (km s$^{-1}$)     \dotfill & \nodata            & 14.5                 \\ 
1096: \enddata 
1097: \end{deluxetable} 
1098: \newpage 
1099:  
1100: % Table 4 
1101: \begin{deluxetable}{lcc} 
1102: %\tabletypesize{\scriptsize} 
1103: \tablewidth{0pc} 
1104: \tablenum{4} 
1105: \tablecaption{Stellar Parameters for FY CMa\label{tab4}} 
1106: \tablehead{ 
1107: \colhead{Parameter} & 
1108: \colhead{Primary} & 
1109: \colhead{Secondary} } 
1110: \startdata 
1111: $T_{\rm eff}$ (kK)        & $27.5 \pm 3.0$ & $45 \pm 5$ \\ 
1112: $\log g$ (cgs)            & $4.0$          & $4.3 \pm 0.6$ \\ 
1113: $V\sin i$ (km s$^{-1}$)   & $340 \pm 40$   & $41 \pm 5$ \\ 
1114: $M/M_\odot$	          & $10 - 13$      & $1.1 - 1.5$ \\ 
1115: $R/R_\odot$	          & $4.1 - 7.6$    & $0.5 - 0.9$ \\ 
1116: \enddata 
1117: \end{deluxetable} 
1118: \newpage 
1119:  
1120: 
1121: % Table 5 
1122: \begin{deluxetable}{lccccc} 
1123: \tabletypesize{\scriptsize} 
1124: \tablewidth{0pc} 
1125: \tablenum{5} 
1126: \tablecaption{Red Spectra Measurements\label{tab5}} 
1127: \tablehead{ 
1128: \colhead{Date}             & 
1129: \colhead{Orbital}          & 
1130: \colhead{$W_\lambda$(H$\alpha$)} & 
1131: \colhead{$V$(H$\alpha$ wings)}   & 
1132: \colhead{$V$(He I peaks)} & 
1133: \colhead{$\triangle V$(He I peaks)}        \\  
1134: \colhead{(HJD--2,400,000)} & 
1135: \colhead{Phase}            & 
1136: \colhead{(\AA )}    & 
1137: \colhead{(km s$^{-1}$)}    & 
1138: \colhead{(km s$^{-1}$)}    & 
1139: \colhead{(km s$^{-1}$)}    } 
1140: \startdata 
1141:  46122.805 \dotfill &  0.399 &  
1142:          $-10.98$ &\phn\phs     $  26.8$ &          \nodata    &        \nodata \\ 
1143:  46152.702 \dotfill &  0.201 &  
1144:          $-13.32$ &\phn\phs     $  21.8$ &\phs         $  216$ &        \nodata \\ 
1145:  46734.931 \dotfill &  0.829 &  
1146:          $-11.22$ &\phn\phs     $  71.1$ &             $ -154$ &        \nodata \\ 
1147:  46902.728 \dotfill &  0.333 &  
1148:          $-11.76$ &\phn\phs     $  30.2$ &          \nodata    &        \nodata \\ 
1149:  46903.703 \dotfill &  0.359 &  
1150:          $-12.39$ &\phn\phs     $  25.2$ &          \nodata    &        \nodata \\ 
1151: \enddata 
1152: \tablecomments{Table 5 is available in its entirely in the electronic edition of the 
1153: {\it Astrophysical Journal.}  A portion is shown here for guidance regarding its 
1154: form and content.}
1155: \end{deluxetable} 
1156: 
1157: %\newpage 
1158:  
1159: 
1160: 
1161: \clearpage
1162: 
1163: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
1164: 
1165: % Figures 
1166: 
1167: % Figure 1 
1168: \begin{figure}
1169: \begin{center} 
1170: {\includegraphics[height=16cm]{f1.eps}}
1171: \end{center} 
1172: \caption{The orbital phase variations in the cross-correlation 
1173: functions of the spectra in the $1310-1385$ \AA ~region with 
1174: a hot star template spectrum.  The ccfs are shown as 
1175: linear plots ({\it top panel}) and as a gray-scale image ({\it lower panel}).  
1176: The intensity in the gray-scale image is assigned one of 16 gray levels  
1177: based on its value between the minimum (dark) and maximum (bright) observed values.  
1178: The intensity between observed ccfs is calculated by a linear interpolation  
1179: between the closest observed phases (shown by arrows along the right axis). 
1180: The weak signal from the hot subdwarf spectrum appears as a backwards ``S''
1181: feature in the lower panel.  The scale bar at top right indicates the 
1182: amplitude of the ccf variation relative to the mean sum of the 
1183: squared differences far away from the optimal correlation shift. 
1184: \label{fig1}} 
1185: \end{figure} 
1186: 
1187:  
1188:  
1189: % Figure 2 
1190: \begin{figure} 
1191: \begin{center} 
1192: 
1193: %\plotone{f2.eps}
1194: {\includegraphics[angle=90,height=12cm]{f2.eps}} 
1195: \end{center} 
1196: \caption{Radial velocity curves for the Be star and its companion. 
1197: Orbital phase 0.0 corresponds to the Be star superior conjunction time.  
1198: The solid circles represent the {\it IUE} cross-correlation  
1199: velocities for the Be star while the open circles show the  
1200: {\it IUE} cross-correlation results for the hot subdwarf.  
1201: The plus signs indicate the velocities from measurements of the  
1202: H$\alpha$ wing bisector positions. The typical radial velocity errors are 
1203: 3 and 7 km~s$^{-1}$ for measurements of the Be star and hot subdwarf, 
1204: respectively.
1205: \label{fig2}} 
1206: \end{figure} 
1207:  
1208: % Figure 3 
1209: \begin{figure} 
1210: {\includegraphics[angle=90,height=12cm]{f3.eps}} 
1211: \caption{A comparison of a part of the reconstructed UV spectrum of the 
1212: secondary star ({\it solid line}) with a model spectrum ({\it dotted line}). 
1213: Line identifications for some of the stronger lines are given above the spectra. 
1214: \label{fig3}} 
1215: \end{figure} 
1216:  
1217: % Figure 4 
1218: \begin{figure} 
1219: {\includegraphics[angle=90,height=12cm]{f4.eps}} 
1220: \caption{A comparison of another part of the reconstructed UV spectrum of the 
1221: secondary star ({\it solid line}) with a model spectrum ({\it dotted line}). 
1222: \label{fig4}} 
1223: \end{figure} 
1224:  
1225: % Figure 5 
1226: \begin{figure}
1227: \begin{center} 
1228: {\includegraphics[height=16cm]{f5.eps}}
1229: \end{center} 
1230: \caption{The orbital phase variations in the H$\alpha$ 
1231: emission line in the spectra of FY~CMa are shown in  
1232: linear plots ({\it top panel}) and as a gray-scale image ({\it lower panel}).  
1233: The intensity in the gray-scale image is assigned one of 16 gray levels  
1234: based on its value between the minimum (dark) and maximum (bright) observed values.  
1235: The intensity between observed spectra is calculated by a linear interpolation  
1236: between the closest observed phases (shown by arrows along the right axis). 
1237: The scale bar at top right indicates the spectral flux 
1238: relative to the local continuum flux.
1239: \label{fig5}} 
1240: \end{figure} 
1241:  
1242: % Figure 6 
1243: \begin{figure} 
1244: \begin{center} 
1245: {\includegraphics[height=16cm]{f6.eps}} 
1246: \end{center} 
1247: \caption{The orbital phase variations in the \ion{He}{1} $\lambda 6678$  
1248: emission line plotted in the same format as Fig.~5. 
1249: \label{fig6}} 
1250: \end{figure} 
1251:  
1252: % Figure 7 
1253: \begin{figure} 
1254: {\includegraphics[angle=90,height=12cm]{f7.eps}} 
1255: \caption{The radial velocities of the prominent emission peaks in  
1256: the profile of \ion{He}{1} $\lambda 6678$ plotted as  
1257: a function of orbital phase.  Single dots represent measurements  
1258: of one peak (possibly blended) while dots with vertical segments 
1259: represent the midpoint velocity where two prominent peaks  
1260: were observed.  The end points of these line segments correspond  
1261: to the velocity of each resolved peak. The dotted line  
1262: represents the orbital velocity curve of the hot subdwarf.  
1263: \label{fig7}} 
1264: \end{figure} 
1265:  
1266: % Figure 8 
1267: \begin{figure}
1268: \begin{center} 
1269: {\includegraphics[angle=90,height=12cm]{f8.eps}} 
1270: \end{center}
1271: \caption{An empirical model of the system geometry as viewed  
1272: from above the orbital plane.  Each tick mark corresponds to a 
1273: length of $0.5 a$. The Be primary ({\it left}) is surrounded  
1274: by an equatorial disk ({\it lightly shaded region}) that is heated along  
1275: the edge ({\it darker shaded region}) facing the subdwarf secondary ({\it right}).  
1276: The outer boundary of the Be star disk is probably within the critical Roche surface 
1277: ({\it solid curve}). The center-of-mass of the system is indicated by the vertical tick
1278: mark on the line-of-centers. The orbital phases are indicated on the periphery. 
1279: \label{fig8}} 
1280: \end{figure} 
1281:  
1282: % Figure 9 
1283: \begin{figure} 
1284: \begin{center} 
1285: {\includegraphics[height=16cm]{f9.eps}} 
1286: \end{center} 
1287: \caption{The orbital phase variations in \ion{Si}{3} $\lambda 1299$ 
1288: and other nearby \ion{Si}{3} lines  
1289: plotted in the same format as Fig.~5. These features are a composite 
1290: of broad photospheric lines and narrower ``shell'' features formed in the disk. 
1291: \label{fig9}} 
1292: \end{figure} 
1293: 
1294: % Figure 10 
1295: \begin{figure} 
1296: \begin{center} 
1297: {\includegraphics[height=10cm]{f10.eps}} 
1298: \end{center} 
1299: \caption{Selected profiles of the \ion{Fe}{3} $\lambda 1895$ feature, 
1300: all observed near phase 0.75 but at different epochs. Examples include an 
1301: apparently pure, rotationally-broadened photospheric profile ({\it dotted line}),
1302: a profile with weak shell absorption ({\it thin, solid line}), and an observation
1303: made during a prominent infall phase ({\it thick, solid line}). The vertical 
1304: line above the profiles shows the observed line center that includes the velocity 
1305: shift of the Be star. 
1306: \label{fig10}} 
1307: \end{figure} 
1308:   
1309: % Figure 11 
1310: \begin{figure} 
1311: \begin{center} 
1312: {\includegraphics[height=14.5cm]{f11.eps}} 
1313: \end{center} 
1314: \caption{The absorption strength and velocity of the core of the \ion{Fe}{3}
1315: $\lambda 1895$~\AA\, line versus time and orbital phase. 
1316: \textit{Panel 1}: The horizontal {\it dashed line} represents the core intensity 
1317: above which the feature appears to be primarily photospheric. 
1318: \textit{Panel 2}: The double {\it dashed lines} represent the range in photospheric 
1319: velocity of the Be star. The observations shown with the {\it open diamonds} 
1320: showed either no or minimal shell components. All observations above the top 
1321: line were obtained during infall epochs.  
1322: \textit{Panel 3}: The {\it dashed line} has the same meaning as in panel 1. 
1323: Enhanced shell absorption is seen between phases 0.3 and 0.7.
1324: \textit{Panel 4}: The velocity of the center-of-mass of the system 
1325: is indicted by a {\it dashed line} and the radial velocity curve of the Be star
1326: with a {\it solid line}. Only observations showing a 
1327: shell component are included. Note the obvious presence of infalling material at
1328: most phases except perhaps near phase 0.5 where material outflow is commonly seen 
1329: in close binaries with early type primaries (Peters~2001). The errors for all 
1330: observations are less than or equal to the dimension of the plotted point.    
1331: \label{fig11}} 
1332: \end{figure} 
1333: 
1334: % Figure 12 
1335: \begin{figure} 
1336: \begin{center} 
1337: {\includegraphics[height=16cm]{f12.eps}} 
1338: \end{center} 
1339: \caption{The orbital phase variations in \ion{Fe}{3} $\lambda 1895$ line  
1340: plotted in the same format as Fig.~5.  A narrow ``shell'' feature can be seen  
1341: around orbital phase 0.5. 
1342: \label{fig12}} 
1343: \end{figure} 
1344:  
1345: % Figure 13 
1346: \begin{figure} 
1347: \begin{center} 
1348: {\includegraphics[height=16cm]{f13.eps}} 
1349: \end{center} 
1350: \caption{The orbital phase variations in the \ion{N}{5} $\lambda 1240$  
1351: doublet plotted in the same format as Fig.~5 (in the velocity frame
1352: of the blue member of the doublet).  The strong absorption occurs 
1353: when the Be star is viewed through the wind of the hot subdwarf.  
1354: \label{fig13}} 
1355: \end{figure} 
1356:  
1357:  
1358: % Figure 14 
1359: \begin{figure} 
1360: \begin{center} 
1361: {\includegraphics[height=16cm]{f14.eps}} 
1362: \end{center} 
1363: \caption{The temporal variations in the \ion{N}{5} $\lambda 1240$  
1364: doublet observed in a daily sequence of {\it IUE} observations obtained 
1365: in 1995 March. The lower portion shows the average profile while the upper
1366: panel shows a gray-scale representation of the deviations from
1367: the average as a function of time.  A shell feature 
1368: appeared near orbital phase 0.0 at HJD~2449796.4. 
1369: \label{fig14}} 
1370: \end{figure} 
1371: 
1372: % Figure 15 
1373: \begin{figure} 
1374: \begin{center} 
1375: {\includegraphics[height=16cm]{f15.eps}} 
1376: \end{center} 
1377: \caption{Selected profiles of the \ion{N}{5} doublet observed during the 1995 March
1378: sequence show the details of the development and fading of the \ion{N}{5} shell 
1379: component. The feature was strongest at inferior conjunction of the subdwarf and 
1380: essentially absent near the quadrature points. The vertical {\it dotted} lines show 
1381: the locations of the observed line centers in the velocity frame of the center-of-mass 
1382: of the system. The y-axis flux offsets are (from top to bottom) $0, -4, -10, -15,$ and 
1383: $-20$ in the flux units plotted.
1384: \label{fig15}} 
1385: \end{figure} 
1386: 
1387: \clearpage
1388: 
1389: {\Huge 
1390: 
1391: \begin{center}
1392: ONLINE MATERIAL
1393: \end{center}
1394: 
1395: }
1396: 
1397: \clearpage
1398: 
1399: \begin{verbatim}
1400: 
1401: Title: Detection of a Hot Subdwarf Companion to the Be Star FY CMa
1402: Authors: G. J. Peters, D. R. Gies, E. D. Grundstrom, M. V. McSwain
1403: Table: IUE Radial Velocity Measurements
1404: ================================================================================
1405: Byte-by-byte Description of file: tab2.txt
1406: --------------------------------------------------------------------------------
1407:    Bytes Format Units Label  Explanations
1408: --------------------------------------------------------------------------------
1409:    1-  9 F9.3   d       Date   HJD-2400000
1410:   11- 14 I4     --      Obs.y  Observation UT year
1411:   16- 17 I2     --      Obs.m  Observation UT month
1412:   19- 20 I2     --      Obs.d  Observation UT day
1413:   21- 26 I6     --      SWP    IUE SWP number
1414:   27- 32 F6.3   --      Phase  Orbital phase from Be star superior conjunction
1415:   33- 38 F6.1   km/s    HRV1   Heliocentric radial velocity for Be star
1416:   39- 44 F6.1   km/s    OMC1   Observed minus calculated velocity for Be star
1417:   45- 51 A7     km/s    HRV2   Heliocentric radial velocity for subdwarf
1418:   52- 57 A6     km/s    OMC2   Observed minus calculated velocity for subdwarf
1419: --------------------------------------------------------------------------------
1420: 44170.044 1979-10-23  6963 0.985  51.5  16.6    1.6 -17.9
1421: 44330.226 1980-03-31  8617 0.285  23.9   4.4  159.8   3.4
1422: 44853.156 1981-09-05 14910 0.321  22.2   1.6  138.6  -8.4
1423: 44867.935 1981-09-20 15053 0.717  40.5  -7.3 -123.5 -29.2
1424: 44920.747 1981-11-12 15478 0.135  39.0  16.2    ?     ?  
1425: 44972.649 1982-01-03 15933 0.528  36.8   0.7    ?     ?  
1426: 44974.604 1982-01-05 15957 0.580  33.9  -6.7    ?     ?  
1427: 44975.736 1982-01-06 15979 0.611  34.4  -8.5  -67.0 -16.0
1428: 45571.096 1983-08-24 20769 0.591  43.7   2.3  -48.6 -10.7
1429: 45573.015 1983-08-26 20805 0.642  52.5   7.6    ?     ?  
1430: 45576.057 1983-08-29 20837 0.724  57.4   9.5 -105.8 -10.5
1431: 45578.924 1983-09-01 20874 0.801  50.7   3.4    ?     ?  
1432: 46338.986 1985-09-30 26810 0.201  23.1   3.3  146.0  -7.5
1433: 46340.871 1985-10-02 26829 0.252   8.4 -10.8    ?     ?  
1434: 46342.778 1985-10-04 26847 0.303  15.0  -5.0    ?     ?  
1435: 46343.927 1985-10-05 26868 0.334  11.1 -10.0  142.2   0.2
1436: 46430.639 1985-12-31 27426 0.661  52.3   6.4  -74.0   3.7
1437: 46436.586 1986-01-06 27459 0.821  37.1  -9.5  -61.0  23.4
1438: 46440.366 1986-01-09 27479 0.923  40.8   0.4  -24.4   4.4
1439: 46440.609 1986-01-10 27484 0.929  34.0  -5.8  -16.1   8.0
1440: 46440.784 1986-01-10 27488 0.934  29.2 -10.3  -21.3  -0.7
1441: 46441.374 1986-01-10 27496 0.950  38.5   0.4    ?     ?  
1442: 46441.513 1986-01-11 27499 0.953  35.5  -2.3    ?     ?  
1443: 46441.720 1986-01-11 27504 0.959  36.6  -0.7   10.0  11.6
1444: 46590.362 1986-06-08 28457 0.949  41.6   3.4   -8.5   1.1
1445: 46691.971 1986-09-18 29238 0.676  49.1   2.6    ?     ?  
1446: 46705.946 1986-10-02 29349 0.051  27.2  -1.9    ?     ?  
1447: 46709.960 1986-10-06 29386 0.159  25.0   3.5    ?     ?  
1448: 46744.776 1986-11-10 29646 0.093  29.0   3.4    ?     ?  
1449: 46823.765 1987-01-28 30182 0.213  19.3  -0.2  145.4 -10.6
1450: 46823.786 1987-01-28 30183 0.214  20.7   1.2    ?     ?  
1451: 46852.535 1987-02-26 30392 0.985  40.3   5.4   15.9  -3.6
1452: 46861.455 1987-03-06 30443 0.225  11.5  -7.8  163.1   5.3
1453: 46919.233 1987-05-03 30901 0.776  35.9 -12.0    ?     ?  
1454: 47060.818 1987-09-22 31900 0.576  31.4  -8.8    ?     ?  
1455: 47085.860 1987-10-17 32114 0.248  16.8  -2.4    ?     ?  
1456: 47103.822 1987-11-04 32222 0.730  45.6  -2.3  -81.4  14.6
1457: 47106.832 1987-11-07 32265 0.811  47.1   0.1    ?     ?  
1458: 47113.690 1987-11-14 32317 0.995  34.4   0.3   18.8  -8.4
1459: 47125.983 1987-11-26 32398 0.325  21.9   1.2    ?     ?  
1460: 47141.568 1987-12-12 32505 0.743  48.7   0.7    ?     ?  
1461: 47167.580 1988-01-07 32677 0.441  27.5  -0.9    ?     ?  
1462: 47169.553 1988-01-09 32688 0.494  28.6  -4.5    ?     ?  
1463: 47172.522 1988-01-12 32702 0.574  46.5   6.4  -25.6   0.7
1464: 47198.348 1988-02-06 32872 0.267  18.1  -1.1    ?     ?  
1465: 47206.425 1988-02-14 32925 0.484  30.4  -1.8    ?     ?  
1466: 47215.342 1988-02-23 32968 0.723  47.0  -0.9 -103.4  -8.2
1467: 47225.456 1988-03-04 33039 0.995  39.4   5.3   29.1   2.0
1468: 47240.674 1988-03-20 33119 0.403  27.7   2.3   92.9 -11.5
1469: 47258.449 1988-04-06 33221 0.880  42.9  -0.6  -69.2 -12.9
1470: 47262.396 1988-04-10 33245 0.986  29.1  -5.7   13.1  -7.2
1471: 47271.315 1988-04-19 33315 0.226  23.6   4.3  170.4  12.5
1472: 47293.289 1988-05-11 33508 0.816  47.9   1.0  -72.4  13.9
1473: 47305.145 1988-05-23 33616 0.134  17.1  -5.7    ?     ?  
1474: 47307.279 1988-05-25 33640 0.191  10.7  -9.4  171.8  21.1
1475: 47426.980 1988-09-22 34288 0.404  29.6   4.2    ?     ?  
1476: 47460.861 1988-10-26 34606 0.313  19.8  -0.5    ?     ?  
1477: 47478.914 1988-11-13 34736 0.798  48.7   1.3    ?     ?  
1478: 47514.647 1988-12-19 35074 0.757  38.1  -9.9 -110.3 -13.4
1479: 47515.683 1988-12-20 35082 0.785  55.1   7.4    ?     ?  
1480: 47517.675 1988-12-22 35094 0.838  54.9   9.0    ?     ?  
1481: 47518.576 1988-12-23 35101 0.862  36.4  -8.2    ?     ?  
1482: 47519.662 1988-12-24 35112 0.892  38.4  -4.3    ?     ?  
1483: 47537.921 1989-01-11 35320 0.382  18.1  -5.7  108.1  -9.9
1484: 47581.733 1989-02-24 35614 0.558  35.8  -2.9   27.8  42.0
1485: 47619.671 1989-04-03 35919 0.576  44.3   4.1  -17.7   9.9
1486: 47656.154 1989-05-09 36229 0.555  40.2   1.7    ?     ?  
1487: 47908.837 1990-01-17 38034 0.337  18.9  -2.4    ?     ?  
1488: 47928.730 1990-02-06 38148 0.871  44.4   0.3    ?     ?  
1489: 47999.467 1990-04-17 38626 0.770  48.7   0.8 -103.3  -7.3
1490: 48004.512 1990-04-23 38655 0.905  43.1   1.4    ?     ?  
1491: 48175.728 1990-10-11 39806 0.501  21.4 -12.3   42.8  12.3
1492: 48232.662 1990-12-07 40284 0.029  36.1   5.1   36.2 -18.3
1493: 48297.434 1991-02-09 40846 0.768  42.1  -5.9    ?     ?  
1494: 48349.522 1991-04-03 41276 0.166  10.4 -10.7    ?     ?  
1495: 49313.649 1993-11-22 49290 0.043  29.7  -0.0   67.6   1.8
1496: 49442.353 1994-03-30 50426 0.498  31.3  -2.1    ?     ?  
1497: 49470.349 1994-04-27 50635 0.249  22.7   3.5    ?     ?  
1498: 49686.641 1994-11-30 52938 0.055  29.0   0.3    ?     ?  
1499: 49762.511 1995-02-14 53907 0.091  29.1   3.3   90.1 -10.7
1500: 49788.326 1995-03-11 54102 0.784  56.3   8.6 -100.7  -6.6
1501: 49789.386 1995-03-12 54111 0.813  54.9   7.9  -93.4  -6.2
1502: 49790.383 1995-03-13 54123 0.839  52.7   6.9  -92.8 -15.5
1503: 49791.385 1995-03-14 54134 0.866  45.9   1.5    ?     ?  
1504: 49792.457 1995-03-15 54147 0.895  39.0  -3.5    ?     ?  
1505: 49793.349 1995-03-16 54155 0.919  51.6  10.9    2.7  34.0
1506: 49794.351 1995-03-17 54163 0.946  46.0   7.6   26.9  38.4
1507: 49795.397 1995-03-18 54173 0.974  18.6 -17.4    ?     ?  
1508: 49796.460 1995-03-19 54183 0.002  31.7  -1.7   32.8  -0.4
1509: 49797.356 1995-03-20 54187 0.027  20.9 -10.3   49.8  -2.7
1510: 49798.385 1995-03-21 54202 0.054  32.2   3.4   72.6  -1.4
1511: 49799.568 1995-03-23 54210 0.086  35.8   9.6   87.2  -9.9
1512: 49800.577 1995-03-24 54215 0.113  25.6   1.4    ?     ?  
1513: 49801.415 1995-03-24 54220 0.135  35.6  12.9    ?     ?  
1514: 49802.348 1995-03-25 54223 0.161  31.7  10.3  142.7   3.0
1515: 49803.365 1995-03-26 54230 0.188  20.1  -0.2    ?     ?  
1516: 49804.296 1995-03-27 54238 0.213  23.5   3.9  154.1  -1.8
1517: 
1518: \end{verbatim}
1519: 
1520: \clearpage
1521: 
1522: \begin{verbatim}
1523: 
1524: Title: Detection of a Hot Subdwarf Companion to the Be Star FY CMa
1525: Authors: G. J. Peters, D. R. Gies, E. D. Grundstrom, M. V. McSwain
1526: Table: Red Spectra Measurements
1527: ================================================================================
1528: Byte-by-byte Description of file: tab5.txt
1529: --------------------------------------------------------------------------------
1530:    Bytes Format Units Label      Explanations
1531: --------------------------------------------------------------------------------
1532:    1-  9 F9.3   d       Date     HJD-2400000
1533:   10- 15 F6.3   --      Phase    Orbital phase from Be star superior conjunction
1534:   16- 22 F7.2   0.1nm   EQW      H-alpha equivalent width (Angstroms)
1535:   23- 28 F6.1   km/s    HRV(HA)  Heliocentric radial velocity H-alpha wings
1536:   29- 33 A5     km/s    HRV(HE)  Heliocentric radial velocity He I peaks
1537:   34- 37 A4     km/s    DVR(HE)  He I peak velocity separation
1538: --------------------------------------------------------------------------------
1539: Note (1): HRV(HA) values omitted in Fig. 1 and in the orbital solution of 
1540: Table 3, column 2, for HJD-2400000 = 47639.671 and 47983.611.
1541: --------------------------------------------------------------------------------
1542: 46122.805 0.399 -10.98  26.8   ?   ? 
1543: 46152.702 0.201 -13.32  21.8  216  ? 
1544: 46734.931 0.829 -11.22  71.1 -154  ? 
1545: 46902.728 0.333 -11.76  30.2   ?   ? 
1546: 46903.703 0.359 -12.39  25.2   ?   ? 
1547: 46905.682 0.412 -13.46  34.7   ?   ? 
1548: 46917.654 0.733 -13.94  56.7  -81  56
1549: 46918.676 0.761 -14.15  42.9  -88  81
1550: 46919.674 0.787 -13.37  56.3  -79  94
1551: 46920.675 0.814 -11.87  52.9  -75 102
1552: 46921.670 0.841 -11.76  46.4  -70 115
1553: 46921.684 0.841 -12.86  43.3  -70 120
1554: 47033.012 0.830  -8.52  62.2  -68 107
1555: 47034.004 0.856 -12.34  53.6  -63 113
1556: 47302.650 0.067 -15.87  33.1   ?   ? 
1557: 47303.636 0.093 -15.65  32.5  143  ? 
1558: 47303.642 0.093 -14.93  40.4  147  ? 
1559: 47304.622 0.120 -13.67  35.4  160  ? 
1560: 47304.631 0.120 -14.61  39.6  158  ? 
1561: 47305.623 0.147 -11.51  45.5  171  ? 
1562: 47305.633 0.147 -13.28  42.0  171  ? 
1563: 47306.622 0.173 -10.57  44.4  173  ? 
1564: 47306.632 0.174 -12.96  41.2  175  ? 
1565: 47469.823 0.554 -13.56  49.0  -94  ? 
1566: 47469.831 0.554 -13.43  47.0  -99  ? 
1567: 47469.914 0.556 -13.15  52.8  -98  ? 
1568: 47469.997 0.559 -13.47  54.1  -92  ? 
1569: 47470.026 0.559 -13.63  54.6  -96  ? 
1570: 47470.826 0.581 -11.53  57.0 -100  ? 
1571: 47471.014 0.586 -11.66  54.7 -107  ? 
1572: 47471.846 0.608 -11.87  57.9 -111  ? 
1573: 47471.995 0.612 -12.00  62.8 -124  ? 
1574: 47472.848 0.635 -11.39  66.6 -128  ? 
1575: 47473.033 0.640 -11.66  66.3 -132  ? 
1576: 47560.654 0.992 -11.66  49.9   ?   ? 
1577: 47561.681 0.019 -15.26  46.5   ?   ? 
1578: 47561.792 0.022 -14.90  46.4   ?   ? 
1579: 47636.683 0.033 -10.94  36.0   ?   ? 
1580: 47639.671 0.113 -10.92 -13.9  203  ? 
1581: 47639.680 0.113 -11.53  33.1  207  ? 
1582: 47640.635 0.139 -12.13  32.1  207  ? 
1583: 47939.622 0.164 -10.96  18.8  226  ? 
1584: 47940.860 0.197 -11.92  10.1  233  ? 
1585: 47981.708 0.293  -5.32  10.3   ?   ? 
1586: 47983.611 0.344 -10.74   4.4   ?   ? 
1587: 48313.702 0.204 -13.26  28.5   ?   ? 
1588: 48313.788 0.207 -13.20  29.4  175  ? 
1589: 48314.619 0.229 -12.31  35.4  173  ? 
1590: 48318.615 0.336 -12.59  20.8  138 113
1591: 48319.800 0.368 -15.66  26.3  122 124
1592: 48320.616 0.390 -12.22  26.3  113 132
1593: 48321.610 0.416 -12.08  43.9  113 137
1594: 48514.969 0.606 -12.09  50.5  -66  ? 
1595: 48515.030 0.608 -10.42  53.9  -68  ? 
1596: 48515.972 0.633 -11.30  58.2  -96  ? 
1597: 48516.017 0.634 -11.93  53.9  -92  ? 
1598: 48517.021 0.661 -12.21  58.6   ?   ? 
1599: 49058.749 0.202 -14.58  34.4  179  ? 
1600: 49061.769 0.283 -14.06  39.0  162  94
1601: 49061.773 0.283 -15.28  30.5  160  94
1602: 49443.771 0.536 -14.61  42.9  -52  ? 
1603: 49444.729 0.562 -10.59  49.0  -68  ? 
1604: 49445.749 0.589 -10.22  52.3  -77  ? 
1605: 49447.755 0.643 -13.23  51.7  -96  ? 
1606: 51123.027 0.608 -16.17  52.1   ?   ? 
1607: 51123.034 0.609 -15.94  53.7   ?   ? 
1608: 51124.020 0.635 -16.40  56.9   ?   ? 
1609: 51126.033 0.689 -16.83  55.3   ?   ? 
1610: 51191.897 0.457 -13.75  41.1   ?   ? 
1611: 51192.937 0.485 -14.95  41.7   ?   ? 
1612: 51193.927 0.512 -13.53  46.0   ?   ? 
1613: 51196.937 0.592 -13.35  53.9  -68  ? 
1614: 51197.933 0.619 -14.21  55.1  -73  ? 
1615: 51502.936 0.806 -13.04  59.8 -141  ? 
1616: 51503.977 0.833 -14.12  59.4 -137  ? 
1617: 51504.994 0.861 -13.71  57.7 -112  ? 
1618: 51506.025 0.888 -13.94  54.6 -132  ? 
1619: 51507.019 0.915 -14.53  49.8 -130  ? 
1620: 51508.023 0.942 -14.07  47.1  -53 124
1621: 51509.013 0.969 -13.77  46.8  -14 100
1622: 51510.006 0.995 -15.13  46.0   -6 120
1623: 51510.053 0.997 -15.52  44.6   -2 111
1624: 51510.798 0.017 -15.91  43.7   72  ? 
1625: 51510.847 0.018 -15.76  42.2   73  ? 
1626: 51510.918 0.020 -15.41  44.6   78  ? 
1627: 51510.969 0.021 -15.33  43.5   83  ? 
1628: 51511.946 0.047 -16.08  40.7   97  ? 
1629: 51512.015 0.049 -15.34  41.3   98  ? 
1630: 51512.037 0.050 -15.39  40.7  100  ? 
1631: 51612.632 0.750 -12.91  69.1 -162  ? 
1632: 51612.693 0.751 -13.83  69.8 -160  ? 
1633: 51612.802 0.754 -12.84  69.0 -159  ? 
1634: 51612.822 0.755 -11.87  66.8 -154  ? 
1635: 51613.617 0.776 -12.95  66.5 -157  ? 
1636: 51613.703 0.779 -12.81  67.5 -152  ? 
1637: 51613.712 0.779 -13.04  67.3 -160  ? 
1638: 51613.751 0.780 -13.07  69.2 -158  ? 
1639: 51614.612 0.803 -13.18  69.5 -156  ? 
1640: 51614.829 0.809 -13.68  69.7 -160  ? 
1641: 51615.607 0.830 -13.24  68.1 -158  ? 
1642: 51615.717 0.833 -13.08  67.0 -160  ? 
1643: 51615.759 0.834 -14.07  67.1 -160  ? 
1644: 51615.801 0.835 -14.01  65.8 -152  ? 
1645: 51615.842 0.836 -14.39  65.8 -146  ? 
1646: 51616.623 0.857 -14.39  60.0 -163  ? 
1647: 51617.608 0.883 -13.96  58.7 -138  ? 
1648: 51817.984 0.262 -13.96  35.3  193  ? 
1649: 51818.977 0.288 -14.02  33.6  158  77
1650: 51819.980 0.315 -14.27  35.5  144  96
1651: 51820.996 0.342 -14.20  37.9  140 100
1652: 51821.996 0.369 -13.28  40.9  140 117
1653: 51822.982 0.396 -14.00  44.4  133 122
1654: 51823.971 0.422 -13.46  46.9  131 126
1655: 51824.981 0.449 -12.68  53.2  123 126
1656: 51830.997 0.611 -14.50  57.6  -96  ? 
1657: 51849.861 0.117 -15.69  43.4  175  ? 
1658: 51849.914 0.119 -15.47  40.1  179  ? 
1659: 51849.961 0.120 -14.25  39.9  177  ? 
1660: 51851.046 0.149 -17.06  39.3  181  ? 
1661: 51851.055 0.149 -16.71  40.9  183  ? 
1662: 51851.839 0.170 -16.05  38.9  184  ? 
1663: 51851.854 0.171 -15.72  35.1  184  ? 
1664: 51889.975 0.194 -15.07  29.4  187  ? 
1665: 51890.901 0.219 -14.71  34.3  203  ? 
1666: 51892.904 0.273 -14.60  31.6  195  ? 
1667: 51893.902 0.299 -14.78  32.1  161  83
1668: 51894.901 0.326 -14.63  35.1  147  98
1669: 51895.969 0.355 -14.29  40.9  144 122
1670: 51896.942 0.381 -14.66  40.7  131 135
1671: 51897.937 0.408 -14.56  46.4  126 141
1672: 51898.948 0.435 -14.27  44.7  116 135
1673: 51899.942 0.461 -13.67  48.4  127 143
1674: 51900.935 0.488 -13.71  53.7  112 160
1675: 51901.944 0.515 -13.92  51.0   11  ? 
1676: 51913.671 0.830 -13.54  72.2 -160  ? 
1677: 51913.717 0.831 -13.25  72.8 -164  ? 
1678: 51913.761 0.832 -13.98  72.2 -160  ? 
1679: 51913.807 0.834 -14.11  72.6 -162  ? 
1680: 51913.831 0.834 -14.14  71.6 -160  ? 
1681: 51913.851 0.835 -14.19  71.5 -162  ? 
1682: 51913.931 0.837 -13.91  72.0 -160  ? 
1683: 51913.969 0.838 -13.81  69.8 -160  ? 
1684: 51914.014 0.839 -12.79  69.0 -159  ? 
1685: 51914.706 0.858 -14.23  66.8 -158  ? 
1686: 51914.808 0.860 -14.34  68.7 -157  ? 
1687: 51915.693 0.884 -13.88  66.5 -154  ? 
1688: 51915.709 0.885 -13.94  66.8 -152  ? 
1689: 51915.766 0.886 -13.82  67.5 -151  ? 
1690: 51916.919 0.917 -14.32  64.7 -147  ? 
1691: 51916.960 0.918 -14.32  60.8 -147  ? 
1692: 51916.976 0.919 -14.22  60.9 -152  ? 
1693: 51917.757 0.940 -13.90  57.3 -144  ? 
1694: 53291.988 0.825 -15.21  52.0  -73  98
1695: 53293.014 0.853 -15.12  49.4  -64 107
1696: 53294.994 0.906 -15.09  41.1  -43 132
1697: 53294.996 0.906 -15.00  38.1  -45 132
1698: 54020.984 0.392 -13.53  31.7   ?   ? 
1699: 54024.967 0.499 -13.95  39.9   ?   ? 
1700: 
1701: 
1702: \end{verbatim}
1703: 
1704: \end{document} 
1705: