1: %% The command below calls the preprint style
2: %% which will produce a one-column, single-spaced document.
3: \documentclass[12pt,preprint]{aastex}
4:
5: %% manuscript produces a one-column, double-spaced document:
6: % \documentclass[manuscript]{aastex}
7:
8: %% preprint2 produces a double-column, single-spaced document:
9: %\documentclass[preprint2]{aastex}
10:
11: %% If you are submitting to a journal that translates manuscripts
12: %% into SGML, you need to follow certain guidelines when preparing
13: %% your macros. See the AASTeX v5.x Author Guide
14: %% for information.
15:
16: \newcommand\5{{\footnotesize V}}
17: \newcommand\4{{\footnotesize IV}}
18: \newcommand\3{{\footnotesize III}}
19: \newcommand\2{{\footnotesize II}}
20: \newcommand\1{{\footnotesize I}}
21: \newcommand\lam{$\lambda$}
22: \newcommand{\sref}[1]{\S\ref{#1}}
23: \newcommand{\eref}[1]{(\ref{#1})}
24: \newcommand{\cf}{{\em c.f.}}
25: \newcommand{\ea}{{et al.}}
26: \newcommand{\eg}{{\em e.g.}}
27: \newcommand{\ie}{{\em i.e.}}
28: \newcommand\kms{\ensuremath{\mbox{km s}^{-1}}}
29: \newcommand\vsini{\ensuremath{v_{\rm e}\sin{i}}}
30: \newcommand{\p}{\ensuremath{\phantom{:}}}
31: \newcommand{\pp}{\ensuremath{\phantom{-}}}
32: \newcommand{\ph}[1]{\ensuremath{\phantom{#1}}}
33: \newcommand\Teff{$T_{\rm eff}$}
34:
35: %\usepackage{rotate}
36: %\usepackage{epsf}
37: %\usepackage{epsfig}
38:
39: %% You can insert a short comment on the title page using the command below.
40:
41: \slugcomment{Accepted by PASP}
42: \shorttitle{UV Spectra of B Stars in the SMC}
43: \shortauthors{Evans et al.}
44:
45: %% This is the end of the preamble. Indicate the beginning of the
46: %% paper itself with \begin{document}.
47:
48: \begin{document}
49:
50: %% LaTeX will automatically break titles if they run longer than
51: %% one line. However, you may use \\ to force a line break if
52: %% you desire.
53:
54: \title{The ultraviolet and optical spectra of luminous
55: B-type stars in the Small Magellanic Cloud}
56:
57: %% Use \author, \affil, and the \and command to format
58: %% author and affiliation information.
59: %% Note that \email has replaced the old \authoremail command
60: %% from AASTeX v4.0. You can use \email to mark an email address
61: %% anywhere in the paper, not just in the front matter.
62: %% As in the title, you can use \\ to force line breaks.
63:
64: \author{ C. J. Evans\altaffilmark{1},
65: D. J. Lennon\altaffilmark{1},
66: N. R. Walborn\altaffilmark{2},
67: C. Trundle\altaffilmark{1,3},
68: S. A. Rix\altaffilmark{1}
69: }
70:
71: \altaffiltext{1}{Isaac Newton Group of Telescopes,
72: Apartado de Correos 321,
73: 38700 Santa Cruz de la Palma,
74: Canary Islands,
75: Spain}
76: \altaffiltext{2}{Space Telescope Science Institute,
77: 3700 San Martin Drive,
78: Baltimore,
79: MD 21218,
80: USA}
81: \altaffiltext{3}{Dept. of Pure and Applied Physics,
82: The Queen's University of Belfast,
83: Belfast,
84: BT7 1NN,
85: N. Ireland}
86:
87: \begin{abstract}
88: We present ultraviolet spectra from the Space Telescope Imaging
89: Spectrograph (STIS) of 12 early B-type stars in the Small Magellanic
90: Cloud (SMC), comprising 9 supergiants and 3 giants. A morphological
91: comparison with Galactic analogues is made using archival data from
92: the {\it International Ultraviolet Explorer (IUE)}.
93: In general, the intensity of the P Cygni emission in the UV resonance
94: lines is greater, and seen to later spectral types, in the Galactic
95: spectra than in their metal-deficient SMC counterparts; we attribute these effects as
96: most likely arising from weaker stellar winds in the
97: SMC targets, as predicted by radiatively driven wind theory.
98:
99: We also include unpublished STIS observations of two late O-type stars
100: in the SMC. In combination with the B-type observations, and published O-type
101: data, we now have an extensive ultraviolet spectral library of
102: metal-deficient stars, of use in the study of unresolved starbursts
103: and high redshift, star-forming galaxies. In this context, we present
104: empirical measurements for the B-type spectra of the new `1978 index'
105: suggested by Rix et al. as a probe of metallicity in such systems.
106:
107: \end{abstract}
108:
109: \keywords{Galaxies: individual: Magellanic Clouds -- stars: fundamental parameters --
110: stars: winds, outflows -- stars: early-type}
111:
112: \section{Introduction}
113: \label{intro}
114:
115: Ultraviolet (UV) spectra are useful in the study of relatively nearby
116: blue galaxies undergoing intense star formation
117: (so-called starburst galaxies), and in high redshift star-forming
118: galaxies in which the stellar UV features are shifted into the
119: optical. Studies of the composite spectra from such unresolved,
120: star-forming systems generally rely on spectral libraries to
121: investigate the stellar populations therein, with the libraries
122: compiled from UV observations of individual early-type
123: stars (e.g. Leitherer et al. 2001).
124:
125: Distant star-forming systems are often less chemically evolved than
126: the Milky Way (e.g. Pettini et al. 2002). With a metallicity of
127: roughly one-fifth solar (e.g.~Venn, 1999), the Small Magellanic Cloud
128: (SMC) offers an excellent opportunity for compilation of UV spectral
129: libraries more appropriate to the accurate interpretation of
130: metal-poor unresolved systems than if Galactic templates were used.
131: Libraries of high-quality UV data also enable detailed comparisons
132: with the synthetic spectra calculated from stellar atmosphere codes,
133: with a view to including theoretical spectra into population synthesis
134: models.
135:
136: Walborn et al (2000; hereafter Paper~1) presented a high-resolution UV
137: spectral atlas of O-type stars in the SMC using data obtained with the
138: {\it Hubble\/ Space\/ Telescope (HST)} Space Telescope Imaging
139: Spectrograph (STIS) from Guest Observer program GO7437. The
140: incorporation of these spectra into the population synthesis code {\sc
141: starburst}{\footnotesize 99}, and an assessment of their impact on the
142: resultant integrated spectra has been presented by Leitherer et
143: al. (2001). However, as De Mello et al. (2000) demonstrated, the
144: contribution from B-type stars in integrated UV spectra can be
145: important in young stellar populations. With this in mind we obtained
146: new STIS spectroscopy of 12 B-type stars in the SMC, with the sample
147: selected to cover a range of effective temperatures and luminosities.
148: These data have already been used to constrain the age of the stellar
149: population observed in a `super-star cluster' in the starburst galaxy
150: NGC 1705 (V$\acute{\rm a}$zquez et al. 2004), illustrating their
151: usefulness in the study of unresolved populations.
152:
153: The new B-type STIS spectra have a much greater wavelength
154: coverage than those in Paper~1 (\lam\lam1150-2350
155: cf. \lam\lam1150-1750), achieved by observing each target
156: using the echelle gratings at two central wavelengths (see Section
157: \ref{obsdetails}). Our motivation behind this strategy was to obtain
158: high resolution observations of the iron `forest' of absorption lines
159: from Fe$^{2+}$, which lies in the \lam\lam1850-2100 region. Since
160: these observations, Rix et al. (2004) have published a theoretical
161: study in which synthetic UV spectra from the WM-Basic code (Pauldrach
162: et al. 2001) are included in {\sc starburst}{\footnotesize 99}. They
163: advocate an equivalent-width index centered at 1978 \AA\/ as a useful
164: metallicity indicator, and the new STIS data presented here offer an
165: opportunity to begin to calibrate this index empirically.
166:
167: Aside from their use as spectral templates, UV observations of
168: individual early-type stars in the SMC enable detailed studies of
169: stellar evolution and radiatively driven winds in a metal-poor
170: environment. For instance, some of the O-type spectra presented in
171: Paper~1 were used in detailed model atmosphere analyses by Bouret et
172: al. (2003) and Hillier et al. (2003). Similarly, four of the B-type
173: STIS spectra discussed here were used by Evans et al. (2004a) to
174: determine stellar wind terminal velocities ($v_\infty$), that were
175: then used in the analysis of the optical spectra by Trundle et al. (2004).
176:
177: Here we primarily concern ourselves with a discussion of the
178: morphological phenomena seen in the UV spectra of
179: our SMC targets, and compare them with Galactic analogues observed
180: with the {\it International\/ Ultraviolet\/ Explorer (IUE)} telescope.
181: We also discuss the correlation between the new UV data and
182: high-resolution optical data of our targets, in particular including
183: observations of the H$\alpha$ Balmer line.
184:
185:
186: \section{Observations}
187: \label{obsdetails}
188:
189: \subsection{UV spectroscopy}
190:
191: For our $HST$ General Observer program (GO9116) we observed
192: 12 B-type stars in the SMC using the STIS; observational parameters of
193: our targets are given in Table~\ref{targets}. Each star was observed
194: at two central wavelengths in echelle mode; with the far-UV MAMA
195: (Multi-Anode Microchannel Array) detector and the E140M grating
196: (centered at 1425 \AA), and with the near-UV MAMA and E230M grating
197: (centered at 1978 \AA). The echelle observations were made through
198: the $0{\farcs}2~\times~0{\farcs}2$ entrance aperture and the effective
199: spectral resolving powers of the E140M and E230M gratings are $R =
200: 46,000$ and 30,000 respectively. The exposure times were set
201: from consideration of low-resolution $IUE$ spectra and are given in
202: Table \ref{targets}.
203:
204: The primary {\sc calstis} processing steps are described in Paper~1,
205: with the difference here that the two-dimensional inter-order
206: background correction performed manually in those reductions has now
207: been incorporated into the standard pipeline as {\sc sc}{\footnotesize
208: 2}{\sc dcorr}. As described by Evans et al. (2004a), after pipeline
209: reduction the individual echelle orders were extracted and merged to
210: form a continuous spectrum. Prior to merger (typically) 25 pixels at
211: both ends of each order are clipped; the spectral overlap between
212: orders is generally large enough to accommodate this and the final
213: result has a more consistent signal-to-noise ratio than that obtained
214: otherwise. Blueward of 1650 \AA\/ there is a significant decrease in
215: the signal-to-noise ratio of the E230M data. therefore this region is
216: removed. The E140M and E230M spectra are then merged to form a final
217: spectrum covering 1150-2350 \AA. For display purposes (and for
218: consistency with Paper~1) the spectra are binned to 0.25\,\AA~per data
219: point. The signal-to-noise ratio (on the basis of photon statistics)
220: is typically in the range of 20-30 per binned data point.
221:
222: In Table \ref{targets} we also give H~\1 column densities, $N$(H~\1),
223: toward our targets, found assuming that Lyman-$\alpha$ has a pure
224: damping profile (\eg~Shull \& Van Steenberg 1985). To avoid possible
225: contamination by the N~\5 \lam\lam1239, 1243 doublet, we consider only
226: the blueward wing when deriving the tabulated values. The column
227: densities are consistent with those for other stars in the SMC,
228: e.g. Tumlinson et al. (2002).
229:
230: There are archival spectra from the {\it Far Ultravoiolet
231: Spectroscopic Explorer (FUSE)} of the type Ia stars, however many
232: suffer from large column densities of H$_2$ and, in the context of the
233: present discussion, reveal little further information such that they
234: are not considered here (A. Fullerton, private communication, 2004).
235:
236:
237: \subsubsection{Additional UV observations of O-type stars}
238: \label{7437}
239:
240: Observations for the GO7437 program were incomplete when Paper 1
241: was written. We now briefly discuss additional STIS
242: observations with the E140M grating of AzV\,423 (Azzopardi \& Vigneau
243: 1975, 1982) and NGC~346 MPG~487 (Massey et al. 1989). The optical
244: spectrum of AzV~423 was classified as O9.5 II(n) by Walborn et
245: al. (2002), in their discussion of its spectrum from $FUSE$; the $HST$
246: data for this star have not been published previously. NGC~346
247: MPG~487 was classified as O8 V by Evans (2001) and Bouret et al.
248: (2003), with the latter presenting a model atmosphere analysis
249: in which parts of the STIS spectrum were shown (their Figure~9)
250: These two stars were not incorporated in {\sc starburst}{\footnotesize 99}
251: with the main O-type sample (Leitherer et al., 2001) and offer
252: additional early-type template spectra for population synthesis work.
253:
254: The two STIS spectra are shown in Figure~\ref{extras}.
255: The spectrum of NGC~346 MPG~487 is primarily photospheric in origin
256: and complements the data in Paper~1 well, providing better sampling of
257: the main sequence. The new spectrum of AzV~423 is more interesting,
258: particularly in comparison to that of AzV~327 from Paper~1 (also shown
259: in Figure~\ref{extras}; the Si~\4
260: and C~\4 profiles are more strongly in emission in AzV~423, although
261: the C~\4 absorption is weaker. In contrast, the N~\5 doublet displays
262: relatively strong emission in AzV~327, compared to none in AzV~423.
263: We also have unpublished, red optical data for these two stars from
264: the University College London echelle spectrograph (UCLES) at the 3.9-m
265: Anglo-Australian Telescope (AAT); the H$\alpha$ Balmer profiles of
266: both stars are in absorption and are practically identical.
267: Comparisons of the blue optical data (see Paper~1 and Walborn et al. 2002)
268: are complicated somewhat by the slightly broader lines in AzV~423, but
269: it appears that C~\3 is stronger, and N~\3 weaker, in AzV~423. We
270: suggest that the differences in the UV spectra may arise from
271: different amounts of mixing of CNO-processed material into the
272: atmosphere and winds of the two stars. Note that the spectrum of
273: AzV~327 in Figure~\ref{extras} appears to be noisier than that of
274: AzV~423; the exposure times were identical and the cosmetic difference
275: arises from the (apparently) faster rotation of AzV~423.
276: Considering their similar magnitudes (suggesting similar physical
277: luminosities) future quantitative analysis of these stars would
278: be of interest.
279:
280:
281: \subsection{Optical data}
282: In addition to the UV data we also consider high-resolution optical
283: spectra of our targets. Blue and red region optical spectra of the
284: eight targets from the AzV catalogue were obtained with the UV-Visual
285: Echelle Spectrograph (UVES) at the ESO Very Large Telescope (VLT) in
286: November 2001. The effective resolving power of the UVES set-up was
287: $R \sim 20000$ and the signal-to-noise ratio of the data was in excess
288: of 100. Full observational details are given by Trundle et
289: al. (2004). For AzV\,22 and 362, the H$\alpha$ spectra are those from
290: the New Technology Telescope (NTT) using the ESO Multi Mode Instrument
291: (EMMI). These are preferred over the VLT data following the
292: discussion of nebular contamination by Trundle et al.
293:
294: Blue region spectra for the four targets in NGC\,330 were obtained
295: using the CASPEC spectrograph at the ESO 3.6-m telescope in August 1994
296: and September 1996. The resolution of the spectra is comparable to the UVES
297: data and the signal-to-noise ratio ranges from 40 to 90; further details
298: are given by Lennon et al. (2003).
299:
300: \section{Morphological comparison with Galactic $IUE$ spectra}
301: \label{morph}
302:
303: \subsection{Luminosity class Ia stars}
304: The STIS spectra of our class Ia SMC targets (with the exception of
305: AzV~104, see Section \ref{optical}) are shown in Figure \ref{fig1}. Galactic
306: analogues to these six stars from the Walborn et al. (1995a) $IUE$
307: atlas are shown in Figure \ref{fig2}. For display purposes the $IUE$
308: spectra have been smoothed using a 5-pixel `boxcar' median
309: filter. There are extant high-resolution, long-wavelength camera
310: observations (i.e., redward of 1950\,\AA) in the $IUE$ archive for the
311: majority of the stars here, however the spectra are typically of
312: low quality and are not included in our discussions. When compiling
313: Galactic analogues we have deliberately omitted HD\,152236
314: ($\zeta^1$~Sco, B1.5 Ia$+$), because of its peculiar nature, and identification as
315: a candidate luminous blue variable (e.g., Sterken et al. 1997).
316: Instead we include HD\,148688 (B1~Ia) and HD\,96248 (BC1.5~Iab) as the
317: most suitable stars for comparison with Sk~191 and AzV~210 (both
318: B1.5~Ia). As noted by Evans et al. (2004a; Fig.~2 therein), the morphology of the
319: Si~\4 doublet in AzV~215 is peculiar; there is a relatively flat
320: `shelf' in the profiles, resembling those seen in $\kappa$~Ori
321: (HD~38771, B0.5 Ia, Walborn et al. 1995a), the origins of which are
322: currently unclear.
323:
324: As one progresses from B0 to B5, there are a number of clear trends.
325: In the SMC data the P Cygni profiles of the Si~\4 and C~\4 lines
326: weaken at later types (see Figure \ref{fig1}); there
327: is evidence of a weak stellar wind in the Si~\4 lines of AzV~18
328: (B2~Ia), with only photospheric features seen in AzV~362 (B3 Ia).
329: The $HST$ Faint Object Spectrograph (FOS) observations of AzV~488
330: (B0.5~Ia) presented by Walborn et al. (1995b) fit neatly into this
331: sequence. The SMC data contrast with the Galactic data (see Figure \ref{fig2}),
332: in which obvious wind signatures are seen through to at least
333: B3~Ia (HD\,111973).
334:
335: Furthermore, at a given spectral type, the P Cygni emission
336: in the SMC targets is generally weaker than in the Galactic
337: data. This is a well documented phenomenon in early-type spectra
338: (e.g., Hutchings, 1982; Bruhweiler et al. 1982) and is most striking
339: in the current comparisons for C~\2 \lam\lam1334, 1336 and Al~\3
340: \lam\lam1855, 1863. In contrast, the Si~\4 \lam\lam1394, 1403 lines
341: in HD~96248 and 148688 are interesting in that, although they display
342: weak emission, the intensity is less than in the SMC spectra. In the
343: case of HD~96248 this could be potentially explained as arising from
344: its lower luminosity, with the significant C~\2 and C~\4 emission due
345: to an enhanced carbon abundance (the star is classified as BC-type).
346: However, the Al~\3 features in HD~96248 are comparable to those seen
347: in HD~148688, suggesting that abundance anomalies are not solely
348: responsible.
349:
350: It is apparent from Figures \ref{fig1} and \ref{fig2} that the lower
351: metallicity of the SMC is manifested by significantly weaker Fe~\3
352: absorption in the `forest' redward of the Al~\3 doublet, although the
353: extent of the $IUE$ data in this region is relatively limited. Figure
354: \ref{fig1} also neatly illustrates the change of the dominant
355: ionization stage of iron in early B-type stars. In AzV~215 the
356: multitude of Fe~\4 lines in the \lam1550-1700 region are clearly
357: visible, with relatively weak Fe~\3 lines; this situation is reversed
358: in Sk~191 and AzV~210. These features are also present in the
359: Galactic data, although it is less obvious given the greater
360: metallicity and reduced wavelength coverage. Walborn \&
361: Nichols-Bohlin (1987) noted that there appeared to be a distinct
362: difference in the stellar wind features between B0.5~Ia and B0.7~Ia
363: spectra. This correlates with the change in iron ionization seen
364: here, suggested by Vink et al. (2001) as a physical explanation
365: for the so-called `bi-stability' of stellar winds.
366:
367:
368: \subsection{Lower luminosity stars}
369: We now move to the less luminous $HST$ targets,
370: shown in Figure \ref{fig3}. In Figure \ref{fig4} we again show Galactic
371: analogues drawn from the atlas of Walborn et al. (1995a).
372:
373: Classified as B0.5 Ia by Lennon (1997), AzV~104 appears to be
374: exceptional in that its physical properties do not appear to correlate
375: well with its blue-region optical morphology. A critical
376: re-examination of the optical spectra of this star is presented in
377: Section \ref{optical}. For now we compare our STIS spectrum with that
378: of HD\,213087 (B0.5~Ib). In their analysis of terminal velocities,
379: Evans et al. (2004a) were unable to apply the `Sobolev with exact integration'
380: (SEI) method to the UV resonance lines of AzV~104,
381: although from inspection of the Si~\4 and C~\4 profiles it appears
382: that there is some wind present, albeit particularly weak. In
383: contrast, the wind in HD~213087 appears more significant (ignoring for
384: the moment any potential luminousity effects arising from mis-matching
385: classifications).
386:
387: The remainder of the lower luminosity STIS spectra are largely
388: photospheric in origin, with no evidence of stellar winds in their
389: resonance lines. Similarly, in the Galactic spectra, HD~218376
390: (B0.5~III), HD~147165 (B1~III), and HD~141381 (B2~II) also contain
391: only photospheric features. However, in the two class Ib Galactic
392: spectra, i.e., HD~51309 and 36371, weak P Cygni profiles are seen in
393: the C~\4 doublet. As in the class Ia stars, evidence of stellar winds
394: is seen to later spectral types for class Ib in the Galaxy than in the
395: SMC. Again, the dominant stage of iron is revealed by the spectra,
396: the \lam\lam1550-1700 Fe~\4 lines are clearly visible in AzV~104, with the
397: \lam\lam1850-2000 Fe~\3 lines strongly in absorption in NGC330-B37 and
398: NGC330-A02.
399:
400: \section{Morphological comparisons of the ultraviolet and optical data}
401: \label{optical}
402: In Figure \ref{opt} we show optical spectra for the seven class Ia stars.
403: The data have been smoothed and rebinned to an effective resolution of
404: 1.5~\AA\/ (cf. Walborn \& Fitzpatrick, 1990) and have also been
405: corrected for the measured recession velocities.
406:
407: Given their excellent signal-to-noise ratio, the blue-region
408: optical data offer a wonderfully clear illustration of the trends in
409: early B-type spectra. The principal temperature diagnostic in these
410: stars is the ionization balance of Si~\4, Si~\3, and Si~\2. The
411: optical lines marked in Figure \ref{opt} are well complemented by
412: several UV photospheric lines (e.g., Si~\2 \lam1265, and Si~\3
413: \lam\lam1294, 1299, 1417, as marked in Figure \ref{fig1}) that were
414: suggested by Massa (1989) and Prinja (1990) as potential UV
415: temperature diagnostics. Indeed, from simple inspection of the Si~\2
416: \lam1265 line alone in AzV~210, one would be tempted to suggest a cooler
417: temperature than Sk~191, even though the spectral types are identical.
418: These morphological arguments are confirmed by the model atmosphere
419: analysis of Trundle et al. (2004) which found a temperature of 20,500~K
420: for AzV~210, some 2,000~K lower than that found for Sk~191.
421:
422: Similarly, the optical spectra neatly illustrate the increase and then
423: decrease in the intensity of the main O~\2 features (\lam\lam4072-76,
424: 4345-49, 4415-17) as one moves from
425: B0$\rightarrow$B1.5$\rightarrow$B3. The H$\alpha$ Balmer profiles of
426: the stars are also shown in Figure \ref{opt}, with a wide range of
427: morphologies displayed, i.e., pure absorption (albeit very slightly
428: `filled-in' by wind effects, AzV\,104), `filled-in' absorption
429: (AzV\,210), weak P-Cygni (AzV\,215) and broad emission (Sk\,191).
430:
431: From comparisons with the UV spectra in Figure~\ref{fig1} (and AzV\,104
432: in Figure~\ref{fig3}), the optical data highlight the strong
433: correlation of the UV and H$\alpha$ profiles for the earliest B-type
434: spectra, i.e., AzV\,215, AzV\,104, AzV\,210 and Sk\,191. In contrast, the
435: significant H$\alpha$ emission in AzV~18 (B2~Ia), AzV~362 (B3~Ia), and
436: AzV~22 (B5~Ia) appear initially to contradict the UV spectra; the
437: H$\alpha$ emission suggests some degree of stellar wind, yet the UV
438: spectra display little or no evidence. This is attributable to the
439: range of temperatures covered by our sample, e.g., the dominant
440: ionization stage of silicon changes from Si$^{3+}$ $\rightarrow$
441: Si$^{2+}$ $\rightarrow$ Si$^{1+}$ as one progresses from B0 to B5.
442: Thus, even in the presence of a stellar wind, P Cygni emission is not
443: expected in the Si~\4 doublet at the latest types, although weak
444: photospheric features are still seen.
445:
446:
447: \subsection{AzV~104}
448: Particularly striking in Figures \ref{fig4} and \ref{opt} are the
449: spectra of AzV\,104 (B0.5 Ia), in which the H$\alpha$ absorption tallies with
450: the lack of obvious wind signatures in the UV; in this case it would
451: appear that the luminosity has not been predicted correctly by
452: morphological considerations.
453: The optical classification employed here is that from Lennon
454: (1997), in which luminosity class was assigned using the H$\gamma$
455: equivalent width calibrations of Azzopardi (1987). From photometric
456: and intrinsic colour arguments, Trundle et al. (2004) gave $M_v =
457: -$5.82, more in keeping with that expected from a B0.5 Ib star (e.g.,
458: Humphreys \& McElroy, 1984) than one classified as B0.5 Ia.
459:
460: The overestimated luminosity is not obviously a metallicity effect.
461: The principle advantage of using H$\gamma$ as a luminosity diagnostic
462: is that it is expected to be relatively unaffected by changes in metallicity.
463: In contrast, if Galactic criteria involving metal-line intensities are
464: applied directly to B-type SMC spectra, the luminosity class is often
465: underpredicted (e.g. Walborn 1983).
466: We do not suggest that the luminosity class be
467: revised on the basis of this discussion (such a change would against the
468: philosophy of the `MK process'), but merely comment that in this case
469: the H$\gamma$ calibrations appear to incorrectly predict the intrinsic
470: luminosity.
471:
472:
473: \section{Discussion}
474: \label{discussion}
475:
476: \subsection{Weaker winds at low metallicity?}
477: A long-standing challenge of observational studies of massive stars
478: has been to test conclusively the theoretical prediction that
479: mass-loss scales with metallicity such as $\dot{M}(Z) \propto
480: Z^{0.5\rightarrow0.7}$ (Kudritzki et al. 1987; Vink et al. 2001).
481:
482: Recent results from Evans et al. (2004b) and Trundle et al. (2004)
483: have highlighted that the predictions from the mass-loss `recipe' of
484: Vink et al. (2001) do not tally with the observationally derived rates
485: in the Magellanic Clouds. However, contemporary results from Crowther
486: et al. (2004) suggest similar differences between observational and
487: theoretical mass-loss rates for Galactic B-type supergiants, i.e.,
488: there appears to be tentative evidence for observational confirmation
489: of some metallicity dependence, with the absolute values differing to
490: those of the Vink et al. (2001) purely as a consequence of the
491: theoretical methods employed.
492:
493: The comparison of the Galactic and SMC spectra in Section \ref{morph}
494: offers some insight in this context. In general, from the more
495: intense P Cygni emission at a given spectral type, a stronger stellar
496: wind can be inferred in the Galactic stars than in our SMC targets.
497: At first, such behaviour in the Clouds was attributed solely to the
498: lower abundances (e.g., Bruhweiler et al. 1982), with wind effects
499: thought to play a less significant role. However, in a star such as
500: AzV~215, the resonance lines are strongly saturated and the line is
501: only weakly dependent on abundance changes, even over the large range
502: of SMC to solar metallicity; the stellar wind dominates the formation
503: of the resulting profile.
504:
505:
506: Stronger winds in the Galactic targets can also be inferred from the
507: presence of P Cygni features at later spectral types than those seen
508: in the SMC spectra. This point is made most strongly by the saturated
509: C~\4 doublet in HD~51309 (B3~Ib, see Figure \ref{fig4}) compared to
510: the weak photospheric absorption seen in NGC330-B37 (also B3~Ib); such
511: a difference cannot be explained by abundance effects alone.
512:
513: It is unlikely that different temperatures can account for these
514: effects; the results found by Trundle et al. (2004) for their SMC
515: targets are comparable to those found for Galactic B-type supergiants
516: by Crowther et al. (2004). Similarly, though a relatively wide range
517: of physical luminosity is possible for stars classified as type Ia,
518: the Galactic and SMC samples do not appear to be particularly
519: mismatched. For instance, Kudritzki et al. (1999) give the luminosity
520: of HD~14818 (B2 Ia) as log(L/L$_\odot$) = 5.47, cf. 5.44 for AzV~18
521: (also B2 Ia, Trundle et al. 2004).
522:
523: Interestingly, Evans et al. (2004a) found no evidence for a significant
524: scaling in the wind terminal velocities in early B-type UV spectra;
525: four of their targets were drawn from the STIS data presented here
526: (namely AzV\,18, AzV\,210, AzV\,215, and Sk~191). So, in comparison to Galactic
527: stars, the early B-type SMC supergiants appear to have weaker winds,
528: but with comparable terminal velocities, in agreement with the predictions
529: of radiatively driven wind theory.
530:
531:
532: \subsection{Ultraviolet Iron Indices and their application to unresolved systems}
533: \label{Fe_index}
534:
535: The STIS data have a further application in the context of population
536: synthesis models. Rix et al. (2004) have proposed an equivalent-width
537: index centered at 1978~\AA\/ as a useful probe of metallicity in
538: high redshift, star-forming galaxies. This index is effectively the
539: summed photospheric absorption between 1935 and 2020
540: \AA\/ (which largely arises from Fe~\3 lines, e.g., Swings et
541: al. 1976) and, in the case of constant star formation, was found to be
542: more sensitive than the indices suggested by Leitherer et al. (2001)
543: centered at 1370~\AA\/ and 1425~\AA.
544:
545: The data presented here provide an opportunity to begin to
546: investigate this index empirically. Following correction for the
547: stellar radial velocities given in Table \ref{targets}, each of the
548: STIS spectra was treated in the same manner as the models used by Rix
549: et al., i.e., they were degraded to an effective resolution of 2.5
550: \AA\/ (to ensure compatibility with contemporary observations of high
551: redshift systems) and then rectified using the same continuum windows. We
552: include the equivalent widths of the 1978 index in Table
553: \ref{targets}. The measurement uncertainties largely
554: arise from the difficulty in accurate placement of the continuum and
555: are typically $\pm$10$\%$ when W$_\lambda$(1978) $>$ 5 \AA, increasing
556: to 15-20$\%$ when W$_\lambda$(1978) $<$ 5 \AA. The measurements for
557: the class Ia stars provide a further illustration of the temperature
558: sequence in early B-type spectra. The equivalent width is greatest at
559: B2, and is seen to increase significantly between B0.5 and B1.5, the
560: region in which the `bi-stability jump' in the behaviour of stellar
561: winds is found.
562:
563: As one might expect, there is some degeneracy between the index for
564: the early-B, class Ia objects and later-type, less luminous stars
565: (i.e., AzV\,215 and AzV\,104 cf. NGC330-B22); similar conclusions were
566: reported by Swings et al. (1976) from their low-resolution spectra,
567: obtained using the European TD1 satellite. In a brief episode of star
568: formation this degeneracy (effectively the age/metallicity degeneracy)
569: can be quite limiting but, coupled with other information from the
570: spectrum such as the appearance of the resonance lines, the dominant
571: stellar population can be better constrained. For example, the
572: dominant population in an integrated spectrum displaying significant P
573: Cygni emission in the resonance lines, with a large 1978 equivalent
574: width, is likely to be much less evolved than one in which there is no
575: evidence of wind features, coupled with a smaller 1978 index. Similar
576: morphological arguments were used by V${\rm \acute{a}}$zquez et
577: al. (2004), in combination with theoretical evolutionary isochrones to
578: constrain the age of the super-star cluster NGC~1705-1.
579:
580:
581:
582: \subsection{The combined SMC sample}
583:
584: In Table \ref{hrdtab} we compile the full list of stars observed with
585: $HST$ for our O- and B-type programs (i.e., GO7437 and GO9116) and in
586: Figure \ref{hrd} we present an H-R diagram for the sample. The
587: evolutionary tracks (dotted lines) are from Charbonnel et al. (1993),
588: with $Z$~=~0.004 and standard mass-loss rates; for clarity the 120,
589: 85, 60 and 40 M$_\odot$ models are truncated at the onset of helium
590: burning. Our intent here is merely to guide the reader to the
591: approximate mass range of our targets so for simplicity we prefer
592: these over newer evolutionary tracks (e.g. those of Maeder \& Meynet,
593: 2001), in part to remove the need to assume initial rotational
594: velocities for each star.
595:
596: Where available, temperatures and luminosities have been taken from
597: published model atmosphere analyses, as indicated in Table
598: \ref{hrdtab}. For the remaining stars, temperatures have been interpolated
599: between recent results from line-blanketed model atmospheres (Martins
600: et al. 2002, Crowther et al. 2002, and Herrero et al. 2002), with
601: luminosities calculated from the photometry in Paper~1 and, in the
602: case of AzV~423, Azzopardi \& Vigneau (1975). Intrinsic colours were
603: taken from Fitzgerald (1970) and, for simplicity, the bolometric
604: corrections were calculated using the relation given by Vacca et
605: al. (1996; their Eqn. 6). Although their study employed atmospheres
606: that did not consider the effects of line blanketing, the lower
607: temperatures derived for O-type stars from its inclusion will
608: yield sensibly smaller bolometric corrections using their relation,
609: sufficiently accurate for our `cosmetic' H-R diagram.
610:
611:
612:
613: \section{Summary}
614:
615: We have compared our STIS observations of early B-type stars in the
616: SMC with Galactic spectra from the $IUE$ archive. The two main
617: morphological features of note are the generally weaker P Cygni
618: emission profiles in the SMC spectra at given spectral type, and the
619: presence of P Cygni profiles in the Galactic spectra to later types
620: than those seen in the SMC. The UV and optical spectra of our most
621: luminous SMC targets were compared, neatly illustrating the correlation
622: between diagnostic features in both regions.
623:
624: We have also begun attempts to calibrate empirically the new 1978
625: index suggested by Rix et al. (2004). Given the relative dearth of
626: high-resolution observations in this region, these data will also
627: enable valuable tests of WM-Basic (Pauldrach et al. 2001) or
628: {\sc cmfgen} (Hillier \& Miller 1998), necessary if model spectra
629: are to be routinely used in population synthesis models.
630:
631: As demonstrated by Figure \ref{hrd}, the full SMC OB-type sample now
632: covers a significant part of the upper H-R diagram, providing an
633: unprecedented library of high-resolution, metal-poor spectra. One
634: issue that we have not addressed at the current time is the wealth of
635: interstellar information in the combined OB-type sample; these data
636: provide a wide variety of sight-lines to the SMC and will form
637: the basis of a future study.
638:
639: \section{Acknowledgements}
640: CJE (under grant PPA/G/S/2001/00131), DJL, and SAR acknowledge
641: financial support from the UK Particle Physics and Astronomy Research
642: Council ({\sc pparc}). NRW acknowledges support through grant
643: GO-9116.09 from the STScI. Based in part on observations with the
644: NASA/ESA {\it Hubble Space Telescope} obtained at the Space Telescope
645: Science Institute (STScI), which is operated by the Association of
646: Universities for Research in Astronomy, Inc., under NASA contract NAS
647: 5-26555; and on INES data from the $IUE$ satellite. We thank Paul
648: Crowther for providing his results in advance of publication,
649: Linda Smith for the red AAT observations, Alex Fullerton for the
650: extracted $FUSE$ spectra, and the referee for their helpful comments.
651:
652: \begin{thebibliography}{100}
653: %
654: \bibitem{av}
655: Azzopardi M. \& Vigneau J., 1975, A\&AS, 22, 285
656: %
657: \bibitem{av2}
658: Azzopardi M. \& Vigneau J., 1982, A\&AS, 50, 291
659: %
660: \bibitem{a87}
661: Azzopardi M., 1987, A\&AS, 69, 421
662: %
663: \bibitem{bc74}
664: Balona L. \& Crampton D., 1974, MNRAS, 166, 203
665: %
666: %\bibitem{b02}
667: %Bresolin F., Kudritzki R.-P., Lennon D. J. et al., 2002, ApJ, 580, 213
668: %
669: \bibitem{b82}
670: Bruhweiler F. C., Parsons S. B., Wray J. D., 1982, ApJ, 256, L49
671: %
672: \bibitem{b03}
673: Bouret J.-C., Lanz T., Hillier D. J. et al., 2003, ApJ, 595, 1182
674: %
675: \bibitem{char93}
676: Charbonnel C., Meynet G., Maeder, A. et al., 1993, A\&AS, 101, 415
677: %
678: \bibitem{c02}
679: Crowther P. A., Hillier D. J., Evans C. J. et al., 2002, ApJ, 579, 774
680: %
681: \bibitem{pac04}
682: Crowther P. A., Walborn N. R., Lennon D. J., 2004, in prep.
683: %
684: \bibitem{d00}
685: De Mello D. F., Leitherer C., Heckman T. M., 2000, ApJ, 530, 251
686: %
687: \bibitem{cjethesis}
688: Evans C. J., 2001, PhD thesis, Univ. of London
689: %
690: \bibitem{vinf}
691: Evans C. J., Lennon D. J., Trundle C. et al., 2004a, ApJ, 607, 451
692: %
693: \bibitem{e04}
694: Evans C. J., Crowther P. A., Fullerton A. W., Hillier D. J., 2004b, ApJ, 610, 1021
695: %
696: \bibitem{f70}
697: Fitzgerald M. P., 1970, A\&A, 4, 234
698: %
699: \bibitem{f90}
700: Fitzpatrick E. L., Garmany C. D., 1990, ApJ, 363, 119
701: %
702: \bibitem{g87}
703: Garmany C. D., Conti P. S., Massey P., 1987, AJ, 93, 1070
704: %
705: %\bibitem{g88}
706: %Garmany C. D., Fitzpatrick E. L., 1988, ApJ, 332, 711
707: %
708: %\bibitem{g57}
709: %Gould N. L., Herbig G. H., Morgan W. W., 1957, PASP, 69, 242
710: %
711: %\bibitem{g89}
712: %Groenewegen M. A. T., Lamers H. J. G. L. M., 1989, A\&AS, 79, 359
713: %
714: %\bibitem{tim}
715: %Harries T. J., Hilditch R. W., Howarth I. D., 2003, MNRAS, 339, 157
716: %
717: %\bibitem{h95}
718: %Haser S. M., 1995, Ph.D. thesis, Univ. of Munich
719: %
720: %\bibitem{hlk95}
721: %Haser S. M., Lennon D. J., Kudritzki R.-P. et al., 1995, A\&A, 295, 136
722: %
723: %\bibitem{h01}
724: %Herrero A., Puls J., Corral L. J. et al., 2001, A\&A, 366, 623
725: %
726: \bibitem{h02}
727: Herrero A., Puls J., Najarro F., 2002, A\&A, 396, 949
728: %
729: \bibitem{hm98}
730: Hillier D. J., Miller D. L., 1998, ApJ, 496, 407
731: %
732: \bibitem{h03}
733: Hillier D. J., Lanz T., Heap S. R. et al., 2003, ApJ, 588. 1039
734: %
735: %%\bibitem{h97}
736: %%Howarth I. D., Siebert K. W., Hussain G. A. J., Prinja R. K., 1997, MNRAS, 284, 265
737: %
738: \bibitem{hm84}
739: Humphreys R. M., McElroy D. B., 1984, ApJ, 284, 565
740: %
741: \bibitem{h82}
742: Hutchings J. B., 1982, ApJ, 255, 70
743: %
744: %\bibitem{j66}
745: %Johnson H. L., 1966, ARA\&A, 4, 193
746: %
747: \bibitem{k87}
748: Kudritzki R.-P., Pauldrach A. W. A., Puls J., 1987, A\&A, 173, 293
749: %
750: \bibitem{k99}
751: Kudritzki R.-P., Puls J., Lennon D. J. et al., 1999, A\&A, 350, 970
752: %
753: %\bibitem{k00}
754: %Kudritzki R.-P., Puls J., 2000, ARA\&A, 38, 613
755: %
756: %\bibitem{k02}
757: %Kudritzki, R.-P., 2002, ApJ, 577, 389
758: %
759: \bibitem{l95}
760: Lamers H. J. G. L. M., Snow T. P., Lindholm D. M., 1995, ApJ, 455, 269
761: %
762: %\bibitem{l92}
763: %
764: %Leitherer C., Robert, C., Drissen, L., 1992, ApJ, 401, 596
765: %
766: \bibitem{l01}
767: Leitherer C., Le$\tilde{\rm a}$o J. R. S., Heckman T. M. et al., 2001, ApJ, 550, 724
768: %
769: \bibitem{star99}
770: Leitherer C., Schaerer D., Goldader J. et al., 1999, ApJS, 123, 3
771: %
772: \bibitem{l97}
773: Lennon D. J., 1997, A\&A, 317, 871
774: %
775: %\bibitem{l99}
776: %Lennon D. J., 1999, RMxAC, 8, 21
777: %
778: \bibitem{l03}
779: Lennon D. J., Dufton P. L., Crowley C., 2003, A\&A, 398, 455
780: %
781: \bibitem{mm01}
782: Maeder A. \& Meynet G., 2001, A\&A, 373, 555
783: %
784: \bibitem{martins}
785: Martins F., Schaerer D., Hillier D. J., 2002, A\&A, 382, 999
786: %
787: \bibitem{m89}
788: Massa D., 1989, A\&A, 224, 131
789: %
790: %\bibitem{massa}
791: %Massa D., Fullerton A. W., Sonneborn G., Hutchings J.B., 2003, ApJ, 586, 996
792: %
793: \bibitem{mpg}
794: Massey P., Parker J. W., Garmany G. D., 1989, AJ, 98, 1305
795: %
796: \bibitem{m02}
797: Massey P., 2002, ApJS, 141, 81
798: %
799: \bibitem{m04}
800: Massey P., Bresolin F., Kudritzki R.-P. et al. 2004, ApJ, 608, 1001
801: %
802: %\bibitem{mcer99}
803: %McErlean N. D., Lennon D. J., Dufton P. L., 1999, A\&A, 349, 553
804: %
805: \bibitem{wmbasic}
806: Pauldrach A. W. A., Hoffmann T. L., Lennon M., 2001, A\&A, 375, 161
807: %
808: \bibitem{p02}
809: Pettini M., Ellison S. L., Bergeron J., Petitjean P., 2002, A\&A, 391, 21
810: %
811: %\bibitem{p87}
812: %Prinja R. K., 1987, MNRAS, 228, 173
813: %
814: \bibitem{p90}
815: Prinja R. K., 1990, MNRAS, 246, 392
816: %
817: %\bibitem{p90}
818: %Prinja R. K., Barlow M. J., Howarth I. D., 1990, ApJ, 361, 607
819: %
820: %\bibitem{p98}
821: %Prinja R. K., Crowther P.A., 1998, MNRAS, 300, 828
822: %
823: %\bibitem{p02}
824: %Prinja R. K., Massa D., Fullerton A. W., 2002, A\&A, 388, 587
825: %
826: %\bibitem{p00}
827: %Puls, J., Springmann, U., Lennon, M., 2000, A\&AS, 141, 23
828: %
829: %\bibitem{p03}
830: %Puls J. \ea~2003.
831: %
832: %\bibitem{r03}
833: %Repolust T., Puls J., Herrero A., 2003, A\&A, in press
834: %
835: \bibitem{rix}
836: Rix S., Pettini M. et al., 2004, ApJ, in press
837: %
838: \bibitem{r74}
839: Robertson J. W., 1974, A\&AS, 15, 261
840: %
841: \bibitem{sk}
842: Sanduleak N., 1968, AJ, 73, 246
843: %
844: %\bibitem{fast}
845: %Santolaya-Rey A. E., Puls J., Herrero A., 1997, A\&A, 323, 488
846: %
847: \bibitem{sch}
848: Schal$\acute{\rm e}$n C., 1975, A\&A, 42, 251
849: %
850: \bibitem{sv85}
851: Shull J. M., Van Steenberg M. E., 1985, ApJ, 294, 599
852: %
853: %\bibitem{kaj}
854: %Siebert K. W., 1999, Ph.D. thesis, Univ. of London
855: %
856: \bibitem{s97}
857: Sterken C., de Groot M., van Genderen A. M., 1997, A\&AS, 326, 640
858: %
859: \bibitem{swings}
860: Swings J. P., Klutz M., Vreux J. M., Peytremann E., 1976, A\&AS, 25, 193
861: %
862: \bibitem{trundle03}
863: Trundle C., Lennon D. J., Puls J., Dufton P. L., 2004, A\&A, 417, 217
864: %
865: \bibitem{t02}
866: Tumlinson J., Shull J. M., Rachford B. L. et al., 2002, ApJ, 566, 857
867: %
868: %\bibitem{u02}
869: %Urbaneja M. A., Herrero A., Kudritzki R.-P. et al., 2002, A\&A, 386, 1019
870: %
871: \bibitem{v96}
872: Vacca W. D., Garmany C. D., Shull M. J., 1996, ApJ, 460, 914
873: %
874: \bibitem{v04}
875: V$\acute{\rm a}$zquez G. A., Leitherer C., Heckman T. M. et al., 2004, ApJ, 600, 162
876: %
877: \bibitem{v99}
878: Venn K., 1999, ApJ, 518, 405
879: %
880: \bibitem{v01}
881: Vink, J. S., de Koter, A., Lamers, H. J. G. L. M., 2001, A\&A, 369, 574
882: %
883: %\bibitem{wnp85}
884: %Walborn N. R., Nicols-Bohlin J., Panek R. J., 1985,
885: %International Ultraviolet Explorer Atlas of O-type Spectra from 1200 to 1900\AA,
886: %(NASA RP1155; Washington; NASA)
887: %
888: \bibitem{w83}
889: Walborn N. R., 1983, ApJ, 265, 716
890: %
891: \bibitem{wn87}
892: Walborn N. R., Nichols-Bohlin J., 1987, PASP, 99, 40
893: %
894: \bibitem{wf00}
895: Walborn N. R., Fitzpatrick, E. L., 1990, PASP, 102, 379
896: %
897: \bibitem{wpn95}
898: Walborn N. R., Parker J. W., Nichols J., 1995,
899: International Ultraviolet Explorer Atlas of B-type Spectra from 1200 to 1900\AA,
900: (NASA RP1363; Washington; NASA)
901: %
902: \bibitem{w95}
903: Walborn N. R., Lennon D. J., Haser S. M. et al., 1995b, PASP, 107, 104
904: %
905: \bibitem{w00}
906: Walborn N. R., Lennon D. J., Heap S. R. et al., 2000, PASP, 112, 1243 (Paper~1)
907: %
908: \bibitem{w02}
909: Walborn N. R., Fullerton A. W., Crowther P. A. et al., 2002, ApJS, 141, 443
910: %
911:
912: \end{thebibliography}
913:
914: %%%%-----TABLES-----
915:
916: \begin{deluxetable}{lllcccccccccccc}
917: \rotate
918: \tabletypesize{\scriptsize}
919: \tablewidth{0pc}
920: \tablecolumns{15}
921: \tablecaption{Observational parameters of GO9116 target stars \label{targets}}
922: \tablehead{
923: \colhead{Star} & \colhead{Alias} & \colhead{Sp. Type} & \colhead{Ref.} & \colhead{$V$} & \colhead{$B-V$} & \colhead{Ref.} & \colhead{$M_{V}$} &
924: \colhead{\Teff} & \colhead{$v_r$} & \multicolumn{2}{c}{Exp. [min]} & log $N$(H~{\tiny I}) & \colhead{Optical} & \colhead{W$_\lambda$(1978)}\\
925: & & & & & & & & \colhead{[kK]} & \colhead{[\kms]} & \colhead{E140M} & \colhead{E230M} & \colhead{[cm$^{-2}$]} & \colhead{Telescope} & \colhead{[\AA]}
926: }
927: \startdata
928: AzV 215 & Sk 76 & BN0 Ia & 2 & 12.69 & $-$0.09 & 4 & $-$6.6 & 27.0 & 159 & 106 & 37.5 & 21.8 & VLT & 5.7 \\
929: AzV 104 & -- & B0.5 Ia & 2 & 13.17 & $-$0.16 & 4 & $-$5.8 & 27.5 & 163 & 106 & 37.5 & 21.4 & VLT & 5.0 \\
930: NGC330-A01 & -- & B0.5 III & 3 & 14.70 & $-$0.18 & 3 & $-$4.5 & 24.0 & 151 & 237 & 145 & 21.5 & ESO 3.6-m & 4.2 \\
931: AzV 216 & -- & B1 III & 5 & 14.22 & $-$0.13 & 4 & $-$5.1 & 26.0 & 203 & 183 & 108 & 21.6 & VLT & 3.1 \\
932: Sk 191 & -- & B1.5 Ia & 2 & 11.86 & $-$0.04 & 1 & $-$7.4 & 22.5 & 130 & 54 & 37.5 & 21.4 & VLT & 15.1 \\
933: AzV 210 & Sk 73 & B1.5 Ia & 2 & 12.60 & $-$0.02 & 4 & $-$6.7 & 20.5 & 173 & 124 & 60.5 & 21.6 & VLT & 16.3 \\
934: AzV 18 & Sk 13 & B2 Ia & 2 & 12.46 & \pp0.03 & 1 & $-$7.0 & 19.0 & 138 & 216 & 75 & 21.8 & VLT & 18.5 \\
935: NGC330-B22 & -- & B2 II & 3 & 14.29 & $-$0.13 & 3 & $-$4.9 & 20.0 & 157 & 308 & 243 & 21.5 & ESO 3.6-m & 4.7 \\
936: AzV 362 & Sk 114 & B3 Ia & 2 & 11.36 & $-$0.03 & 1 & $-$7.8 & 14.0 & 208 & 54 & 37.5 & 21.3 & VLT/NTT & 14.2 \\
937: NGC330-B37 & -- & B3 Ib & 3 & 13.19 & $-$0.07 & 3 & $-$6.0 & 18.0 & 156 & 183 & 108 & \p21.4: & ESO 3.6-m & 13.9 \\
938: NGC330-A02 & -- & B4 Ib & 3 & 12.90 & $-$0.05 & 3 & $-$6.3 & 16.0 & 151 & 237 & 145 & 21.5 & ESO 3.6-m & 9.8 \\
939: AzV 22 & Sk 15 & B5 Ia & 2 & 12.25 & $-$0.10 & 1 & $-$6.6 & 14.5 & 139 & 162 & 75 & 21.5 & VLT/NTT & 11.3 \\
940: \enddata
941: \tablecomments{Stellar identifications are those of
942: Azzopardi \& Vigneau (AzV; 1975, 1982), Robertson (NGC330; 1974) and Sanduleak (Sk; 1968).
943: Absolute magnitudes ($M_v$) are calculated using intrinsic colours
944: from Fitzpatrick \& Garmany (1990) and Fitzgerald (1970), adopting a
945: distance modulus of 18.9 (Harries \ea, 2003) and taking the ratio of
946: total to selective extinction $R = 3.1$ (\eg~Schal$\acute{\rm e}$n,
947: 1975). Effective temperatures and radial velocities are from Lennon et al. (2003) and Trundle et al. (2004).}
948: \tablerefs{(1) Garmany et al. (1987); (2) Lennon (1997); (3) Lennon et al. (2003) -- although in the case of
949: NGC330-A01 the luminosity class has been revised from III/V to III following comparisons of the H$\gamma$
950: equivalent width with the calibrations of Balona \& Crampton (1974); (4) Massey (2002);
951: (5) Trundle et al. (2004)}
952: \end{deluxetable}
953:
954: \clearpage
955:
956: \begin{deluxetable}{llccc}
957: \tabletypesize{\scriptsize}
958: \tablewidth{0pc}
959: \tablecolumns{5}
960: \tablecaption{Physical parameters of the combined SMC sample\label{hrdtab}}
961: \tablehead{
962: \colhead{Star} & \colhead{Spectral Type} & \colhead{\Teff (kK)} & \colhead{log(L/L$_\odot$)} & \colhead{Ref.}
963: }
964: \startdata
965: NGC 346 MPG 355 & ON2 III(f$^\ast$)& 52.5 & 6.04 & 1 \\
966: NGC 346 MPG 324 & O4 V((f)) & 41.5 & 5.52 & 1 \\
967: NGC 346 MPG 368 & O4-5 V((f)) & 40.0 & 5.41 & 1 \\
968: %NGC 346 MPG 435 & O4-6 III(f) &&&\\
969: AzV 80 & O4-6n(f)p & {\it 40.5} & {\it 5.9} & -- \\
970: %AzV 75 & O5 III(f$+$) & {\it 40.5} & {\it 6.1} & -- \\
971: AzV 75 & O5 III(f$+$) & 40.0 & 6.2 & 5 \\
972: NGC 346 MPG 113 & OC6 Vz & 40.0 & 5.15 & 1 \\
973: AzV 220 & O6.5 f?p & {\it 37.3} & {\it 5.2} & -- \\
974: AzV 15 & O6.5 II(f) & {\it 35.0} & {\it 5.7} & -- \\
975: AzV 95 & O7 III((f)) & {\it 34.7} & {\it 5.3} & -- \\
976: AzV 83 & O7 Iaf$+$ & 32.8 & 5.54 & 2 \\
977: AzV 69 & OC7.5 III((f)) & 33.9 & 5.61 & 2 \\
978: AzV 47 & O8 III((f)) & {\it 32.6} & {\it 5.5} & -- \\
979: NGC 346 MPG 487 & O8 V & 35.0 & 5.15 & 1 \\
980: %NGC 346 MPG 682 & O9-9.5 V &&&\\
981: AzV 327 & O9.5 II-Ibw & {\it 28.4} & {\it 5.4} & -- \\
982: AzV 423 & O9.5 II(n) & {\it 28.4} & {\it 5.4} & -- \\
983: NGC 346 MPG 12 & O9.5-B0 V (N str)& 31.0 & 4.93 & 1 \\
984: AzV 170 & O9.7 III & {\it 30.0} & {\it 5.1} & --\\
985: AzV 215 & BN0 Ia & 27.0 & 5.63 & 4 \\
986: AzV 104 & B0.5 Ia & 27.5 & 5.31 & 4 \\
987: NGC330-A01 & B0.5 III/V & 24.0 & 4.67 & 3 \\
988: AzV 216 & B1 III & 26.0 & 5.00 & 4 \\
989: Sk 191 & B1.5 Ia & 22.5 & 5.77 & 4 \\
990: AzV 210 & B1.5 Ia & 20.5 & 5.41 & 4 \\
991: AzV 18 & B2 Ia & 19.0 & 5.44 & 4 \\
992: NGC330-B22 & B2 II & 20.0 & 4.65 & 3 \\
993: AzV 362 & B3 Ia & 14.0 & 5.50 & 4 \\
994: NGC330-B37 & B3 Ib & 18.0 & 4.98 & 3 \\
995: NGC330-A02 & B4 Ib & 16.0 & 4.98 & 3 \\
996: AzV 22 & B5 Ia & 14.5 & 5.04 & 4 \\
997: \enddata
998: \tablerefs{(1) Bouret et al. (2003); (2) Hillier et al. (2003); (3) Lennon et al. (2003);
999: (4) Trundle et al. (2004); (5) Massey et al. (2004)}
1000: \tablecomments{Italicized values denote values interpolated from published results, see
1001: text for details}
1002: \end{deluxetable}
1003: \clearpage
1004:
1005:
1006: %%%%-----FIGURES----
1007:
1008:
1009: \begin{figure*}
1010: \begin{center}
1011: \includegraphics{fig1.eps}
1012: \caption{STIS spectra of AzV~423 and NGC~346 MPG~487 from {\it HST} program GO7437.
1013: The spectrum of AzV~327 was published previously by Walborn et al. (2000) and is
1014: included here for comparison with AzV~423 (see discussion in Section \ref{7437}).}
1015: \label{extras}
1016: \end{center}
1017: \end{figure*}
1018:
1019: \clearpage
1020:
1021: \begin{figure*}
1022: \begin{center}
1023: \includegraphics[width=22cm, height=16cm, angle=-270]{fig2b.ps}
1024: \caption{\footnotesize STIS spectra of class Ia SMC stars. The lines
1025: identified in AzV~215 are, from left to right, N~\5 \lam\lam1239,
1026: 1243; Si~\4 \lam\lam1394, 1403; and C~\4 \lam\lam1548, 1551.
1027: Additional lines marked in Sk~191 are, C~\2 \lam\lam1334, 1336; and
1028: Al~\3 \lam\lam1855, 1863; and in AzV~18, Si~\2 \lam1265 and Si~\3
1029: \lam\lam1294, 1299, 1417. Also note the iron `forests' from Fe~\4 in
1030: AzV~215 (\lam1550-1750) and Fe~\3 (\lam1850-2100) in the other
1031: spectra. Each spectrum is offset by two continuum flux units.}
1032: \label{fig1}
1033: \end{center}
1034: \end{figure*}
1035:
1036: \clearpage
1037:
1038: \begin{figure*}
1039: \begin{center}
1040: \includegraphics[width=15.7cm, height=16cm, angle=-270]{fig3b.ps}
1041: \caption{$IUE$ spectra of Galactic analogues to the SMC stars shown in
1042: Figure \ref{fig1}; the line identifications are the same.}
1043: \label{fig2}
1044: \end{center}
1045: \end{figure*}
1046:
1047: \clearpage
1048:
1049: \begin{figure*}
1050: \begin{center}
1051: \includegraphics[width=22cm, height=16cm, angle=-270]{fig4b.ps}
1052: \caption{STIS spectra of additional B-type SMC targets.}
1053: \label{fig3}
1054: \end{center}
1055: \end{figure*}
1056:
1057: \clearpage
1058:
1059: \begin{figure*}
1060: \begin{center}
1061: \includegraphics[width=15.7cm, height=16cm, angle=-270]{fig5b.ps}
1062: \caption{$IUE$ spectra of Galactic analogues to
1063: those in Figure \ref{fig3}; the luminosity type of HD~213087 is
1064: deliberately mis-matched to that of AzV~104 (see text for discussion).}
1065: \label{fig4}
1066: \end{center}
1067: \end{figure*}
1068:
1069: \clearpage
1070:
1071:
1072: \begin{figure*}
1073: %\begin{center}
1074: \includegraphics[width=22cm, height=16cm, angle=-270]{fig6b.ps}
1075: \caption{\footnotesize Optical spectra of class Ia SMC stars.
1076: Lines identified in the AzV~215 spectrum are, from left to right by
1077: species, He~\1 \lam\lam4009, 4026, 4121, 4144, 4388, 4471, 4713; and
1078: the N~\3 \lam4640 and C~\3 \lam4650 blends. Lines identified in
1079: AzV~210 are N~\2 \lam3995; O~\2 \lam\lam4072-4476, 4415-4417; Mg~\2
1080: \lam4481; and Si~\3 \lam\lam4553, 4568, 4575. Lines
1081: marked in AzV~22 are Si~\2 \lam4128-32; and the Balmer H$\alpha$ line
1082: (\lam6563). For display purposes the data have been smoothed to an
1083: effective resolution of 1.0 \AA~and each spectrum is offset by one
1084: continuum flux unit.}
1085: \label{opt}
1086: \end{figure*}
1087:
1088:
1089: \clearpage
1090:
1091: \begin{figure*}
1092: \begin{center}
1093: \includegraphics[scale=1.0, angle=0]{fig7.eps}
1094: %\includegraphics[width=18cm, height=18cm, angle=0]{hrd.eps}
1095: \caption{H-R diagram of the full {\it HST} STIS sample. Evolutionary tracks ({\it dotted lines})
1096: are from Charbonnel et al. (1993) and are labelled with initial stellar masses
1097: in $M_\odot$. Stars marked with open symbols are those for which
1098: physical parameters have been interpolated from published results (see
1099: text for details).}
1100: \label{hrd}
1101: \end{center}
1102: \end{figure*}
1103:
1104:
1105: \end{document}
1106: