1:
2:
3: \documentclass[12pt,preprint]{aastex}
4:
5:
6:
7:
8:
9: \newcommand {\asp}{\mbox{$.\!\!^{\prime\prime}$}}
10: \newcommand{\vdag}{(v)^\dagger}
11: \newcommand{\myemail}{pmcc@stardial.com}
12: \newcommand{\irassrc}{{\it IRAS} 06337+1051}
13:
14: \newcommand{\IRAS}{{\it IRAS}}
15: \newcommand{\COBE}{{\it COBE}}
16: \newcommand{\MAP}{{\it MAP}}
17:
18: \newcommand{\LPH}{LPH~201.663+1.643}
19: \newcommand{\HI}{H\,{\scriptsize I}}
20: \newcommand{\HII}{H\,{\scriptsize II}}
21: \newcommand{\halpha}{H$\alpha$}
22: \newcommand{\brg}{Br$\gamma$}
23: \newcommand{\MJysr}{MJy~sr$^{-1}$}
24: \newcommand{\etal}{{\it et al.}~}
25:
26:
27:
28: \shorttitle{\LPH}
29: \shortauthors{McCullough and Chen}
30:
31:
32: \begin{document}
33:
34: \title{An Alternative to Spinning Dust for the Microwave Emission of \LPH:
35: an Ultracompact \HII\ Region}
36:
37: \author{P. R. McCullough\altaffilmark{1} and R. R. Chen\altaffilmark{}}
38: \affil{Dept. of Astronomy, Univ. of Illinois, Urbana IL 61801, USA}
39: \altaffiltext{1}{Cottrell Scholar of Research Corporation.}
40: \email{pmcc@astro.uiuc.edu, raychen@astro.uiuc.edu}
41:
42:
43: \begin{abstract}
44:
45: The microwave spectral energy distribution of the dusty, diffuse \HII\ region
46: \LPH\ has been interpreted by others as tentative evidence for
47: microwave emission from spinning dust grains. We present an alternative
48: interpretation for that particular object; specifically, that an ultracompact
49: \HII\ region embedded within the dust cloud would explain the available
50: observations as well or better than spinning dust.
51: Parameters for the size, surface brightness, and flux density of the putative
52: ultracompact \HII~region,
53: derived from the microwave observations, are within known ranges.
54: A possible candidate for such an ultracompact \HII\ region is
55: \irassrc, based upon its infrared colors. However, \irassrc's infrared flux
56: appears to be too small to be consistent with the microwave flux required
57: for this alternative model to explain the observations.
58:
59: \end{abstract}
60:
61:
62: \keywords{
63: cosmic microwave background ---
64: diffuse radiation ---
65: dust, extinction ---
66: ISM: clouds ---
67: \HII\ regions ---
68: radiation mechanisms: thermal ---
69: radio continuum: ISM
70: }
71:
72:
73:
74:
75: \section{Introduction} \label{sec_intro}
76:
77: Studies of the cosmic microwave background are affected by foreground
78: emissions from the Milky Way and other sources. Principal mechanisms
79: are thermal emission from warm dust, synchrotron from electrons gyrating
80: in magnetic fields, and free-free (a.k.a. bremsstrahlung) emission from
81: ionized plasma. An elusive forth foreground is attributed to spinning dust
82: \citep[]{draine98a} or perhaps magnetic dust \citep[]{draine99}.
83: A small but growing number of statistical comparisons
84: of infrared tracers of dust and microwave observations of ``CMB quality''
85: have demonstrated repeatedly, but in each case at low significance,
86: an excess of microwave emission, correlated with dust on the sky,
87: and possibly due to spinning or magnetized dust.
88: Finkbeiner {\it et al.} (2002, Paper I)
89: list such statistical comparisons and add to them two tentative
90: detections of the spinning-dust mechanism in specific astronomical sources,
91: \LPH\ and Lynds 1622, from a sample of 10 sources observed by them.
92:
93: \LPH\ is the topic of this {\it Letter}. It differs from the other nine dust
94: clouds observed by Finkbeiner {\it et al.} (2002) in that it is known to be
95: a diffuse \HII\ region whereas the others are not ionized.
96: Finkbeiner {\it et al.}
97: included \LPH\ in their sample despite their prejudice that free-free
98: emission might overwhelm any emission from spinning dust.
99: Alternatively, one might anticipate that ionized regions would be good
100: places to find microwave emission from spinning dust, because in those regions,
101: ion collisions with grains are expected to be
102: the largest contributory factor by far in maintaining the spin of the grains
103: \citep[]{draine98b}.
104:
105: Because of its special status as the sole \HII\ region in the list, and
106: because \LPH\ provided the only very significant detection and was indeed
107: much brighter than the spinning-dust theory would have predicted, we
108: thought that a different explanation might exist.
109: The alternative model for \LPH\ that we present in this {\it Letter}
110: is the superposition of one source with very large emission measure
111: and very small angular scale with a low emission-measure, extended source.
112: The combination of a small, optically thick source and a large optically thin
113: source, both emitting by free-free, can create the rising microwave spectrum
114: observed in Paper I.
115:
116:
117: \section{Observations} \label{sec_obs}
118:
119: The three observations from Table 3 of Paper I relevant to this {\it Letter} are
120: $\rm I_\nu = 0.397 \pm 0.046,\ 0.516 \pm 0.019,\ and\ 0.667 \pm 0.027$\
121: \MJysr\ at $\nu =$ 5 GHz, 8.25 GHz, and 9.75 GHz. Note that the brightness
122: increases with frequency.
123:
124: The envelopes of newly forming stars in close proximity to an O-type star
125: can have emission measures as large as $\rm 3\times 10^8~cm^{-6}~pc$
126: over angular scales $^{<}_{\sim} 1$\arcsec\ (McCullough \etal\ 1995).
127: An ensemble of externally ionized stellar envelopes
128: could match the angular extent of the dust cloud, but the combined flux
129: from the ensemble would be insufficient to match the
130: observed microwave flux from \LPH\, which at 9.75 GHz is $\sim 3$ Jy.
131: For example, the total 15 GHz flux density from the the ensemble of 24 ionized
132: envelopes illuminated by $\theta^1$C Orion is 0.175 Jy (Felli \etal\ 1993).
133: A better idea would be an
134: ultracompact \HII\ region, which would intercept all of the UV flux
135: from an OB star, whereas an ensemble of stellar envelopes intercepts only a tiny
136: fraction of the OB star's flux.
137:
138:
139: In order to estimate the distance to \LPH\,
140: we searched its vicinity in the SIMBAD database for hot stars
141: that could ionize it and found two,
142: GSC 00737-01170 (O5:; V=11.29; B-V=0.75) and GSC 00737-00898 (O5; V=10.27;
143: B-V=1.11), where the spectral types and magnitudes are from
144: Voroshilov \etal\ (1985).
145: For an O5 V star, $\rm M_V = -5.3$ and for an O5 Ia star, $\rm M_V = -6.4$
146: (Vacca, Garmany, \& Shull 1996). Regardless of luminosity class, an unreddened
147: O5 star has B-V = $-0.33$ (Cox 2000).
148: For $A_V = 3.2 E(B-V)$, the implied distances to
149: GSC 00737-01170 and GSC 00737-00898 are 4.2 kpc and 1.6 kpc, respectively
150: if each is O5 V, and 7.0 kpc and 2.6 kpc, respectively
151: if each is O5 Ia. At $l=201.6$\arcdeg\ and $b=+1.6$\arcdeg, \LPH\ is
152: $\sim$22\arcdeg\ from the anticenter direction, and its Galactic height
153: $Z < 200$ pc for d $< 7.0$ kpc.
154:
155:
156:
157: \section{Modeling} \label{sec_modeling}
158:
159: The optical depth for free-free emission,
160: \begin{equation}
161: \rm \tau_{ff} \approx 2.6~T_4^{-1.35}~(EM/10^9~cm^{-6}~pc)~(\nu/10~GHz)^{-2.1},
162: \label{eq-tauff}
163: \end{equation}
164: where
165: $\rm T_4 \equiv T_e / 10^4 K$,
166: $\rm T_e$ is the electron temperature,
167: $\rm EM$ is the emission measure, and
168: $\nu$ is the radio frequency.
169: As observed with an antenna of beam solid angle $\Omega$,
170: the apparent microwave surface brightness
171: of a uniform disk of radius $\theta_s$ is
172: \begin{equation}
173: \rm I_1 = B_\nu(T_e) ( 1 - e^{-\tau_{ff}} ) \pi \theta_s^2 \Omega^{-1}
174: \label{eq-i1}
175: \end{equation}
176: where $\rm B_\nu(T_e)$ is the Planck function and we assume that
177: the background is negligible and that $\pi \theta_s^2 << \Omega$.
178: For the 140-foot antenna, $\Omega = 4.59, 1.75,$ and 1.31
179: microsteradians for observing frequencies of 5 GHz, 8.25 GHz, and 9.75 GHz
180: respectively, but in our model we use $\Omega = 4.59\times 10^{-6}$ sr
181: for all frequencies because in Paper I the higher frequency data were smoothed
182: to match the 5 GHz beam.
183:
184: To get the total emission of the model, we add $\rm I_1$ to the
185: free-free emission from a diffuse, optically thin component.
186: The diffuse component cannot be very diffuse, because
187: if it was larger than the 0.2\arcdeg\ chop used in Paper I
188: at 8.25 and 9.75 GHz, it would be
189: eliminated by the chopping. Although we do not know the angular distribution
190: of free-free emission in this region, we assume that the diffuse free-free
191: component contributes to the brightnesses at each
192: frequency equally except for the $\nu^{-0.1}$ spectral index of
193: optically thin free-free. Thus, our model is
194: \begin{equation}
195: \rm I_\nu = I_2 (\nu/5~GHz)^{-0.1} + B_\nu(T_e) ( 1 - e^{-\tau_{ff}} ) \pi \theta_s^2 \Omega^{-1},
196: \label{eq-inu}
197: \end{equation}
198: where the model parameters are $\rm I_2$
199: and the two parameters of the ultracompact \HII\ region (its radius
200: $\theta_s$ and its emission measure, EM).
201:
202: Fig. \ref{fig_spectrum} compares the observed brightnesses
203: to those predicted by the model
204: with a few combinations of parameters for illustration.
205: Fig. \ref{fig_contours} illustrates $\chi^2$ contours in the phase space of the parameters
206: of the ultracompact \HII\ region, $\rm \theta_s\ and\ EM$,
207: with $\rm I_2$ fixed at 0.255 \MJysr.
208: The allowed parameters are similar to those of known
209: ultracompact \HII\ regions (Wood \& Churchwell 1989b).
210: For $\rm EM = 10^9$ cm$^{-6}$~pc, $\theta_s = 1$\arcsec,
211: the number of ionizing photons required to power the ultracompact
212: \HII\ region,
213: $N_c^\prime~^{>}_{\sim} 4\times10^{47} (d/1~{\rm kpc})^2 {\rm s}^{-1}$
214: (Kurtz, Churchwell, \& Wood 1994),
215: which corresponds to a main-sequence spectral type B0 or earlier
216: for $d = 1.6$ kpc
217: or O8 or earlier for $d = 4.2$ kpc (Vacca, Garmany, \& Shull 1996).
218:
219: \section{A Possible Candidate: \irassrc} \label{sec_candidate}
220:
221: \irassrc\ stands out in the \IRAS\ ISSA images as the brightest 25 $\mu$m
222: source in the vicinity of \LPH.
223: \irassrc\ has the infrared colors of an ultracompact \HII\ region (Fig.
224: \ref{fig_wood1b}).
225: Its PSC position (Beichman \etal\ 1988),
226: $\rm 6^h36^m29^s.47~+10\arcdeg49\arcmin5\arcsec.1$ [2000],
227: with an error ellipse $38\arcsec \times 7\arcsec$\
228: oriented with the major axis at a position angle of 94\arcdeg,
229: is compatible with the Paper I's linear scans, which were
230: centered on $\rm 6^h36^m40^s~+10\arcdeg46\arcmin28\arcsec$
231: and oriented at position angle 292.5\arcdeg.
232: The nominal scan's closest approach to the PSC position of \irassrc\
233: is 1.4\arcmin,
234: and the peak along the nominal 8.25 GHz scan occurs 2.4\arcmin\ or
235: 0.6\arcmin\ east of the nominal position of \irassrc, for the forward
236: and reverse scan directions respectively (Paper I, Fig. 3b).
237: For comparison, the resolution (FWHM) of the microwave data was 6\arcmin.
238: The antenna temperature at 8.25 GHz of the
239: ultracompact \HII\ region is $^{<}_{\sim} 55$\% of the total for the
240: models in Section \ref{sec_modeling},
241: so the peak position at 8.25 GHz need not be centered
242: at the position of the ultracompact \HII\ region.
243:
244:
245: A potential flaw in \irassrc\ as the putative ultracompact \HII\ region
246: is that our model requires parameters for the
247: ultracompact \HII\ region that would make it brighter than most ultracompact
248: \HII\ regions at radio wavelengths (cf. Wood \& Churchwell 1989a and
249: Kurtz, Churchwell, \& Wood 1994) but at $60\mu$m
250: \irassrc\ is four times fainter than
251: the median of embedded OB star candidates selected by Wood \& Churchwell
252: (1989a).
253: For the models depicted in
254: Fig. \ref{fig_contours}, the ultracompact \HII\ regions' flux density,
255: $F_\nu(2 cm)$ ranges from 1.6 Jy to 4.4 Jy.
256: Ultracompact \HII\ regions are more than a thousand times brighter
257: in the infrared than in the radio (Kurtz, Churchwell, \& Wood 1994),
258: so the $3\sigma$ upper limit to the \IRAS\ PSC flux at 100 $\mu$m,
259: $F_\nu({\rm 100\mu m}) < 383$ Jy would imply that the radio flux density
260: $F_\nu(2 cm) < 0.4 $ Jy, i.e. in conflict with the model fluxes.
261: Stated another way, the luminosity is
262: \begin{equation}
263: L_* = 4 \pi d^2 S_\nu \Delta\nu f_\nu^{-1},
264: \label{eq-lum}
265: \end{equation}
266: where $f_\nu$ is the fraction of total luminosity $L_*$ emitted
267: at frequency $\nu$ in the passband $\Delta\nu$, and $S_\nu$ is
268: the associated flux density.
269: For \irassrc\ at 60 $\mu$m, $S_\nu = 33$ Jy,
270: $\Delta\nu = 2.58\times10^{12}$ Hz (Beichman et al. 1988),
271: and $f_{\rm 60 \mu m} = 0.25$ for ultracompact \HII\ regions
272: (Wood \& Churchwell 1989a), so $L_* = 100 (d /1 kpc)^2 L_\odot$.
273: Because ionizing OB stars have $L_* > 10^4 L_\odot$,
274: the implied distance $d > 10$ kpc.
275: At such large distances from the Galactic center,
276: massive stars do exist but are rare (de Geus \etal\ 1993).
277: However, if the distance to \irassrc\ $d > 10$ kpc, then it is too
278: far behind \LPH\ for them to be
279: physically associated, and their alignment would be highly improbable
280: {\it a priori}.
281:
282: \section{Conclusions} \label{sec_dis}
283:
284: An optically thick ultracompact \HII\ region can provide a
285: rising microwave spectrum
286: with a spectral index $\approx 2$ which is similar at these
287: frequencies to the spinning-dust model's spectral index $\approx 2.8$
288: (Ferrara \& Dettmar 1994; Draine \& Lazarian 1998b).
289: For free-free emission,
290: for $\tau_{ff} \ ^{>}_{\sim} 1$ at $\nu ^{<}_{\sim} 10$ GHz, the
291: EM must be $^{>}_{\sim} 10^9$ cm$^{-6}$~pc.
292:
293: While \irassrc\ is within a fraction of a 6\arcmin\ beam width of \LPH,
294: and has the infrared colors
295: of an ultracompact \HII\ region, its \IRAS\ fluxes are too small unless
296: its distance is improbably large. Also, the distance-independent ratio of
297: the 100 $\mu$m flux of \irassrc\ to our model's 2-cm flux is too small by
298: more than an order of magnitude.
299: The latter conflict is mitigated somewhat if the star is an early B-type star
300: that has an atypically large ratio of ionizing to bolometric luminosity
301: as has been observed in some cases (Cassinelli 1996) and has been predicted by
302: modern stellar atmosphere models (Vacca, Garmany, \& Shull 1996).
303:
304: Paper I concluded that ``as long as the density is low,
305: as in the LPH list, \HII\ regions may be optimal targets for future work for
306: DASI, CBI, and \MAP.''
307: It would be inappropriate to misinterpret this single case study of \LPH\ as
308: contradictory to that recommendation.
309: Instead, attention to its premise, that the density be low,
310: is especially appropriate. The all-sky observations made by \MAP\ will allow
311: astronomers to conveniently avoid any regions that might be unsatisfactory
312: in that regard.
313:
314: The hypothesis presented in this {\it Letter} can be tested directly with radio
315: interferometric imaging, which we plan to complete in early 2002.
316: If that test shows not even one dense ionized knot in
317: \LPH, then this {\it Letter}'s remaining value
318: will be the analysis of Section \ref{sec_modeling}, and the
319: microwave emission from \LPH\ will still be a mystery.
320:
321: \acknowledgments
322:
323: We thank Doug Finkbeiner for sending us a pre-publication manuscript
324: that initiated our thoughts about \LPH. A conversation with Bruce Draine
325: during his visit to Illinois on October 30, 2001, was especially
326: helpful; as referee, B. D. made additional valuable comments,
327: most importantly that \irassrc\ be in its own section.
328:
329: P. R. M. wishes to acknowledge two grants that funded this research,
330: a CAREER AWARD from the National Science Foundation (AST-9874670) and a Cottrell
331: Scholarship from the Research Corporation,
332: and to thank the persons that championed the associated proposals.
333:
334:
335:
336: \begin{thebibliography}{}
337:
338: \bibitem[Cox(2000)]{2000asqu.book.....C} Cox, A.~N.\ 2000, Allen's astrophysical quantities, 4th ed.~ Publisher: New York: AIP Press; Springer, 2000.~ Edited by Arthur N.~Cox.~ ISBN: 0387987460,
339:
340: \bibitem[Beichman et al.(1988)]{iras88} Beichman, C.~A., Neugebauer, G., Habing, H.~J., Clegg, P.~E., \& Chester, T.~J.\ 1988, NASA Reference Publication, 1190, 1
341: \bibitem[Cassinelli(1996)]{1996aeu..conf..367C} Cassinelli, J.~P.\ 1996,
342: IAU Colloq.~152: Astrophysics in the Extreme Ultraviolet, 367
343:
344: \bibitem[de Geus, Vogel, Digel, \& Gruendl(1993)]{1993ApJ...413L..97D} de Geus, E.~J., Vogel, S.~N., Digel, S.~W., \& Gruendl, R.~A.\ 1993, \apjl, 413, L97
345:
346: \bibitem[Draine \& Lazarian 1998a]{draine98a}
347: Draine, B. T., \& Lazarian, A. 1998a, \apj, 494, L19
348:
349: \bibitem[Draine \& Lazarian 1998b]{draine98b}
350: Draine, B. T., \& Lazarian, A. 1998b, \apj, 508, 157
351:
352: \bibitem[Draine \& Lazarian 1999]{draine99}
353: Draine, B. T., \& Lazarian, A. 1999, \apj, 512, 740
354:
355: \bibitem[Felli et al.(1993)]{1993A&AS..101..127F} Felli, M., Taylor, G.~B., Catarzi, M., Churchwell, E., \& Kurtz, S.\ 1993, \aaps, 101, 127
356:
357: \bibitem[Ferrara \& Dettmar 1994]{ferrara94}
358: Ferrara, A., \& Dettmar, R. -J. 1994, \apj, 427, 155
359:
360: \bibitem[Finkbeiner \etal\ 2002]{finkbeiner01}
361: Finkbeiner, D. P., Schlegel, D. J., Frank, C., \& Heiles, C. 2002, \apj, in press
362:
363: %\bibitem[Gaustad, J. E. \etal\ (2001)]{gaustad01}
364: %Gaustad, J. E., McCullough, P. R., Rosing, W., \& Van Buren, D. 2001,
365: %PASP, 113, 1326
366:
367:
368: \bibitem[Kurtz, Churchwell, \& Wood(1994)]{1994ApJS...91..659K} Kurtz, S.,
369: Churchwell, E., \& Wood, D.~O.~S.\ 1994, \apjs, 91, 659
370:
371:
372: \bibitem[McCullough et al.(1995)]{1995ApJ...438..394M} McCullough, P.~R.,
373: Fugate, R.~Q., Christou, J.~C., Ellerbroek, B.~L., Higgins, C.~H.,
374: Spinhirne, J.~M., Cleis, R.~A., \& Moroney, J.~F.\ 1995, \apj, 438, 394
375:
376:
377: \bibitem[Wood \& Churchwell(1989a)]{1989ApJ...340..265W} Wood, D.~O.~S.~\&
378: Churchwell, E.\ 1989a, \apj, 340, 265
379:
380: \bibitem[Wood \& Churchwell(1989b)]{1989ApJS...69..831W} Wood, D.~O.~S.~\&
381: Churchwell, E.\ 1989b, \apjs, 69, 831
382:
383: \bibitem[Vacca, Garmany, \& Shull(1996)]{1996ApJ...460..914V} Vacca, W.~D., Garmany, C.~D., \& Shull, J.~M.\ 1996, \apj, 460, 914
384:
385: \bibitem[Voroshilov et al.(1985)]{1985cbvm.book.....V} Voroshilov, V.~I., Guseva, N.~G., Kalandadze, N.~B., Kolesnik, L.~N., \& Kuznetsov, V.~I.\ 1985, Kiev, Izdatel'stvo Naukova Dumka, 1985, 140 p.~In Russian.
386:
387: \end{thebibliography}
388:
389: \clearpage
390:
391:
392: \figcaption{\label{fig_spectrum}
393: The brightness of \LPH\ at three microwave frequencies (Paper I)
394: is compared to that of combined emission from a diffuse \HII\ region
395: and an ultracompact \HII\ region (Equation \ref{eq-inu}) for various
396: model parameters (solid lines). For models A, B, C, and D respectively, with
397: $\rm \theta_s\ and\ EM$\ as indicated in Fig. \protect{\ref{fig_contours}},
398: and $\rm I_2 = 0.255$ \MJysr\ in all four models,
399: the $\rm \chi^2 \approx 2, 4, 10, and\ 20$.
400: The dashed line shows model A with the additional contribution from
401: spinning dust appropriate for a warm ionized medium (Ferrara and Dettmar 1994;
402: Draine \& Lazarian 1998b)
403: and a dust column $\rm N(H) = 2.7 \times 10^{22} cm^{-2}$ (Paper I).
404: }
405:
406: \figcaption{\label{fig_contours}
407: Contours of $\chi^2$ are plotted in the phase space of the parameters
408: of the ultracompact \HII\ region, $\theta_s$ and $\rm EM$,
409: for a representative value for the parameter that describes
410: the diffuse \HII\ region $\rm I_2$ = 0.255 \MJysr.
411: For values of EM $^{>}_{\sim} 10^9$
412: cm$^{-6}$~pc, the ultracompact \HII\ region is optically thick at even
413: the highest observed frequency, so
414: the $\chi^2$ contours are open at the top of the plot.
415: Models with
416: $\rm \theta_s\ and\ EM$\ as indicated by the letters A, B, C, and D
417: are compared to the data in Fig. \protect{\ref{fig_spectrum}}.
418: }
419:
420: \figcaption{\label{fig_wood1b}
421: \irassrc~is compared to known ultracompact \HII~regions ({\it filled circles})
422: in an \IRAS\ color-color plot from Wood \& Churchwell (1989a).
423: Two solid-lined rectangles are centered on the most-likely colors of
424: \irassrc,
425: ${\rm Log(F_\nu(25 \mu m) / F_\nu(12 \mu m)) = 0.96 \pm 0.049}$, and
426: ${\rm Log(F_\nu(60 \mu m) / F_\nu(12 \mu m)) = 1.59 \pm 0.094}$.
427: The smaller rectangle has a $1\sigma$ half-width and half-height;
428: the other rectangle is three times larger.
429: That the rectangles lie in the region of the diagram populated by
430: ultracompact \HII~regions and that other types of sources very rarely
431: populate that region make \irassrc\ likely to be an ultracompact \HII\ region.
432: The dashed lines indicate
433: the boundary prescribed by Wood \& Churchwell (1989a) to discriminate
434: embedded OB stars from other sources in the \IRAS\ point source catalog.
435: Other representative sources are those within a $2\arcdeg \times 2\arcdeg$\
436: box in the Galactic plane ({\it open squares}) and a group of \IRAS\ sources
437: that lie between ${\rm 13^h00^m\ and\ 13^h10^m}$ in right ascension
438: ({\it crosses}).
439: }
440:
441: \begin{figure}
442: \plotone{f1.ps}
443: \end{figure}
444:
445: \begin{figure}
446: \plotone{f2.ps}
447: \end{figure}
448:
449: \begin{figure}
450: \plotone{f3.eps}
451: \end{figure}
452:
453: \end{document}
454:
455: