Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

RevMexAA (Serie de Conferen ias), 9, 238{245 (2000)THE WISCONSIN H� MAPPER:A NEW LOOK AT THE WARM IONIZED MEDIUML. M. Ha�ner, R. J. ReynoldsDepartment of Astronomy, University of Wis onsin{Madison, USAandS. L. TufteDepartment of Physi s, Lewis & Clark College, USARESUMENEl relevamiento en H-Alpha de Wis onsin (WHAM) ha ubierto todo el ielodel norte en H� desde Kitt Peak, Arizona. Usando un espe tr�ometro Fabry-Perotde alta transmisividad, on etal�on doble de 15 m y un CCD de alta sensibilidad,el relevamiento WHAM ha obtenido el primer mapa on velo idades alibradas enH� en emisi�on de nuestra Galaxia. Una zona amplia, que in luye el brazo Lo al(Ori�on) y el brazo de Perseo, tambi�en ha sido observada por el WHAM en las l��neasde [S II℄ y [N II℄. Estos nuevos datos muestrean dire tamente y por primera vez,las ondi iones f��si as a gran es ala del medio ionizado tibio (WIM). Las tenden- ias de los o ientes de l��neas sugieren que las varia iones en la temperatura sonmuestreadas por los mapas de [N II℄/H� y [S II℄/H�. Dado que estos o ientes sein rementan fuertemente en regiones alejadas del plano Gal�a ti o, revelan un in re-mento sustan ial en la temperatura del WIM del halo. Adem�as de este resultado,los datos revelan una nueva region H II ex itada por una estrella B, dan una medidade la es ala de altura de los ele trones del WIM y, del o iente [S II℄/[N II℄, proveennueva informa i�on sobre la ioniza i�on del WIM.ABSTRACTThe Wis onsin H-Alpha Mapper (WHAM) has surveyed the entire northernsky in H� from Kitt Peak, Arizona. Using a high-throughput, 15- m diameterdouble-etalon Fabry-Perot spe trometer and a sensitive CCD dete tor, the WHAMsurvey provides the �rst alibrated, velo ity-resolved map of H� emission in ourGalaxy. A large portion of the Galaxy, whi h samples regions of the Lo al (Orion)spiral arm and the more distant Perseus arm, has been also been observed withthe WHAM in lines of [S II℄ and [N II℄. These new data dire tly probe the physi al onditions of the Warm Ionized Medium (WIM) on a global s ale for the �rsttime. Trends in these line ratios over this large region of the sky suggest thattemperature variations are tra ed by the [N II℄/H� and [S II℄/H� maps. Sin e theseratios in rease dramati ally away from the Gala ti plane, they reveal a substantialtemperature rise in WIM halo gas. In addition to this striking new result, thedata set also reveals new information about the ionization in the WIM throughthe [S II℄/[N II℄ ratio, un overs a previously undis overed B-star H II region, andprovides an a urate measurement of the ele tron s ale height of the WIM.Key Words: GALAXY: HALO | H II REGIONS | ISM: ATOMS |ISM: STRUCTURE 238

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

THE WISCONSIN H� MAPPER 2391. INTRODUCTIONWith a s ale height of 1 kp , mass surfa e density equal to one-third that of H I, and power input requirementequal to the supernovae output rate of our Galaxy (Reynolds 1993), the Warm Ionized Medium (WIM) isjusti�ably a major omponent of the Interstellar Medium (ISM). Although, originally dis overed in the 1960sthrough radio observations (Hoyle & Ellis 1963; Gu�elin 1974), the WIM is typi ally studied through opti alemission lines, parti ularly H�, [S II℄ �6716, and [N II℄ �6583. Chara terizing the details of the WIM isimportant not only for improving the pi ture of the ISM and the intera tion among its omponents, butalso for understanding the nature of the foreground gas through whi h all extra-gala ti and osmologi alobservations are made.Several authors have presented ideas for the sour e of the ionizing power of the WIM. Most re ently, Miller& Cox (1993), Dove & Shull (1994), and Dove, Shull, & Ferrara (2000) present a pi ture of a pervasive, warm,(originally) neutral medium being ionized by radiation from early-type stars leaking between old H I louds orout of density bounded H II regions. This pi ture results in regions of fully ionized gas and \shadowed" neutralgas. Other theories suggest that the H� emission should be positively orrelated with H I features. Norman(1991) and Koo, Heiles, & Rea h (1992) asso iate the pervasive H� emission with the walls of superbubbles,\ himneys," and \worms," while M Kee & Ostriker (1977) suggest that prominent H� emission is on�ned tothe surfa e of neutral louds bathed in a hot medium. Alternatively, Spitzer & Fitzpatri k (1993) and S iama(1990) have suggested that the H+ should be well-mixed with the neutral material.Major advan es in dete tor te hnology have re ently allowed breakthrough studies of the WIM. Dire tmeasurement of extremely faint, opti al diagnosti lines in the WIM, in luding He I �5876 (Tufte 1997) and[O I℄ �6300 (Reynolds et al. 1998), are beginning to pla e important onstraints on the nature and sour eof the ionization and heating of this gas in our Galaxy. In addition to our proje t, several imaging surveysare now underway to map the details of the stru ture of faint H� emission near the plane of the Milky Way(Dennison, Simonetti, & Topasna 1998; Gaustad, M Cullough, & Van Buren 1996; Parker & Phillips 1998).There have a been quite a few intensive studies on the nature of the WIM in other galaxies|often referredto in this ontext as \Di�use Ionized Gas" (DIG). Re ent opti al emission line imaging studies of galaxies byGreenawalt et al. (1998) and Hoopes, Rand, & Walterbos (1999) explore the detailed relationship between theWIM and H II regions in the plane and show that the fra tion of total H� luminosity in spiral galaxies arisingfrom the WIM is typi ally around 50%. The extended layer of WIM at large distan es from the gala ti planeoften shows networks of shells and �laments as well as an overall di�use omponent. The most omprehensivespe tros opi work on an edge-on spiral galaxy similar to the Milky Way has been undertaken by Rand (1997;1998), who has measured numerous emission lines in the halo of NGC 891. Domg�orgen & Dettmar (1997) andGolla, Dettmar, & Domg�orgen (1996) have also probed similar lines in the edge-on galaxies NGC 2188 andNGC 4631. All found a pronoun ed in rease in the intensity ratios [S II℄/H� and [N II℄/H� in regions withfainter H� emission. For spe tra obtained perpendi ular to the galaxies, these ratios rise systemati ally withdistan e from the gala ti plane. Ferguson, Wyse, & Gallagher (1996) and Otte & Dettmar (1999) studied thehighly-in lined SBm galaxy NGC 55 and found a smooth transition in the emission-line ratios of [O I℄, [O II℄,[N II℄ and [S II℄ to H� from bright H II regions to fainter WIM gas. In many of these studies, the emissionline trends in the brighter regions of the WIM an be explained fairly well by photoionization models, butdis repan ies often reep in at lower emission measures where [S II℄ and [N II℄ (most notably) emission be omesequal to that of H�. Here urrent photoionization models have diÆ ulty emulating the WIM.To further our understanding of this phase of the ISM, we have designed and built a dedi ated surveyinstrument, the Wis onsin H-Alpha Mapper (WHAM), to produ e the �rst deep, velo ity-resolved maps of theWIM in our Galaxy. Below we highlight the features of the H� survey in x2, review urrent results derivedfrom WHAM data in x3, and dis uss future s ien e to be explored with the instrument in x4.2. THE WHAM SURVEYAlthough the on-going H� imaging surveys are starting to provide striking details of WIM features withtheir moderate angular resolution (� 10) and deep overage (� 1 R; 1 R = 106 ph m�2 s�1), they la k velo ityresolution. WHAM omplements these imaging surveys ni ely by providing a data set with 12 km s�1 spe tralresolution and one-degree angular resolution. The addition of this apability allows the WHAM survey to add

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

240 HAFFNER, REYNOLDS, & TUFTE

Fig. 1. A preliminary portion of the WHAM H� Sky Survey. Grays ale represents total integrated H� emissionwithin vLSR = �100 to +100 km s�1. White 'point sour es' are bright stellar absorption features that will beremoved during the �nal survey redu tion.two important pie es to the di�use H� pi ture.First, at this velo ity resolution, Gala ti emission is separated from H� emission from the earth's upperatmosphere, the \geo oronal" H� line. Sin e Gala ti emission from most regions of the sky is shifted 20{30km s�1 away from the atmospheri line at some time during the year (due to orbital motion of the earth),prudent observing ta ti s and a urate geo oronal subtra tion make an absolute intensity alibration of theWHAM survey possible. Sin e the intensity of this foreground emission varies with both time and dire tion onthe sky, normal �lter imaging te hniques annot inherently provide a onsistent zero-point referen e for absolute alibration. The spe tral separation of the two H� lines also prevents the sky line from adding unwanted noiseto the Gala ti spe tral region, allowing WHAM to unambiguously dete t Gala ti emission down to 0.1 R(EM � 2 p m�6).More importantly, velo ity resolution gives us the unique opportunity to explore the a tual kinemati s ofthe WIM and other extended di�use H� sour es. Without this resolution, omparing stru tures found in theWIM with other surveys, parti ularly H I, is diÆ ult (e.g., Reynolds et al. 1995). Our hope is that su h omparisons will reveal the relationships between the WIM and other omponents of the interstellar mediumand help dis riminate among powering sour es for the WIM.The H� survey onsists of more than 37,000 pointings of the one-degree beam with �b = 0:Æ85 and �l =0:Æ98= os b with Æ � �30Æ. Ea h 30-se ond exposure results in a 200 km s�1 wide spe trum entered near theLo al Standard of Rest (LSR). More details of the WHAM instrument and survey strategy an be found inTufte (1997) and Ha�ner (1999). At the time of this writing, we are �nishing the �nal data redu tion passand the survey is expe ted to be released in a few months. The data presented here are from a preliminaryredu tion whi h did not a ount for numerous faint (< 0:1 R) atmospheri lines that are being removed duringthe �nal pass.Even without the full alibration planned for our �nal release, Figure 1 already shows the impressive omplexity of H� emission in our Galaxy. The region of the sky we show here was omposed from over 13,400spe tra, about a third of the northern sky survey. The more obvious extra-planar emission features in lude the

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

THE WISCONSIN H� MAPPER 241

Fig. 2. A magni�ed view of a se tion of Figure 1 that fo uses on two prominent H� �laments dis overed in theearly survey redu tion.Orion-Eridanus superbubble (` = 180Æ to 230Æ, b = 0Æ to �60Æ), the � Per (` = 140Æ to 180Æ, b = �10Æ to �25Æ)and � Cam (` = 145Æ, b = +15Æ) extended H II regions, the Monogem Ring supernova remnant (` = 180Æ to220Æ, b = 0Æ to +40Æ), and the verti al �lament rising out of an H II region in the plane at ` = 225Æ (b = 0Æ to+50Æ). 3. CURRENT RESULTS3.1. H� FilamentsFigure 2 is a magni�ed se tion of the upper-left orner of Figure 1. The long, verti al �lament that risesout of the plane at ` = 225Æ and the shorter, horizontal ar between b = +30Æ and +40Æ do not appear tobe asso iated with any other phase of the ISM tra ed by urrent all-sky surveys. They both have intensitieswhi h range from 0.5 to 1.0 R (EM � 1.0 { 2.0 p m�6). The velo ity at the base of the verti al �lament oin ides with the H II region near ` = 225Æ, b = 0Æ, CMa OB1, whi h is lo ated about 1 kp from the sun.At this distan e, the �lament rises to 1.2 kp above the plane before appearing to bend ba k down toward theplane. There is also a signi� ant velo ity gradient along its length, from vLSR = +14 km s�1 at the base tovLSR = �25 km s�1 at the apex. Mu h more detail on these H� �laments an be found in Ha�ner, Reynolds,& Tufte (1998).We present several possible s enarios in Ha�ner et al. (1998), but none has been spe i� ally modeled. Sin ethe published letter, we have be ome intrigued by the possibility that su h �laments ould be generated bylarge-s ale magneti �elds. Observations of the stellar polarization from stars behind these stru tures shouldprovide some information on the validity of su h an idea. We hope to un over more of these unique stru tureswhen the H� survey is fully redu ed!

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

242 HAFFNER, REYNOLDS, & TUFTE3.2. Perseus ArmA taste of the ri h spe tros opi information that is provided by the WHAM survey an be seen in Figure 3.Here a small portion of the Galaxy toward the Perseus arm (` = 123Æ to 164Æ, b = �35Æ to �5Æ) is displayedin various integrated velo ity bands. Due to Gala ti rotation, the Perseus arm emission is well separated fromLo al emission in Figures 3e through 3h and an be studied with little ontamination from nearby gas.We exploited this velo ity separation in Ha�ner et al. (1999) to make a new measurement of the verti alextent of the WIM layer. This new, high-quality data results in a measured exponential s ale height of 1:0�0:1kp if the Perseus arm is at a distan e of 2.5 kp from the sun. This new value is in ex ellent agreement withprevious estimates (Reynolds 1991; Reynolds 1997).As data taking for the H� survey began to �nish up near the end of 1997, we began mapping portions ofthe sky in [S II℄ and [N II℄ as well. This same region toward the Perseus arm was hosen as our �rst target. Asdes ribed in Ha�ner et al. (1999), new information about the global properties of the WIM an be obtained by ombining emission line observations. In parti ular, we �nd that S seems to be about 40{60% in the form ofS+ with notable oherent spatial variations and di�eren es between the Lo al and Perseus arms. Su h valuesare typi ally expe ted in a s enario where the WIM is predominantly photoionized, but the un ertain valueof S abundan e may ome into play as noted by Semba h et al. (2000). However, the temperature trends ofthe H� emitting gas do not seem to be explained well by urrent models (Reynolds, Ha�ner, & Tufte 1999;Reynolds, these pro eedings).In Ha�ner et al. (1999), we �nd that both the [N II℄/H� and [S II℄/H� ratios rise as the H� intensityde reases. This trend, ombined with the fa t that the [S II℄/[N II℄ ratio is remarkably onstant, leads us to on lude that rising temperatures rather than hanges in the ionization stru ture of the gas are produ ingelevated ratios in these fainter H� regions. The most striking example of this e�e t is revealed when looking atthe verti al stru ture of the Perseus arm where de reasing H� intensity with height above the plane is simplydue to the de reasing ele tron density. Our estimates suggest that at 1.75 kp , the ele tron temperature is over10,000 K. 3.3. Other WHAM S ien eHeiles has been studying the Orion-Eridanus superbubble in great detail for the last few years. The WHAMdata has helped by allowing us to explore the full range of ISM phases present in the region and to startuntangling the omplex kinemati s of the superbubble (Heiles, Ha�ner, & Reynolds 1999). The H� emissionfrom several di�use stru tures in the region has also been ombined with radio survey observations to measuretemperatures and to derive the extin tion of the H� emission (Heiles et al. 2000).A re ent de omposition of FIRAS far-infrared emission implied that dust in the WIM is a substantial ontributor (Laga he et al. 1999). The addition of WHAM data to this de omposition has indeed on�rmedthis fa t in a few small regions of the sky (Laga he et al. 2000). Sin e there will be a substantial amountof WHAM H� survey data from high Gala ti latitudes, where mole ular and H II regions are less abundant,future appli ation of these orrelation te hniques should prove quite valuable in extra ting details about dustin the WIM and the Cosmi Far-Infrared Ba kground.4. FUTURE WHAM SCIENCEAs multiwavelength observations in the Perseus arm show (x3.2; Ha�ner et al. 1999; Reynolds et al. 1998),the H� survey is only a starting point for studies of di�use emission with WHAM. We are ontinuing to olle tdata and have started several additional proje ts.� Although even the brightest of the nebular [O III℄ lines at 5007�A is quite faint in the WIM, Rand (1998)has tra ed this emission out of the plane in NGC 891 and �nds that [O III℄/H� in reases at high jzj. Weare investigating if this trend is similar in the Milky Way. With an ionization potential around 35 eV,O++ is an important tra er of ionization onditions in the WIM.� In a joint STIS program with Jenkins & Tripp, we plan to ombine the wealth of information from highsignal-to-noise, high resolution absorption line spe tra with multiwavelength WHAM emission spe tra

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

THE WISCONSIN H� MAPPER 243

Fig. 3. These images show the integrated H� emission (dark) in eight sele ted velo ity bands. The axes areGala ti longitude and latitude. The overplotted box shows the lo ation of a previous H� survey in this region(Reynolds 1980; Reynolds et al. 1995). (a) shows the total intensity of H� from the region by integrating overvLSR = �100 to +100 km s�1. (b) through (h) show 12 km s�1 integration sli es as labeled in the �gure.

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

244 HAFFNER, REYNOLDS, & TUFTEtoward a sele t group of stars in and behind the Perseus arm. In addition to deriving further onstraintson the physi al onditions in WIM gas, we hope this te hnique will be useful in de oupling the ontributionof the WIM and Warm Neutral Medium for ertain spe ies that are prevalent in both phases (Semba het al. 2000).� In early 2000 we plan to begin a survey of H� emission near the Gala ti Plane (perhaps up to jbj = 20{30Æ) and toward brighter di�use targets su h as the Orion-Eridanus superbubble. Combining H� spe trawith the H� survey should provide interesting information about the extin tion toward these H� emittingregions.� Sin e O and early B stars are able to provide a substantial amount of the ionization for the WIM, we have olle ted multiwavelength data on several fainter, extended H II regions away from the Gala ti plane.At least one new H II has been found in the H� survey (Ha�ner 1999). The B0.5e + sdO binary system� Per lo ated at ` = 131:Æ3, b = �11:Æ3 is surrounded by a faint (< 10 R), ir ular (d � 50 p ) H� nebula(Figures 3b and 3 ). The line ratios are mu h di�erent than the nearby O-star H II region reated by �Per. We plan to use these observations to help the understanding of UV emission from su h stars.� As shown in Ha�ner et al. (1999), [S II℄ and [N II℄ emission are bright enough to be tra ed to high Gala ti latitudes and seem to provide signi� ant information on the ionization and temperature of the WIM. Weplan to extend the maps of these other emission lines to further explore the global properties of the WIM.� Finally, WHAM has an imaging mode whi h an provide one-degree, few-ar minute resolution imageswith a very narrow spe tral bandpass (12{200 km s�1). Although time onsuming, we an use this modeto explore features smaller than the one-degree survey resolution. With multiwavelength apability andvelo ity-resolution, this observing mode may open up a whole new area of WHAM s ien e.We en ourage the reader to visit the WHAM website at http://www.astro.wis .edu/wham/ for informationabout these proje ts as well as the status of the WHAM H� survey, whi h will be released in early 2000.LMH would like to thank the organizers for the invitation to speak and for hosting a wonderful andenlightening onferen e! The authors would also like to re ognize ontinuing support for the WHAM proje tthrough grant AST9619424 from the National S ien e Foundation.REFERENCESDennison, B., Simonetti, J. H., & Topasna, G. A. 1998, PASA, 15, 147Domg�orgen, H., & Dettmar, R.-J. 1997, A&A, 322, 391Dove, J. B., & Shull, J. M. 1994, ApJ, 430, 222Dove, J. B., Shull, J. M., & Ferrara, A. 2000, ApJ, 531, 846Ferguson, A. M. N., Wyse, R. F. G., & Gallagher, J. S. 1996, AJ, 112, 2567Gaustad, J., M Cullough, P., & Van Buren, D. 1996 PASP, 108, 351Golla, G., Dettmar, R.-J., & Domg�orgen, H. 1996, A&A, 313, 439Greenawalt, B., Walterbos, R. A. M., Thilker, D., & Hoopes, C. G. 1998, ApJ, 506, 135Gu�elin, M. 1974, in IAU Symp. 60, Gala ti Radio Astronomy, ed. F. J. Kerr & S. C. Simonson, (Dordre ht: Reidel),51Ha�ner, L. M. 1999, Ph.D. thesis, University of Wis onsin{MadisonHa�ner, L. M., Reynolds. R. J., & Tufte, S. L. 1998, ApJ, 501, L83. 1999, ApJ, 523, 223Heiles, C., Ha�ner, L. M. & Reynolds, R. J. 1999, in ASP Conf. Ser. Vol. 168, New Perspe tives on the InterstellarMedium, , ed. A. R. Taylor, T. L. Lande ker, & G. Jon as (San Fran is o: ASP), 211Heiles, C., Ha�ner, L. M., Reynolds, R. J., & Tufte, S. L. 2000, ApJ, submittedHoopes, C. G., Rand, R. J., & Walterbos, R. A. M. 1999, ApJ, 522, 669Hoyle, F., & Ellis, G. R. 1963, Australian J. Phys, 16, 1Koo, B.-C., Heiles, C., & Rea h, W. T. 1992, ApJ, 390, 108Laga he, G., Abergel, A., Boulanger, F., D�esert, F. X., & Puget, J.-L. 1999, A&A, 344, 322Laga he, G., Ha�ner, L. M., Reynolds. R. J., & Tufte, S. L. 2000, A&A, in press

Ast

rop

hy

sic

al P

lasm

as:

Co

de

s, M

od

els

, a

nd

Ob

serv

atio

ns

(Me

xic

o C

ity

, 2

5-2

9 O

cto

be

r 1999)

Ed

ito

rs: Ja

ne

Art

hu

r, N

an

cy

Bri

ck

ho

use

, &

Jo

sé F

ran

co

THE WISCONSIN H� MAPPER 245M Kee, C. F., & Ostriker, J. P. 1977, ApJ, 218, 148Miller, W. W., III, & Cox, D. P. 1993, ApJ, 417 579Norman, C. A. 1991, in IAU Symp. 144, The Interstellar Disk-Halo Conne tion in Galaxies, ed. H. Bloemen(Dordre ht: Kluwer), 337Otte, B., & Dettmar, R.-J. 1999, A&A, 343, 705Parker, Q. A., & Phillips, S. 1998, PASA, 15, 28Rand, R. J. 1997, ApJ, 474, 129. 1998, ApJ, 501, 137Reynolds, R. J. 1980, ApJ, 236, 153. 1991, ApJ, 372, L17. 1993, in AIP Pro . Vol. 278, Ba k to the Galaxy, ed. S. S. Holt & F. Verter (New York: AIP), 156. 1997, in The Physi s of Gala ti Halos, ed. H. Les h, R.-J. Dettmar, U. Mebold, & R. S hli keiser (Berlin:Akademie-Verlag), 57Reynolds, R. J., Ha�ner, L. M., & Tufte, S. L. 1999, ApJ, 525, L21Reynolds, R. J., Hausen, N. R., Tufte, S. L. 1998, & Ha�ner, L. M. 1998, ApJ, 494, L99Reynolds, R. J., Tufte, S. L., Kung, D. T., M Cullough, P. R., & Heiles, C. 1995, ApJ, 448, 715Semba h, K. R., Howk, J. C., Ryans, R. S. I., & Keenan, F. P. 2000, ApJ, in pressS iama, D.W. 1990, ApJ, 364, 549Spitzer, L., Jr., & Fitzpatri k, E. L. 1993, ApJ, 409, 299Tufte, S. L. 1997, Ph.D. thesis, University of Wis onsin{Madison

L. M. Ha�ner and R. J. Reynolds: Department of Astronomy, University of Wis onsin{Madison, 475 NorthCharter Street, Madison, WI 53703 (ha�ner, reynolds�astro.wis .edu).S. L. Tufte: Department of Physi s, Lewis & Clark College, 0615 SW Palatine Hill Road, Portland, OR 97219(tufte�l lark.edu).