------------------------------------------------------------------------ From: Harvey Liszt hliszt@nrao.edu To: gcnews@aoc.nrao.edu Subject: submit SgrRing.tex A&A, submitted 9 May 2009 %astro-ph/0905.1412 \documentclass[]{aa} \usepackage{natbib} \usepackage{psfig} \usepackage{graphicx} \usepackage{txfonts} \bibsep 0pt \def\EBV{\mbox{E$_{\rm B-V}$}} \def\AF{\mbox{A$_{F\prime \rightarrowF\prime\prime}$}} \def\Nmu{\mbox{N$_\mu$}} \def\AV{\mbox{A$_{\rm V}$}} \def\AMSX{\mbox{A$_{\rm MSX}$}} \def\RV{\mbox{R$_{\rm V}$}} \def\AB{\mbox{A$_{\rm B}$}} \def\HH{\mbox{H$_2$}} \def\HHHP{\mbox{H$_3$}^+} \def\XCH{\mbox{X$_{\rm CH}$}} \def\nH2{{\rm n}({\rm H}_2)} \def\NH2{{\rm N}({\rm H}_2)} \def\pccc{~{\rm cm}^{-3}} \def\pcc {~{\rm cm}^{-2}} \def\NCO{N(\rm CO)} \def\Tstar#1 {\mbox{${\rm T}_{\rm #1}^*$}} \def\Tsub#1 {\mbox{$T_{\rm #1}$}} \def\TK {\Tsub K } \def\TB {\Tsub B } \def\Tbg {\Tsub bg } \def\TD {\Tsub D } \def\TL {\Tsub L } \def\Tsp {\Tsub sp } \def\TA {\Tsub A } \def\Texc {\Tsub exc } \def\vturb{\mbox{v$_{turb}$}} \def\Tmb {\Tsub mb } \def\Tcmb{\Tsub cmb } \def\Tcont{\Tsub cont } \def\Snu{S_{\nu}} \def\arcsec{\mbox{$^{\prime\prime}$}} \def\arcmin{\mbox{$^{\prime}$}} \def\omet{\mbox{$(1-{\rm e}^{-\tau})$}} \def\ometov{$(1-{\rm e}^{-\tau({\rm v})})$} \def\Romet{(1-{\rm e}^{-R\tau})} \def\iomet{$\int{1-e^{-\tau}}dv$} \def\itau{$\int \tau dv$} \def\degr{\mbox{$^{\rm o}$}} \def\p{\mbox{$^+$}} \def\cotw {\mbox{$^{12}$CO}} \def\coth {\mbox{$^{13}$CO}} \def\coei {\mbox{C$^{18}$O}} \def\as{^{\prime\prime} } \def\hcop{\mbox{{HCO\p}}} \def\hocp{\mbox{{HOC\p}}} \def\chp{\mbox{CH\p}} \def\cth{\mbox{C$_3$H}} \def\cch{\mbox{C$_2$H}} \def\cfh{\mbox{C$_4$H}} \def\CnHm{\mbox{C$_n$H$_m$}} \def\hhco{\mbox{H$_2$CO}} \def\h13cop{\mbox{{H$^{13}$CO\p}}} \def\nnhp{\mbox{N$_2$H\p}} \def\c3h2{\mbox{C$_3$H$_2$}} \def\ad{^{\rm{o}} } \def\am{^{\prime}} \def\Vphi{$V_\theta(R)$} \def\R0{R$_0$} \def\Jl{{{\rm J}_l}} \def\oneskip{\vskip\baselineskip} \def\mc{\mu\rm m} \def\kpc{\rm kpc} \def\eg{\it e.g.} \def\etal{\mbox{\it et al.}} \def\deg{{}^\circ} \def\ddeg{{}^\circ\kern-.1em} \def\Rsun{{R_0}} \def\Lsun{{L_0}} \def\Msun{M$_{\rm sun}$} \def\pc{\rm pc} \def\th{{\Theta}} \def\thsun{{\Theta_0}} \def\kms{\mbox{km\,s$^{-1}$}} \def\ps{\mbox{s$^{-1}$}} \def\bll{BL Lac} \def\fight{10.8} \def\E#1 {$10^{#1}$} \def\E#1 {E{#1}} \def\P#1,{$\nH2\TK~=~#1\times~10^4\pccc$~K} \def\ec#1,#2,#3,{#1\,(#2)\E{#3}} \def\xe{x(e)} \def\Cas{Cassiopeia } \def\sac{`spiral-arm' clouds} \def\zoph{$\zeta$ Oph} \def\methCN{\mbox{CH$_3$CN}} \def\H3{\mbox{H$_3$}} \def\Lya{\mbox{Ly-$\alpha$}} \def\ammon{\mbox{N\H3} } \def\zcg{\mbox {$\zeta^\gamma_C$}} \def\zetaH{\mbox{$\zeta_H$}} \def\RH2{\mbox {R$_{\rm G}$}} \def\fH2{\mbox {f$_{\HH}$}} \def\FH2{\mbox {F$_{\HH}$}} \def\W13{\mbox{W$_{13}$}} \def\WCO{\mbox{W$_{\rm CO}$}} \sloppy % \title{HII regions, infrared dark molecular clouds and the local geometry of the Milky Way's nuclear star-forming ring} \author{H. S. Liszt\inst{1} } % \institute{National Radio Astronomy Observatory, 520 Edgemont Road, Charlottesville, VA, USA 22903-2475} % \thesaurus{ % 09 % A&A Section 9 % (09.01.1 %ISM: abundances % 09.03.1 %ISM: clouds % 09.13.2 %ISM: molecules % 09.19.1 %ISM: structure % 13.19.3 %Radio lines: interstellar) % } \begin{document} \date{received \today} \offprints{H. S. Liszt} \mail{hliszt@nrao.edu} % % \abstract{}{}{}{}{} % 5 {} token are mandatory \begin{abstract} % context heading (optional) leave it empty if necessary {Observations of the HII region-molecular cloud complexes Sgr B, C, D and E provide unique insights into the structure of a nuclear star-forming ring.} % aims heading (mandatory) {To interpret the galactic center HII region/molecular cloud complexes as parts of a barred galaxy's nuclear star-forming ring.} % methods heading (mandatory) {We compare observations of 18cm VLA radiocontinuumm, $8-22\mu$ MSX IR and 2.6mm BTL/KP12m CO emission from the Sgr B, C, D and E complexes in the inner few degrees of longitude.} %results heading (mandatory) {The observed IR structure and 8-22 $\mu$ IR source spectra are heavily influenced by extinction in individual compact HII regions and over entire source complexes: in the most striking case, the powerful compact HII regions in Sgr B2 at $l > 0.6$\degr\ are almost undetected by MSX. In a few especially favorable cases the IR extinction from individual galactic center molecular clouds is seen to be wavelength-independent. The kinematics of the HII regions are mostly rotational, as expected for a nuclear star-forming ring but with spectactular departures in Sgr B and Sgr C.} %conclusions headiing (optional), leave it empty if necessary {Owing to high molecular column density and flattening of the extinction curve, the galactic center HII regions would likely be shrouded even if viewed from outside. Slight departures from circular motion and perfect alignment with the galactic equator allow the front (Sgr B) and rear (Sgr C) portions of the Galaxy's nuclear star-forming ring to be distinguished. Much of the volume interior to the ring may be somewhat hollow.} % \end{abstract} \keywords{ interstellar medium -- molecules } % Molecular gas and the Sgr A-E % H II regions reside in a nuclear star-forming ring as is % commonly seen in other galaxies. The near and far portions of % the ring are % slightly separated in galactic latitude and in velocity owing to % a % non-circular component of the motion which overall shows strong gradients % due to galactic rotation. Sgr C and the % 50 \kms\ gas near Sgr A are one side (likely the back) and Sgr B % on the % other. Placement of Sgr E is ambiguous because it is viewed % near one extremum of the ring. % Locally, various portions of the ionized gas traced in the % cm-wave radiocontinuum are visible in various ways in the 8 and % 24 $\mu$ % infrared bands but there is no unique way of predicting IR % morphology % given the radio continuum, beyond the obvious that the % non-thermal % radiocontinuum has no IR counterpart. However, and quite % strikingly, % much or most of Sgr B2 is absent in the IR despite being of % thermal origin: the molecular shell surrounding Sgr B, inside of % which all the radio continuum has formed, is especially % prominent in IR % absorption, suggesting that this portion of the nuclear ring is % in front.} \titlerunning{The shiny ring hidden in the Milky Way Bar} \maketitle \section {Introduction.} This is the third paper of a short series \citep{Lis06Shred,Lis08Multi} discussing the Milky Way's inner galaxy gas distribution in the context of the characteristic features which are present in nominally similar barred external galaxies \citep{ReyDow99,GreRey+99,RegShe+99} and in models of barred galaxy gas flow \citep{Fux99,RegShe+99,RegTeu04,RodCom08}. Inasmuch as possible, but without being overly reductive, we wish to meld the local data, with its unique galactic perspective and overwhelming wealth of detail, with a framework which has proved so successful in elucidating structure in other galaxies. The desirability of doing this seems clear, and an entirely similar view has recently been adopted by \cite{RodCom+06} and \cite{RodCom08}. However, connections between Milky Way observations and those of other systems have been obscured by the complexity and richness of the local observations and by our unique perspective within the galactic disk. Some aspects of of the Milky Way have been made to appear unique simply by our unparalleled vantage point and common features in barred galaxies have been missed in the Milky Way. In the first two papers we discussed material associated with the large-scale bar dust lanes. We suggested \citep{Lis06Shred} that a series of enigmatic, highly-localized broad-lined molecular gas features should be associated with the shredding of molecular clouds at positions of gas uptake into the large-scale galactic dust lane shocks. In other galaxies related phenomena are visible as broad lines and velocity shifts across the dust lane \citep{ReyDow99} but neither the individual uptake events nor their surprising and quite exteme vertical elongation (common to all the Milky Way events) are distinguishable. We also showed \citep{Lis08Multi} that the peculiar kinematic signature commonly identified with the Milky Way's dust lane can be found in at least three distinct features in the northern hemisphere (also see \cite{RodCom+06}) and discussed how the temporal sequence associated with gas uptake and inward flow might be used to understand the vertical structure of the gas flow. In this work, we focus on the structure interior to the inner ends of the prominent bar shock dust lanes, in the more immediate vicinity of the nucleus. In other barred galaxies the star formation in this zone occurs within narrowly-defined nuclear star-forming rings \citep{KorCor04}. In the Milky Way, this phenomenon is represented by the complex of sources Sgr A-E, the giant HII region-molecular cloud complexes at |l| $< 1.2$\degr\ corresponding to projected distance of 170 pc from SgrA$*$. However, owing to our viewing geometry, a ring morphology has been far from obvious; instead, the region has been generally characterized as the ``central molecular zone'' (CMZ) \citep{MorSer96} which in its most extreme form is described as being nearly filled with high density molecular gas. Here we suggest that the CMZ is actually rather hollow, as is typically the case in barred galaxies having prominent nuclear star-forming rings (NSFR). Following the brief discussion in \cite{Lis06Shred} pointing out the similarity of the inner galaxy H II region kinematics to those of extragalactic NSFR, this paper considers the ring structure in detail, both locally within individual HII region/molecular cloud complexes, and overall. We compare the distributions and/or kinematics of three tracers, namely the L-band (18cm) radiocontinuum emission, 8-22$\mu$ IR emission and 3mm CO emission. The organization of this work is as follows. Observational material, some of which is new, is discussed in Sect. 2. Section 3 gives an overview of the distribution of mid-IR and molecular emission in the inner $\pm$200 pc and a detailed discussion of the individual off-nuclear HII region/molecular cloud complexes Sgr D, B, C, and E. In the most extreme case, the strongest radio HII regions in Sgr B2 are lost to extinction in the IR, even at 22$\mu$. In Section 4 we derive the extinction originating locally in a few especially-prominent infrared-dark (molecular) clouds (IRDC): consistent with previous results \cite{IndMat+05,NisNag+08} and a plateau in the model interstellar extinction curves of \cite{WeiDra01}, the extinction shows no decline over the MSX bands. In Sect. 5 we derive IR-radio spectra of the Sgr source complexes and their compact constituents, showing how the IR-radio continuum comparison can be used to infer high extinction when it is not obvious in the underlying source morphology. In Sect. 6 we discuss the global placement of sources in a putative ring, within which the non-circular motion and front-back tilt are just large enough to distinguish the near and far ring segments. Section 7 is a summary. \section{Observational material considered} \subsubsection{New observations} The observational material considered here comes largely from published sources, but it also includes a substantial component of previously unpublished large-scale mapping of J=1-0 \coth\ emission from the NRAO Kitt Peak 12m telescope, observed in 1995 May by the author. Maps were made in the vicinity of Sgr B and D at $0.5\degr < l < 1.2 \degr$ and around Sgr E at $l=-1.2$\degr. These are fully spatially-sampled data with 1\arcmin\ spatial resolution, 1 MHz (2.7 \kms) velocity resolution and typical rms noise 0.1 K. %1 \begin{figure*} %\psfig{figure=RingF1.eps,height=16.9cm} \includegraphics[height=16.9cm]{RingF1.jpg} \caption[]{Mid-infrared continuum and molecular gas in the innermost $\pm$125 pc (R$_0=8$ kpc) of the Milky Way. Top: MSX emission in Band A ($8.3\mu$). Middle: MSX emission in Band E ($21.8\mu$). Bottom and superposed as contours in the middle panel: \coth\ emission integegrated at $-100 < v < 100$ \kms\ from the survey of \cite{BalSta+87}.} \end{figure*} %2 \begin{figure*} %\psfig{figure=RingF2.eps,height=8.5cm} \includegraphics[height=8.5cm]{RingF2.pdf} \caption[]{Views of the Sgr D region. Left: VLA radiocontinuum emission at 1616.4 MHz (grayscale) and overlayed contours of \coth\ emission integrated over $-20 < v < 0$ \kms\ as mapped at the then-NRAO, now-ARO 12m telescope with 1\arcmin\ resolution. Right: Sgr D as seen at MSX Band A (top), Band E (middle) and at 1616.4 MHz.} \end{figure*} \subsection{Previously published CO emission line data} Also shown here are maps drawn from the BTL galactic center CO survey \citep{BalSta+87} at 1.7\arcmin\ resolution. For \cotw\ these data are on a 1\arcmin\ grid with 1 MHz or 2.6 \kms\ velocity resolution and typical rms 0.5 K. For \coth\ the grid spacing is 30\arcsec\ and the typical rms noise level is 0.15 K. \cite{LisSpi95} published 45\arcsec\ maps of molecular gas near Sgr C from the SEST telescope, and these data are used here in Fig.Y. \subsection{Nominal CO-\HH\ conversion} Unless otherwise noted, velocity-integrated intensities of \cotw\ and \coth, denoted respectively by \WCO\ and \W13\ are converted to \HH\ column density N(\HH) assuming a typical galactic disk conversion N(\HH) $= 2\times10^{20}$ \HH\ $\pcc$ \WCO, and \WCO/\W13\ = 6 which is typical of galactic center material in the H II region complexes in our data and the BTL survey. \subsection{VLA L-band continuum} This is the data discussed originally by cite{Lis88}, cite{Lis92} and cite{LisSpi95}. It was taken in a 12.5 MHz band centered at 1616.4 MHz (a region of the spectrum subsequently lost to use by the Iridium satellite constellation) in the VLA C and D configurations and has spatial resolution typically 13\arcsec $\times$ 23\arcsec. \subsection{Radio recombination line data} The data, shown here in Fig Y, were taken in 1990 at the NRAO 43m telescope in Green Bank in 1990, and were first shown by cite{Lis92}. \subsection{MSX maps} The Mid-Course Space Experiment (MSX; \citep{PriEga+01}) mapped the galactic center region at 21\arcsec\ resolution an a 6\arcsec/pixel grid in four bands at center wavelengths ranging from $8\mu$ to $22\mu$. These data are given as specific intensity in units of W (m$^{2}$-sr)${-1}$. These can be expressed in Jy within individual pixels using the conversion factors given in Table 1 corresponding to the solid angle of 6\arcsec\ pixels and the the stated photometric bandwidths of the MSX spectral bands expressed in Hz. Figure A.1 in Appendix A illustrates the 50\%-response MSX bandwidths graphically on curves of the extinction cross-section per H atom for galactic dust \citep{WeiDra01}. %1 \begin{table} \caption[]{MSX Bands and conversion from pixel values to Jy} { %\small \begin{tabular}{lrc} \hline Band & wavelength & X$^1$ \\ &$\mu$ & sr Hz$^{-1}$\\ \hline A & 8.28 & 5.53$\times10^{3}$ \\ C & 12.13 & 2.40$\times10^{4}$ \\ D & 14.65 & 2.66$\times10^{4}$ \\ E & 21.85 & 2.02$\times10^{4}$ \\ \hline \end{tabular}} \\ $^1$ X is the factor by which MSX pixel values or sums of pixel values should be multiplied to convert to Jy. \\ \end{table} \section{Structure of the inner region and individual sources} The top two panels of Fig 1 show the mid-IR brightness in the MSX A and E-bands (around 8.3 and 21.8 $\mu$, respectively, see Table 1 and Fig. A.1). Superposed in each case are contours of the velocity-integrated intensity of \coth, \W13, at $|v| \le 100 $ \kms\ from the survey of \cite{BalSta+87}. Shown at bottom in grayscale is the same distribution of \W13. The 0.9\degr\ radius which contains most of the IR brightness projects to 125.6 pc across the line of sight at a Sun-center distance of 8 kpc (1\arcmin\ corresponding to 2.33pc). The outlying sources Sgr D and Sgr E at l=$\pm$1.1\degr\ ($\pm 154$ pc) which define the outer edge of the nuclear star-forming region are not shown in Fig. 1 but will be discussed separately in this Section along with the other sources. It is well-understood from observations of galactic H II region complexes (see \cite{PovSto+07,WatPov+08} and Appendex C) that the shorter wavelength panel in the middle of Fig. 1 mainly shows very hot dust so that the Sgr B, A, and C HII regions at 0.7\degr, 0.2\degr\ and -0.5 \degr, respectively, appear with greater clarity. The longer wavelength panel at the top of Fig. 1 shows more broadly-distributed material within the ambient neutral gas, and especially (as discussed below) at the HII region/molecular gas interfaces. As shown in the following subsections, all of the Sgr HII regions lie in a remarkably thin band running parallel to the galactic equator between b $= -0.05$\degr\ (the latitude of Sgr A$^*$) and b $= -0.10$\degr). However, the large-scale distributions of molecular gas and IR light are noticeably bowed, with most of the emission occuring well below the galactic plane at the extremes in Fig. 1. The IR distribution in the uppermost panel at 8$\mu$ is clearly affected by infrared dark clouds visible in CO, especially in an extended distribution of foreground extinction at l $> 0.25\degr$ shrouding Sgr B. The downward arc of the northern edge of the 8$\mu$ emission at postive longitude in Fig. 1 at top is more obviously an artifact of this extinction, and is not matched in the molecular gas at bottom. However, extinction is also visible, albeit somewhat more faintly, at the northern edge of the 8$\mu$ light distribution at negative longtitude where it also bows downward. Extinction in the Sgr B complex around l = 0.6\degr\ largely but not exclusively (see Sect. 3.2) arises in molecular gas at 0 \kms\ $<$ v $<$ 50 \kms\ and one especially clear case of extinction by this gas is seen just above the equator between Sgr A and B at l $\approx 0.25\degr$. Another prominent extinction arising in gas at negative velocity appears just above the Sgr C HII region at l $= 359.5\degr$ and both of these cases are discussed in detail in Sect. 4. It is less obvious that emission at the longest wavelength IR map in the middle panel is significantly affected but the extinction of the individual IRDC shows no obvious wavelength dependence from 8-22$\mu$ and some foreground extinction at 22$\mu$ which is not apparent from the morphology may be inferred from the spectra which are discussed in Sect. 5. The general question of how the material represented in Fig. 1 is actually distributed in space is left for Sect. 6 but the appearance of the molecular gas at bottom is clearly rather loopy and somewhat hollow at negative longtiudes. At postive longitudes the molecular gas in the Sgr B complex is so strong and so extended that the z-height of the material overall is otherwise obscured. The expectation is that the star forming regions in a strongly-barred galaxy like the Milky Way will be situated in a radially-narrow, approximately circular, mostly-rotating ring and the inference of a ring geometry for the Milky Way gas is discussed in Sect. 6. The nuclear region in our Galaxy is the only one whose vertical structure is distinguishable and, at least at negative longitudes, it appears that a slight scalloping may have displaced the front and back gas portions, so that they are not projected on top of each other. In Sect. 6 it is also shown that the front and back ring portions are distinguishable kinematically, owing to departures from pure circular motion. The structure of the individual HII region/molecular cloud complexes is discussed in the following subsections, progressing from highest (Sgr D) to lowest (Sgr E) galactic longitude. % 4 % \begin{figure*} % \psfig{figure=RingF4.eps,height=8.9cm} % \includegraphics[height=8.9cm]{RingF4.pdf} % \caption[] {Longitude-velocity plots of \coth\ emission from OTF mapping at the % NRAO 12m telescope at two latitudes in the Sgr B-D region. % The resolution is 1\arcmin\ and 2.7 \kms\ (1 MHz).} % \end{figure*} %3 \begin{figure} %\psfig{figure=RingF3.eps,height=10cm} \includegraphics[height=10cm]{RingF3.jpg} \caption[] {Sgr B as seen at MSX Band A (top), Band E (middle) and at 1616.4 MHz.} \end{figure} \subsection{Sgr D} The radiocontinuum properties of Sgr D were described by \cite{Lis92} and \cite{MehGos+98} and a Band E MSX image appears in \cite{ConCro04}. The source is visible at $\lambda2\mu$ \citep{BluDam99}. Although it lies at the ``right'' galactic latitude, its location within the ring is contentious. Fig. 3 shows the source morphology in the radio and MSX IR bands. Sgr D consists of an H II region to the North, visible at both radio and infrared wavebands, and an adjacent shell radio supernova remnant (radio SNR) which appears only in the radio see also \cite{Gra94}. The H II region is associated with a peak in the \coth\ emission having \W13\ $\approx 50$ K \kms\ and line brightness 5 K at v $\approx -17$ \kms\ (see also Fig. 2) near the recombination line velocity (Fig B.1). The nominal mass of associated molecular material (see Sect. 2.2) is $\approx 2\times10^5$ \Msun. The bright \coth\ peak most likely represents a region of interaction between the ionized and neutral gas distributions; overall the molecular gas has a shell-like appearance which cradles the southern portions of the extended thermal radiocontinuum distribution. Portions of this shell and the wishbone structure to the north are visible in the MSX Band A emission at top right in Fig. 2. Longer wavelength IR emission in MSX Band E is more nearly dominated by hot material in the compact HII region but extended structure is also visible, especially to the north. Although it is not obvious that the IR source morphology is affected by extinction, the global spectrum shows some deficiency of IR emission compared to the radiocontinuum, which in other cases results from foreground extinction of the IR (see Sect. 5). The placement of Sgr D along the line of sight remains somewhat of a puzzle owing to its recombination line velocity of -9 \kms\ (see Fig.B.1). Interpretation of absorption measurements has placed it at \citep{ConCro04} or beyond \citep{Lis91,MehGos+98} the galactic center. In particular, \cite{Lis91} argued for a position outside the bulge based on the narrowness of the mm-wave emission lines from the nearby molecular gas. Examination of molecular emission over a larger region does not provide much context for placement of Sgr D, whose velocity is distinct from those of other similary-narrow but spatially-extended ridges of emission in the region. Owing to the extraordinary kinematic complexity of the region, which falls between the nuclear ring and the innermost uptake event into the dust lane at l=1.3\degr\ \citep{Lis06Shred}, there actually is gas at rather low $|v|$ which is undeniably part of the inner-galaxy gas. Within the ring it is expected that stars and natal gas will separate and phase mix with other ring members after the stars form, making it less likely that the evolved progenitor of the Sgr D SNR and the O-stars in the HII region actually formed from the same cloud. Moreover, it is possible that the SNR is an interloper in the vicinity of the HII region and molecular cloud (if all are actually sited within the ring), so that proximity still does not demand a physical association. However, if an association between the HII region and SNR could be demonstrated, it would have interesting consequences for star formation mechanisms within the ring gas. A physical association between the HII region and SNR could be investigated by searching for 1720 MHz OH maser emission around the SNR. %4 \begin{figure*} %\psfig{figure=RingF4.eps,height=7cm} \includegraphics[height=7cm]{RingF4.jpg} \caption[]{Sgr B at infrared and radio wavelengths. Left: grayscale of 1616.4 MHz radiocontinuum, overlayed by contours of \coth\ at 1\arcmin\ resolution integrated at $50 < v < 100$ \kms\ which includes the radio recombination line velocites (see Fig. 3). Right: MSX Band E emission in grayscale with superposed contours of \coth\ integrated emission at $10 < v < 50$ \kms, which forms a shell within which most of the Sgr B radio and IR continuum is contained. Note the strong absorption associated with the molecular gas at right and the near total absence of Sgr B2 at $l > 0.6\degr$ even at 22$\mu$} \end{figure*} % % Fig. 4 illustrates the complexity of the molecular gas kinematics; % the % patterns of sharp velocity gradients and abrupt velocity reversals % are like nothing outside the bulge region, but have clear % analogues where gas is entering the standing dust % lane shocks at l = 1.3\degr, 3.2\degr\ and 5.5\degr\ % \cite{Fux99,Lis06Shred}; indeed one of these is apparent in the % upper % panel of Fig. 4 and the possible influence of the phenomena % associated % with it \cite{TanKam+07} upon material further inside the nuclear % ring was recently remarked by \cite{RodCom08}. % % The kinematics of Sgr B at l=0.7\degr\ v= 60 \kms\ are discussed % in the next Section. \subsection{Sgr B} The presence of extinction in the Sgr B source complex around l=0.6\degr\ is apparent in Fig. 1 but its full influence can better be seen by comparing source structure in the radio and IR continuum as in Fig. 3. Simply put, the strongest HII regions visible in the radio (at $l \ga 0.6$\degr) are mostly absent at both 8$\mu$ and 22$\mu$. Although the physical structure of Sgr B arises from an interaction between the neutral and ionized gas (which partly overlap in velocity and are physically co-mingled), the absence of the strong radio HII regions in the IR can only result from near-complete attenuation of the IR flux. The radiocontinuum and MSX Band E images are shown in Fig. 4 along with the integrated \coth\ emission at velocities above and below 50 \kms. The higher velocity emission, which overlaps the recombination line velocities (62 \kms; see Fig. B.1), is shown over the radiocontinuum with its clear image of the HII regions: lower velocity emission, from more extended gas responsible for the IR extinction, is overlaid on the MSX Band E image in which the extinction is the most important aspect of the extended neutral gas distribution. The IR and radio continuum emission largely exists inside a cavity in the lower-velocity molecular gas, which is partly filled in by material at the higher velocity. The brightest compact radio HII regions which are so heavily extincted are seen projected squarely against very bright (in CO) material at comparable velocity, and must be behind the molecular gas. The low-velocity shell and high-velocity cap have been interpreted as a cloud-collision in a series of papers by Hasegawa and his collaborators \citep{HasAra+08,SatHas+00,HasSat+94}; this explanation accounts for many of the properties of the regions chemistry and masers. Sgr B2 can probably be identified with the contact region between the star-forming ring and gas inflowing in the dust lanes in the nuclear regions of the Milky Way Bar (e.g. \cite{RegTeu03} and Fig. 9 of \cite{RegShe+99} or recent discussions in \cite{Lis06Shred} and \cite{RodCom08}). \subsection{Sgr C} Although far less active than Sgr B, the Sgr C HII region near l = 359.45\degr, b = -0.1\degr, at v = -60 \kms\ (see Fig. B.1) is positioned nearly symmetrically in space and velocity. The radiocontinuum structure has been discussed by \cite{Lis85}, \cite{TsuKob+91} and \cite{LisSpi95} and maps of the molecular gas were discussed by \cite{LisSpi95} and \cite{StaStu+98}. % 7 % \begin{figure*} % \psfig{figure=RingF7.eps,height=7cm} % \includegraphics[height=7cm]{RingF7.jpg} % \caption[] {As in Fig. 6, but for MSX Band E. Note that Sgr B2 % is also missing at this longer wavelength.} % \end{figure*} The IR and radiocontinuum structure of Sgr C is shown in Fig 5. In the radio, Sgr C consists of extended, round, thermal emission from an H II region powered by one or a small number of O-stars \cite{LisSpi95} and an adjacent,elongated, non-thermal filament which was the first such object to be seen outside Sgr A. Partly based on the structure in Sgr C \cite{Yus03} argued that non-thermal filaments in the galactic center are caused by interaction of young stellar clusters and ambient dense gas, but \cite{Roy03} argued that the H II region and filament are separated along the line of site and not physically related, owing to differences in their H I absorption spectra. The MSX Band E emission coincides with the radio HII region and arises from heated material within it. The more extended MXS Band A emission (whose appearance is strongly affected by extinction) occurs between the lower edge of the more negative-velocity gas to the North and the upper edge of the more positive-velocity emission, at the HII region velocity, to the South; only a rather minor portion of it overlaps the H II region. As such it could apparently be associated with either of the two molecular gas distributions. If the obvious extinction to the North had truncated the MSX Band A emission, an association between the IR and the gas at -125 $<$ v $<$ -85 \kms\ might have been more evident. Although there is weaker emission at the HII region velocity across the prominent Band A extinction north of Sgr C, emission around -100 \kms\ is much stronger and similar in appearance. The extinction likely arises in the lower velocity gas (see just below and Fig. 6). As noted by \cite{Roy03} there is H I absorption at -100 \kms\ toward the H II region. The extinction is discussed in more detail in Sect. 4. Based on the limited \coth\ mapping shown in Fig. 5 \cite{LisSpi95} noted merely that the Sgr C H II region is placed between the two molecular emission distributions at -125 $<$ v $<$ -85 \kms\ and -85 $<$ v $<$ -45 \kms. It has been something of a puzzle why gas at such widely separated velocities should all brighten near Sgr C, but larger-scale imaging of the molecular gas (Fig. 6) shows that the kinematic structure is actually very like that in Sgr B: a smaller core of emission near the recombination line velocity (-65 \kms; Fig. B.1) is surrounded by a partial ring of gas which is displaced in velocity. Apparently, Sgr C is actually physically associated with both kinematic components of the gas and, like Sgr B, it may represent the interaction of inflowing material in the dust lane with pre-existing material in the SFR. \subsection{Sgr E} The radiocontinuum structure and recombination line kinematics of Sgr E at l = -1.3\degr, b -0.1\degr, v = -215 \kms\ (Fig. B.1) have been observed by \cite{Lis92}, \cite{GraWhi93} and \cite{CraCla+96}. At a maximum projected radius of 195 pc corresponding to l = 358.6\degr, Sgr E marks the outer boundary of the galactic center star-forming ring in both galactocentric radius and velocity. Sgr E lacks a symmetric counterpart at positive longitude. Sgr E is unusual among the named galactic center H II region complexes in being devoid of extended emission. The very high observed velocities imply that the sources are observed near the sub-central point, in which case the line of sight velocity gradient is small and the sources could be distributed over a long path where the line of sight is tangent to the edge of the ring. Observing Sgr E may afford the opportunity to study a sample of somewhat older (i.e. B-type) ``field'' stars which have phase-mixed around the ring but the Sgr E source complex is not a physical entity. The IR and radiocontinuum structure of Sgr E is shown in Fig. 7. Nearly all of the radiocontinuum sources have IR counterparts. The radiocontinuum source at l = 358.6\degr, b = 0.06\degr\ lacks such a counterpart and was singled out by \cite{CraCla+96} for lacking a radio recombination line: It is probably an extragalactic interloper and perhaps a useful source for acquiring comparison absorption spectra through the galactic disk. The nominal total \HH\ mass represented in Fig. 7 is $1.1\times10^6$ \Msun\ (see Sect. 2.2). \section{IRDC and the CO-\HH\ conversion factor} Extinction of the IR radiation by darker foreground material is apparent in Fig. 1 and two especially well-defined cases of association between molecular emission and extinction in the MSX bands are illustrated in Fig. 8 and 9. \subsection{The IRDC near Sgr C} A very prominent IRDC seen just north of Sgr C was noted in Sect. 3.3 and illustrated in Fig. 5. It probably originates in gas seen in CO emission at -135 $<$ v $<$ -85 \kms\ (also see Fig. 6). Fig. 8 at top shows a latitude cut across the IRDC in MSX Band A. The peak absorption optical depth is straightforwardly estimated as 0.75, corresponding to an extinction of 0.81 mag. Extinction is probably present at 22$\mu$, illustrated by the transverse cut in the lower panel of Fig. 8 which shows a clear dip at the expected latitude. The extinction at both wavelengths must be comparable but determination of the spectrum of the extinction is better left for the even better-defined extinction between Sgr A and Sgr B. As seen in Fig. 7, the CO brightness associated with the IRDC near Sgr C is \WCO\ = 480 K \kms. \subsection{The prominent IRDC between Sgr A and Sgr B} MSX fluxes over a spatial cut across the especially prominent feature between Sgr A and Sgr B around l = 0.25\degr\ in Fig. 1 are shown in Fig. 9. The path of this cut runs at a 45\degr\ angle with respect to the galactic equator, along the short axis of the absorption, from l = 0.3\degr, b=-0.1\degr\ to l = 0.1\degr, b=+0.1\degr. Shown at top is a strip across the full extent of the cut in MSX Band A, illustrating how a baseline was established to gauge the strength of the absorption. At bottom in Fig. 9 are the absorption and optical depth (inset) profiles which ensue. They are lower limits because we have assumed that the absorbing material is black and we have not accounted for contamination by unrelated foreground emission. %5 \begin{figure} %\psfig{figure=RingF5.eps,height=9.24cm} \includegraphics[height=9.24cm]{RingF5.png} \caption[]{Sgr C. Top: MSX Band A emission with overlaid contours of emission at $-125 < v < -85 $ \kms\ at 1\arcmin\ resolution showing the higher-latitude, lower-velocity branch of the molecular ring (see Figs 1 and 2). Middle: MSX Band E. Bottom 1616.4 MHz radiocontinuum (grayscale) with overlaid contours of \coth\ emission at $-85 < v < -45 $ \kms, which includes the recombination line velocity of the Sgr C HII region (see Fig. 3). } \end{figure} %6 \begin{figure} %\psfig{figure=RingF6.eps,height=9.6cm} \includegraphics[height=9.6cm]{RingF6.pdf} \caption[]{\cotw\ emission near Sgr C. The grayscale represents emission at 1.7\arcmin\ resolution integrated over the range $-135 < v < -85 $ \kms\ and the contours are for emission integrated over the range $-85 < v < -35 $ \kms\ which includes the recombination line velocity of Sgr C as shown in Fig.3} \end{figure} The extinction is rather gray, with no obvious variation with band wavelength and a peak optical depth of 1.3 in all bands, corresponding to extinction of 1.4 mag. Wavelength-independent extinction in this region of the spectrum has recently been noted by \cite{IndMat+05} and \cite{NisNag+08} and is broadly consistent with a plateau in the interstellar extinction in the dust models of \cite{WeiDra01}. The overlap of the MSX 50\% photometric bandwidths with the wavelength-dependent extinction coefficients of \cite{WeiDra01} is shown in Fig. A1. At least approximately, an extinction coefficient C$^{\prime}$/H = $1.3\times10^{-23}\pcc/$H can be attributed across the MSX A-E bands. This implies column densities of $0.75/{\rm C}^{\prime} = 6\times10^{22}\pcc$ and $1.3/{\rm C}^{\prime} = 1.0\times10^{23}\pcc$ for the IRDC near Sgr C and around l = 0.25\degr, respectively, given their optical depths of 0.75 and 1.3. These then correspond to visual extinction \AV\ = $ 6\times10^{22}\RV/5.8\times10^{21} $ = 32.1 (\RV/3.1) mag and 53.4 (\RV/3.1) mag. Alternatively, \AV/\AMSX = 39.4 (\RV/3.1) given the common assumption that \EBV/mag = N(H)$/5.8\times10^{21} \pcc$ \citep{SavDra+77}. As suggested by the images Fig. 1, more detailed images of CO emission (not shown here) indicate clearly that the material responsible for the absorption at l=0.25\degr\ is at 0-50 \kms, i.e. it is the same material responsible for the extinction around the Sgr B complex as discussed above in Sect. 3.2. At the deepest trough in the IRDC, the peak integrated CO brightness in the 1.7\arcmin\ resolution BTL survey is 505 K \kms, which is some 400 K \kms\ higher than at adjacent positions outside the IRDC. This is equivalent to an CO-\HH\ conversion factor N(\HH)/\WCO\ = 0.5 N(H)/\WCO\ = $1.25\times10^{20}$ H (\RV/3.1) $\pcc$. A consistent result is also obtained for the feature near Sgr C. At this level of accuracy the CO-\HH\ conversion factor for galactic center gas is at most modestly below that for material near the Sun. %7 \begin{figure} %\psfig{figure=RingF7.eps,height=12.5cm} \includegraphics[height=12.5cm]{RingF7.jpg} \caption[]{Views of Sgr E. Top: MSX Band A overlaid with contours of \coth\ emission at 1\arcmin\ resolution, integrated over $-225 < v < -200$ \kms. Middle: MSX Band E (grayscale) and \coth\ emission integrated over $-200 < v < -175$ \kms. The recombination line velocities of the radio-brighter sources at l = 358.6\degr\ and 358.8 \degr\ are -210 \kms and -190 \kms, respectively.} \end{figure} \subsection{Sgr B} The IR extinction in Sgr B is so very spatially extended that a reliable baseline flux measurement did not seem feasible. The integrated CO intensities for the lower-velocity gas responsible for the IR extinction are shown in Fig. 6 and Fig. 7 and are somewhat larger than are present in the two IRDC just discussed, 140-160 K \kms\ for \coth\ and 800 - 1000 K \kms\ for \cotw. As discussed in Sect. 6, comparison of radio and IR fluxes suggests that the very strongest radio H II regions in the eastern part of Sgr B are behind material with optical depth of order 4 in the MSX bands. \section{Global and compact source spectra} Figure 10 shows IR-radio spectra of many sources in the galactic center region, extracted from the MSX and 18cm VLA maps, or using the 20cm radiocontinuum fluxes of the MAGPIS survey \citep{HelBec+06} for the template sources M17 and N10. All of the radiocontinuum fluxes have been scaled upward by a uniform factor of 250 to allow convenient viewing and the scaled spectrum of M17 \citep{PovSto+07} is shown in each panel for comparison. Except at the upper left, the flux scale of the M17 spectrum is arbitrary. At the upper left, the M17 fluxes have been rescaled to 8 kpc distance (from 1.6 kpc) and (as noted) further divided by a factor of two to facilitate comparison. The spectrum of N10, a galactic HII region bubble at a distance of 4.9 kpc, is also shown at upper left. The morphology of N10 is generic; it strongly resembles that of Sgr C (see Fig. C1 and Appendix C). Integrated spectra of the extended galactic center sources Sgr B, C, and D are shown at upper left; they resemble that of M17 in form but are somewhat weaker. In the case where the foreground extinction is obviously the greatest, Sgr B2 at l $>$ 0.6\degr, the radiocontinuum flux is conspicuously large (by a factor of five) relative to M17. Sgr D also has a large radio/IR flux ratio, suggesting a higher extinction. Overall the comparison with M17 seems quite fair and M17 should not be too heavily extincted by intervening unrelated gas in the MSX bands. However, given that all of the sources in the galactic center (not just those in Sgr B2) probably suffer from some significant foreground extinction, the similarity of their spectra to that of M17 imples that the extinction must be rather gray, as previously indicated by the individual IRDC discussed in Sect. 4 and and illustrated in Figs. 8 and 9 (see \cite{IndMat+05} and \cite{NisNag+08}). %8 \begin{figure} %\psfig{figure=RingF8.eps,height=14cm} \includegraphics[height=14cm]{RingF8.pdf} \caption[]{MSX intensity (W m$^{-2}$ sr$^{-1}$) along spatial strips through the IRDC near Sgr C. Top: a strip in galactic latitude at l = 359.4736\degr in MSX Band A at 8.2$\mu$. When the intensity distribution is interpolated across the IRDC as shown the optical depth is 0.75 (0.815 mag). Bottom: a strip in longitude in MSX Band E around 22$\mu$. Spatial resolution is 21\arcsec\ = 0.00583\degr.} \end{figure} Spectra of compact sources in Sgr B are shown in the righthand panels of Fig. 12. The systematics of spectra from the more heavily-extincted sources in Sgr B2 at higher longitude (lower left panel) seem especially clear. The weakest MSX source has the flattest IR spectrum and the largest radio/IR disparity; its MSX E-band/VLA 18cm flux ratio is low by a factor about 50 compared to the template spectrum of M17, corresponding to an additional 4 mag of IR extinction. As the MSX fluxes increase in Sgr B2 the IR spectrum steepens, but for the brightest IR source the 18cm flux is unexpectedly small. This pattern is mostly repeated in Sgr B1 in the upper right hand panel of Fig. 13 where again the two brightest MSX sources have comparatively small 18cm flux. Apparently, the brighter MSX E-band sources in Sgr B tend to have steeper IR spectra and higher Band E/18cm flux ratios. Spectra of sources in Sgr E, which is largely devoid of extended flux in the radiocontinuum, are shown at lower left in Fig. 13. IR Spectra were also extracted for three sources (nos 7-9, shown dashed and in blue) which show no obvious 18cm radiocontinuum at all; the radio fluxes reported for these sources are the totals within the boundaries of the respective IR images. These sources are characterized by very flat MSX spectra: the strong MSX sources in Sgr E having flat IR spectra are all radio-weak. For the sources with prominent radiocontinuum most have 18 cm/MSX E band flux ratios which are smaller than those of M17 or in Sgr B2. %9 \begin{figure} %\psfig{figure=RingF9.eps,height=12cm} \includegraphics[height=12cm]{RingF9.pdf} \caption[]{MSX intensities (W m$^{-2}$ sr$^{-1}$) on a spatial strip through the IRDC near l=0.25\degr, running from (l,b) = (0.4\degr,-0.1\degr) to (0.2\degr,0.1\degr). At top is the strip in Band A, as in Fig. 11 and at bottom are the absorption optical depth profiles in MSX bands A, C and E. Strip coordinates are labelled with the galactic latitude at positions along it (the arc length is sqrt(2) larger than the latitude pixel spacing) and the resolution is 21 sqrt(2)\arcsec\ = 0.00825\degr.} \end{figure} The spectra in Fig. 10 express in a quantitative way the visual impression gained from comparing the IR and radio images. To the extent that many objects appear to share the template spectral energy distribution seen in N10 and M17, the radio-IR comparison suggests an added optical depth as large as 4 mag over the easter portions of Sgr B. To the extent that the IR spectral slopes appear to steepen with apparent brightness, it would be inferred that the extinction is actually higher at longer wavelengths in those cases where the extinction is actually largest. To the extent that brighter IR sources (with their steeper IR spectra) also tend to have higher 22$\mu$/18cm flux ratios, it may be that the galactic center sources actually have a somewhat higher innate 22$\mu$/18 cm flux ratio, but are also more heavily extincted, so that the innate flux ratio tends to become more apparent for the less-heavily extincted sources. \section{The NSFR in the MW} The nuclear star-forming rings in barred galaxies are ongoing starbursts fed by the continuing inward flow of material along bar dust lanes into narrow regions of previously-accumulated, actively star-forming gas. Such rings naturally account for many of the properties which have been attributed to molecular gas in the inner regions of the Milky Way -- high internal density and thermal pressure over a very extended region, in an especially turbulent gas showing broad linewidths. The H II regions and molecular gas and their kinematics are not always discussed together but observational descriptions of a kinematic ring can be found in \cite{LisBur77}, \cite{Sof95}, \cite{OkaHas+96} and \cite{SawHas+04}; mechanisms of star formation in a local ring are discussed by \cite{StaMar+04} without reference to the Sgr HII regions and the connection between barring in the Milky Way and an inner ring are noted explicitly by \cite{Lis06Shred}, \cite{RodCom+06} and \cite{RodCom08}. How then is the material in the inner $\pm 175$ pc actually distributed? From our vantage point so near the mid-plane the 8$\mu$ light distribution in Fig. 1 is largely undifferentiated while the 22$\mu$ emission is at the other extreme, showing the seemingly isolated Sgr B, A, and C source complexes. However, the \coth\ distribution in Fig. 1 is somewhat hollow and loop-like at negative longitudes and the kinematics support this impression. Even from the crude sampling of recombination line emission in Fig. B.1 it is apparent that the profiles are narrow compared to the span of the rotational pattern across the central region. Although kinematic projection effects are important, this is the signature of a hollow, ring-like distribution, not a filled disk. %10 \begin{figure*} %\psfig{figure=RingF10.eps,height=13.2cm} \includegraphics[height=13.2cm]{RingF10.pdf} \caption[]{MSX-VLA, IR-radio spectra of galactic center sources. 18cm radio fluxes scaled upward by a factor 250 are shown at the extreme right side of each curve. The global spectrum of M17 \citep{PovSto+07} is shown in each panel as a template, arbitrarily scaled except at the upper left. Upper left: Integrated fluxes over extended regions. In this panel the M17 fluxes have been scaled to 8 kpc distance and divided by 2. Upper right: small sources in Sgr B1. Lower left: small sources in Sgr E. Lower right, small sources in Sgr B2.} \end{figure*} Shown in Fig. 11 are two collapsed longitude-velocity maps of the \coth\ emission (see Fig 1), where at each longitude the emission has been averaged either above or below the galactic equator (see also \cite{Sof95}). Emission associated with the Sgr source complexes generally appears in one of two kinematic ``branches'' which are manifested as nearly straight lines in the l-v plane, with similar velocity gradient (implying similar galactocentric radii) but very different longitude crossings at zero-velocity (implying a substantial component of non-circular motion, approximately 50 \kms). The branch at positive latitude is displaced to negative velocity at l=0\degr, crossing 0-velocity at positive longitude and including the locus of the Sgr B2 recombination lines at l $\ga$ 0.6\degr, v = 60 \kms. Another branch at negative latitude displays roughly symmetric behaviour, being displaced to positive velocity, crossing 0-velocity at negative longitude and meeting the locus of the Sgr C recombination lines at l=-0.5\degr, v = -60 \kms. An economical description of this behaviour puts the gas into a ring (or a pseudo-ring comprised of arm segments, see \cite{SawHas+04}) which is slightly tipped out of the galactic plane, so that the front and back portions are displaced in latitude. The outer size of the ring and its rotation speed are set by the locus of Sgr E at l=-1.25\degr\ corresponding to 175 pc, with a rotation speed of 220 \kms\ (since the ring is viewed tangentially). The line of sight inclination is small, 3.2\degr, i.e. a vertical displacment of $\pm0.07$\degr\ across a diameter of 350 pc . The near and far sides are also separated kinematically by a substantial ($\pm40-50$\kms) component of non-circular motion, perhaps indicating an elliptical shape to the ring. The ring in the Milky Way bar is somewhat on the small side, but not impossibly so, If it can be decided which of the kinematic branches is the near side, the sense of the non-circular motion, the absolute inclination and a full description of the geometry and kinematics are specified. \cite{SawHas+04} placed Sgr B on the near side, implying that the ring is seen above the equator on the near side, and that Sgr C is behind the center. This placement is broadly consistent with all of the observations, but not necessarily required in any aspect. It accounts for the bowed shape to the upper edge of the 8$\mu$ light and the presence of absorption against Sgr C at -100 \kms. Front placement also would enhance the visibility of the extinction associated with the 0-50 \kms\ gas around Sgr B, However it should be recalled that gas below -100 \kms\ is probably indigenous to the Sgr C complex anyway and the prominent extinction around Sgr B is for the most part seen against emission from within the Sgr B complex itself. The ring description is incomplete, especially inasmuch as it begs the question of the gas in Sgr A, much of which is found in the velocity range 20-70 \kms; the presence of gas at 50\kms\ near Sgr A* is clear but not understood. It seems unavoidable that some of the material at 20-70\kms\ which is presently ascribed to the vicinity of the center must instead arise in the ring. Velocities in around 50 \kms\ clearly occur in the ring and portions of the ring are unavoidably projected against Sgr A. Recalling the star formation activity in the nuclear rings of other galaxies, and imagining what how they would appear to observers embedded in the galactic midplane, it is inevitable that unrelated material would be viewed projected against the center. \begin{acknowledgements} %11 \begin{figure} \includegraphics[height=5.2cm]{RingF11.jpg} %\psfig{figure=RingF11.eps,height=10cm} \caption{\coth\ Kinematics of the star-forming ring. Shown are longitude- velocity diagrams of \coth\ emission, where at each longitude the emission has been averaged vertically, at 0\degr\ $<$ b $<$ 0.3\degr\ at top and over -0.3\degr\ $<$ b $<$ 0\degr\ at bottom. The locii of several H II regions and the locus of the lower-latitude kinematic branch of the ring are marked in the bottom panel. Some of this annotation has been carried over to the top panel.} \end{figure} The NRAO is operated by AUI, Inc. under a cooperative agreement with the US National Science Foundation. \end{acknowledgements} \appendix{} \section{MSX bands} The MSX bands and wavelengths are enumerated in Table 1 of the text. Figure A.1 shows how these bands overlap the extinction curves for galactic dust with \RV\ = 3.1 and 5.5 \citep{WeiDra01}. %A1 12 \begin{figure} %\psfig{figure=RingF12.eps,height=7.7cm} \includegraphics[height=7.7cm]{RingF12.pdf} \caption[]{Extinction cross-section per H-particle for galactic dust with \RV\ = 3.1 and \RV\ = 5.5 from \cite{WeiDra01}.} \end{figure} \section{Recombination Line Velocities} 19-GHz H70$\alpha$ recombination line profiles drawn from the work of \cite{Lis92} are shown in Fig. B.1. %B1 13 \begin{figure*} %\psfig{figure=RingF13.eps,height=15.5cm} \includegraphics[height=15.5cm]{RingF13.pdf} \caption[]{19-GHz H70$\alpha$ radio recombination line velocities toward H II regions seen over a range of longitudes within the star-forming ring \citep{Lis92}.} \end{figure*} \section{N10: An exemplary HII region bubble} N10, at a kinematic distance of 4.9 kpc is one of three exemplary HII region bubbles recently discussed by \cite{WatPov+08}. Its resemblance to Sgr C is obvious. %C1 14 \begin{figure*} %\psfig{figure=RingF14.eps,height=13cm} \includegraphics[height=13cm]{RingF14.png} \caption[]{MSX and MAGPIS 20cm images of N10. Upper left: MSX A band. Lower left: MAGPIS 20cm image \citep{HelBec+06}. Upper right: contours of 20cm radiocontinuum emission overlaid on the MSX A band image. Lower left: same for MSX E band. See also \cite{WatPov+08}. } \end{figure*} \bibliographystyle{apj} \bibliography{mnemonic,shredded} \end{document} -- Harvey S. Liszt work:+1 434.296.0344 fax 0278 Scientist & Spectrum Manager home:+1 434.973.3744 National Radio Astronomy Observatory cell:+1 434.227.6356 (+41792953708) 520 Edgemont Road mailto:hliszt@nrao.edu Charlottesville, VA 22903-2475 http://www.cv.nrao.edu/~hliszt "Be Kind To Radio Astronomy" http://www.nrao.edu/~hliszt/RFI/