L1448 IRS 2: A HIRES‐identified Class 0 Protostar

THE ASTROPHYSICAL JOURNAL, 515 : 696È705, 1999 April 20 ( 1999. The American Astronomical Society. All rights reserved. Printed in U.S.A.

L1448 IRS 2 : A HIRES-IDENTIFIED CLASS 0 PROTOSTAR JOANN OÏLINGER,1,2 GRACE WOLF-CHASE,1,3 MARY BARSONY,1,4,5 AND DEREK WARD-THOMPSON6 Received 1998 August 7 ; accepted 1998 December 1

ABSTRACT We present far-infrared and submillimeter images and millimeter-continuum photometry for L1448 IRS 2, a low-luminosity (5.2 L ), embedded young stellar object discovered by IRAS. New far-infrared _ maps were produced from the archival IRAS data via high-resolution (HIRES) processing, the name given to the implementation of the maximum correlation method for image construction from the IRAS data streams. The HIRES-processed images of L1448 IRS 2 presented here have e†ective resolutions of D35@@ ] 28@@, D35@@ ] 37@@, and D45@@ ] 40@@ at 25, 60, and 100 km, respectively, which represent an order of magnitude improvement over previously available far-infrared maps. L1448 IRS 2 was mapped at 450 and 850 km, and we acquired continuum photometry at 1.3 mm. With these new data, we plot the most complete spectral energy distribution for L1448 IRS 2 to date. We also present sensitive 12CO J \ 1 ] 0 emission-line maps of a 47@ ] 7@ area of the L1448 dark cloud, centered on L1448 IRS 2, obtained with the on-the-Ñy imaging capability of the NRAO 12 m telescope. As a direct consequence of this mapping, we have discovered a parsec-scale molecular outÑow associated with L1448 IRS 2. The combined continuum and molecular line data conÐrm the classiÐcation of L1448 IRS 2 as a new member of the rare, short-lived (a few ] 104 yr), Class 0 phase of protostellar evolution. The use of HIRES imaging of IRAS data as a promising tool for identifying protostellar candidates is demonstrated by the example of L1448 IRS 2. Subject headings : circumstellar matter È infrared : stars È stars : imaging È stars : individual (L1448 IRS 2) È stars : preÈmain-sequence 1.

INTRODUCTION

driven by these youngest sources are more highly collimated and are generally an order of magnitude more powerful than the older outÑows powered by the more evolved, Class I protostars (e.g., Bontemps et al. 1996 ; Barsony et al. 1998 ; Wolf-Chase et al. 1998a). A popular method for discovering new Class 0 objects has been via follow-up studies of bipolar outÑows with no known central exciting source (e.g., Sanders & Willner 1985 ; Andre et al. 1990 ; Bachiller et al. 1990 ; WardThompson, Eiroa, & Casali 1995 ; Wolf-Chase et al. 1998a). An alternative method for identifying new Class 0 candidates is by comparison of near-infrared (NIR) images of embedded young clusters with their corresponding largescale, highÈangular resolution millimeter/submillimeter continuum maps. Submillimeter/millimeter continuum sources without NIR counterparts are ideal Class 0 candidates. To date, only a few nearby star-forming regions have been mapped over sufficient areas and at adequate spatial resolutions in both wavelength ranges to make such comparisons possible (e.g., o Oph : Barsony et al. 1997 ; Motte, Andre, & Neri 1998 ; Serpens : Casali, & Eiroa 1992 ; Casali, Eiroa, & Duncan 1993 ; Giovanetti et al. 1998 ; Davis et al. 1998 ; Testi & Sargent 1998). However, the far-infrared (FIR) portion of the electromagnetic spectrum is also important for star formation studies, being the wavelength regime in which the radiation from the youngest protostars reaches its peak. Because of the extreme opacity of the terrestrial atmosphere at FIR wavelengths, FIR observations require either airborne or satellite platforms. The all-sky surveys of IRAS at 12, 25, 60, and 100 km therefore provide a unique resource for identifying and characterizing protostars, especially in view of the recent software advances in high-resolution (HIRES) processing that allow unprecedented spatial resolutions to be achieved at the longest FIR wavelengths (Aumann, Fowler, & Melnyk 1990 ; Surace et al. 1993 ; Terebey &

The advent of submillimeter astronomy allowed the youngest known protostellar stage to be identiÐed, since the youngest protostars (also known as Class 0 sources) generally remain undetected shortward of 10 km (Andre, Ward-Thompson, & Barsony 1993 ; Barsony 1994 ; Ward-Thompson 1996). Class 0 sources have massive dust (A D 1000) envelopes, some of which have been shown to beVinfalling (Walker et al. 1986 ; Zhou et al. 1993 ; Zhou 1995 ; Lehtinen 1997 ; Myers et al. 1995, 1996 ; MoriartySchieven et al. 1995 ; Mardones et al. 1997 ; Hurt, Barsony, & Wootten 1996 ; Ward-Thompson et al. 1996 ; Gregersen et al. 1997). The ages of these objects may be estimated by determining an appropriate range of mass infall rates, M0 , along with constraints on the core masses of the protostars, M , derived from the observed bolometric luminosities *  et al. 1993) ; clearly the age of such an object is on the (Andre order of M /M0 . Class 0 protostars are short-lived (a few * therefore, quite rare (Barsony 1994). Since ] 104 yr) and, the bulk of the zero-age main-sequence mass of a star is assembled during this brief evolutionary phase (Andre & Montmerle 1994), there is much interest in identifying and closely studying members of this class. All known Class 0 sources also display bipolar molecular outÑows. OutÑows 1 University of California, Riverside, Department of Physics, Riverside, CA 92521. 2 Present Address : Jet Propulsion Laboratory, IPAC MS 100-22, 770 South Wilson, Pasadena, CA 91125. 3 Present Address : Adler Planetarium and Astronomy Museum, 1300 South Lake Shore Drive, Chicago, IL 60605. 4 NSF CAREER Award Recipient. 5 NSF POWRE Visiting Professor, Physics Department, Harvey Mudd College, Claremont, CA 91711. 6 University of Wales Cardi†, Department of Physics & Astronomy, P.O. Box 913, Cardi† CF2 3YB, UK.

696

L1448 IRS 2 : HIRES-IDENTIFIED PROTOSTAR Mazzarella 1994 ; Cao et al. 1996, 1997 ; Hurt & Barsony 1996 ; Barsony et al. 1998). In this paper, we suggest HIRES processing of the IRAS data as an alternative method of identifying Class 0 sources. We demonstrate the utility of this technique for the case of L1448 IRS 2 (\IRAS 03222]3034), which was Ðrst identiÐed as a protostellar candidate via HIRES imaging of the L1448 dark cloud (OÏLinger 1997). The L1448 dark cloud in the Perseus molecular complex (distance B300 pc) extends in C18O D 1.3 ] 0.7 pc and contains 100 M (Bachiller & Cernicharo 1986b). The _ ammonia core is restricted to a 1 ] 0.5 pc area containing 50 M (Bachiller & Cernicharo 1986b). Three objects _ with L1448 are listed in the IRAS Point Source associated Catalog (IRAS PSC), two of which reside in the ammonia core, L1448 IRS 2 and L1448 IRS 3 (Bachiller & Cernicharo 1986b). The study of L1448 IRS 2 has largely been neglected because of the proximity of its brighter, more famous neighbors, all associated with the IRAS PSC source L1448 IRS 3 : L1448C, whose high-speed, highly collimated outÑow has been an object of much study ; and the Class 0 binary L1448N(A) ] L1448N(B), each member of which drives its own separate outÑow (see Barsony et al. 1998 and references therein). L1448 IRS 2 remains undetected at 12 km, has a rising spectral energy distribution (SED) from 25 to 60 km, and, because of source confusion, has only a 100 km upper limit listed in the IRAS PSC. A water maser, a hallmark of outÑows from low-luminosity young stellar objects (YSOs), was found to be associated with this source (Anglada et al. 1989 ; Persi, Palagi, & Felli 1994). L1448 IRS 2 was included in the comprehensive NIR imaging survey of embedded YSOs carried out by Hodapp (1994), where no corresponding 2 km point source is found to K@ ¹ 16.5 mag, although a faint cone of nebulosity opens toward the northwest from the IRAS position. In their large Ðeld-of-view, optical imaging survey of L1448 and L1455, Bally et al. (1997) reported the discovery of numerous clusters of Herbig-Haro objects, many at surprisingly large distances from their exciting sources. These authors note that the HH195 complex is excited by a highly collimated outÑow emanating from L1448 IRS 2, with a rough coincidence between the optical emission and a chain of knots seen in shock-excited molecular hydrogen line emission (Hodapp 1994). The cloud rest velocity in the L1448 ammonia core has two components, one at V \ 4.7 km s~1 associated LSR at 4.2 km s~1, which is with the L1448 IRS 3 core, the other associated with the L1448 IRS 2 core (Bachiller & Cernicharo 1986b). Weak, redshifted (V \ 7 km s~1) CO J \ 1 ] 0 emission has previously LSR been reported in the vicinity of L1448 IRS 2, with the bulk of this emission emanating in a southeasterly direction (Bally et al. 1997). In order to illuminate the true nature of this highly embedded protostellar candidate, we have undertaken a comprehensive study of L1448 IRS 2 using HIRES processing of the IRAS data, submillimeter continuum imaging with the Submillimeter Common-User Bolometer Array (SCUBA) on the 15 m James Clerk Maxwell Telescope (JCMT), and millimeter continuum photometry and largescale on-the-Ñy (OTF) molecular line mapping in CO J \ 1 ] 0 at NRAOÏs 12 m telescope at Kitt Peak, Arizona. The observations and data processing are described in ° 2, the results and discussion are presented in ° 3, and the conclusions are summarized ° 4.

2.

697

OBSERVATIONS AND DATA REDUCTION

2.1. FIR Mapping : HIRES Processing of the IRAS Data During 10 months in 1983, IRAS was able to complete three nearly total sky surveys at four di†erent wavelengths : 12, 25, 60, and 100 km. The survey instrument aboard the 0.6 m infrared telescope consisted of two separate arrays of detectors for each of the four wave bands. Each array consisted of either seven or eight detectors, with the average detector sizes being 45@@ ] 267@@, 45@@ ] 279@@, 90@@ ] 285@@, and 180@@ ] 303@@, for 12, 25, 60, and 100 km, respectively. Detector responses were read out every [email protected] along the scan direction : each such data point is called a ““ footprint. ÏÏ Several steps are involved in producing high-resolution images from the IRAS detector responses. First, the raw data are run through the LAUNDR program, which corrects for cosmic-ray hits and detector glitches and calibrates the data. An unwanted, troublesome artifact that results from imperfect detector calibrations across adjacent tracks is ““ striping. ÏÏ Destriping algorithms are now implemented in both the LAUNDR package, and in the image construction software YORIC to minimize this problem (Cao et al. 1996, 1997). The data are also baseline subtracted and a Ñux bias is applied to prevent clipping of negative noise in the subsequent processing. The LAUNDR data are then input to the YORIC program, developed at IPAC,7 which implements HIRES processing using the maximum correlation method (MCM ; Aumann et al. 1990). We used YORIC v1.87 for processing of the L1448 data. At each iteration, YORIC creates simulated IRAS detector responses to an input image, compares these simulated responses with the actual detector responses recorded by IRAS in the area of sky to be imaged, and calculates multiplicative correction factors for each image pixel (the default pixel size being 15A) based on statistical correlations. The calculated correction factors are then applied to the input image, which becomes the next output image for further iterations of the algorithm. Default HIRES processing, which can be requested remotely via electronic mail to IPAC, typically halts at twenty iterations of the MCM algorithm. The default mode does not use the convergence acceleration feature of YORIC, so the iterative procedure converges logarithmically. The result is that most of the spatial information is obtained early on, and later iterations progress very slowly, which is a partial justiÐcation for selecting a general-purpose limit of 20 iterations. When the requesting investigator has more time available to use the additional tools provided in the YORIC program (beam sample maps, prior knowledge input, simulation mode, etc.), then practically any case can be taken beyond 20 iterations. One situation in which iterating further is usually appropriate is that in which the scans di†er considerably in direction on the sky (J. W. Fowler 1998, private communication). Examples of signiÐcant improvements in image resolution past 20 iterations may be seen in the literature : Aumann et al. (1990) ; Hurt & Barsony (1996) ; OÏLinger (1997) ; and Barsony et al. (1998). For the data presented here, we halted HIRES processing at 40 iterations in the 12 km and 25 km bands and at 120 iterations at 60 and 100 km. These images do not represent ““ converged ÏÏ data in a formal sense, as the concept of convergence 7 IPAC is funded by NASA as part of the IRAS extended mission under contract to JPL.

698

OÏLINGER ET AL.

in iterative solutions of nonlinear problems seldom has any absolute meaning (J. W. Fowler 1998, private communication). Our iteration limits in the various wave bands were chosen by looking at the change in the correction factor variance from one iteration to the next and setting an acceptable threshold value (for further information, see Aumann et al. 1990 ; Hurt & Barsony 1996 ; OÏLinger 1997 ; Barsony et al. 1998). The irregular sampling of the sky brightness distribution by subsequent passes of the IRAS detectors results in variable spatial resolution across a HIRES-processed image. We therefore quote ““ e†ective ÏÏ resolutions at each wave band, which is obtained by averaging point-source responses sampled on a rectangular grid throughout each image. The e†ective resolutions of our HIRES-processed images of the L1448 Ðeld are D35@@ ] 28@@ at 12 and 25 km, D35@@ ] 37@@ at 60 km, and D45@@ ] 40@@ at 100 km. The standard deviations from these values range from 14% to 22%. The e†ective resolutions at 100 km are approaching the di†raction limit. The default pixel size of 15A was used. Absolute calibration uncertainty for HIRES-derived Ñuxes is estimated to be 20% (Levine & Surace 1993). 2.2. Submillimeter Continuum Mapping : JCMT Observations Submillimeter continuum observations at 450 km and 850 km were carried out at JCMT8 on Mauna Kea, Hawaii on the morning of 1997 October 3 from UT 11 : 45 until 12 : 45. The detector used was the common-user array camera, SCUBA (Holland et al. 1999), which was used in its 64 position jiggling mode to make a fully sampled image of a [email protected] Ðeld simultaneously at 450 and 850 km. On-source integration times were 10 s point~1 in each of the 64 positions, which corresponds to 40 s beam~1 area at each wavelength. The whole observation was then repeated with a slight o†set so that the 5 pixels with signiÐcantly above average noise could be removed from the subsequent data reduction without leaving areas of the map without data. The observations were carried out while using the secondary mirror to chop 120A in azimuth at around 7 Hz and synchronously to detect the signal, thus rejecting ““ sky ÏÏ emission. The atmospheric opacity at 225 GHz was 0.068, which is typical of fairly good conditions at the site and corresponds to a zenith atmospheric transmission at 850 km of around 80%. The source was above 70¡ in elevation throughout the observations. The opacity at both 850 and 450 km was monitored by ““ skydips, ÏÏ and the opacity at 1.3 mm was monitored by the radiometer located at the Caltech Submillimeter Observatory. Pointing and focusing were checked using the bright radio source 3C84 immediately before and after the observations of L1448, and the pointing was found to be good to D1A. Calibration was performed using the planet Uranus (Griffin et al. 1986 ; Orton et al. 1986 ; Griffin & Orton 1993), which was observed approximately 4 hr earlier, before it set, under almost identical observing conditions, since the sky opacity remained stable throughout this period. The calibration was cross checked using the secondary calibration source GL 618 (Sandell 1994), and a consistent calibration 8 The James Clerk Maxwell Telescope is operated by the Joint Astronomy Center, Hawaii, on behalf of the UK PPARC, the Netherlands NWO, and the Canadian NRC. SCUBA was built by the Royal Observatory Edinburgh.

Vol. 515

was obtained. We estimate the total absolute calibration uncertainty to be D30% at 450 km and D20% at 850 km. The maps of Uranus show that the JCMT error beam at 450 km is only signiÐcant below the 10% level of peak, and above this the beam is essentially circular. At 850 km the error beam is only signiÐcant below the 5% level. The structure we observe in L1448ÈIRS 2 is considerably above this level and is morphologically dissimilar to the structure associated with the error beam. 2.3. NRAO 12 m Data We obtained 1.3 mm continuum data, 115 GHz CO J \ (1 ] 0) and 110 GHz 13CO J \ (1 ] 0) spectral line data toward L1448 IRS 2 during 1997 April and June using NRAOÏs 12 m telescope located on Kitt Peak, near Tucson, Arizona.9 2.3.1. Continuum Photometry

We measured the 1.3 mm continuum Ñux toward L1448 IRS 2 within a 27A beam using a dual-channel, doublesideband, SIS heterodyne receiver system. The receiver had a bandwidth of 600 MHz and was operated at a sky frequency of 231.6 GHz. The subreÑector was nutated at a frequency of 4 Hz using a 2@ beam throw. Data were calibrated by chopping between sky and an ambient temperature load. Because of the inopportune positions of the planets, absolute calibration was achieved by observing L1551 IRS 5 and using published 1.3 mm Ñuxes to scale our data (Sandell 1994 ; Walker, Adams, & Lada 1990). Consequently, we estimate 1.3 mm Ñux uncertainty to be at the 30%È40% level. 2.3.2. Spectral-L ine OT F Mapping

In order to study the high-velocity outÑow gas associated with L1448 IRS 2, we made a 47@ ] 7@ map of the CO J \ (1 ] 0) emission along a position angle of 135¡ (measured east from north) centered on L1448 IRS 2, using the NRAO 12 m telescope in the spectral-line OTF mapping mode. This technique allows the acquistion of large-area, high-sensitivity, spectral line maps with unprecedented speed and pointing accuracy. These observations were made using a dual-channel, single-sideband SIS receiver with a beamwidth B55A at this frequency. The back end consisted of 250 and 500 kHz resolution Ðlterbanks, yielding velocity resolutions of 0.65 and 1.3 km s~1, respectively. Single-pointing observations of the CO J \ (1 ] 0) and 13 CO J \ (1 ] 0) transitions were obtained at selected positions to check CO optical depths in the line wings. Line temperatures at the 12 m are on the T * scale and must be divided by the corrected main-beamR efficiency, g*, to m For convert to the main-beam brightness temperature scale. our very extended source, g* B 1.0. Since the corrected m main-beam efficiency is the fraction of the forward power in the main di†raction beam relative to the total forward power in the main beam plus error beam, contributions from the error beam can make g* [ 1.0. At 115 GHz, the m theoretical error beamwidth is B17@, but the ratio of the error-beam amplitude to the main-beam amplitude is only 6 ] 10~4, suggesting contributions from the error beam can be ignored. The RMS attained in each spectrum was T * B R 0.11 K. 9 NRAO is a facility of the National Science Foundation, operated under cooperative agreement by Associated Universities, Inc.

No. 2, 1999

L1448 IRS 2 : HIRES-IDENTIFIED PROTOSTAR 3.

699

RESULTS AND DISCUSSION

3.1. Continuum Maps and Photometry 3.1.1. IRAS Images The IRAS images of a [email protected] ] [email protected] Ðeld containing L1448 IRS 2 are displayed in Figures 1 and 2. Figures 1a and 1c show the images obtained after one iteration of the HIRES algorithm at 12 km and 25 km, respectively. Such images are also known as FRESCO (full-resolution co-add) images. Figures 1b and 1d show the corresponding HIRESprocessed images after 40 iterations. Figures 2a and 2c show the FRESCO images at 60 km and 100 km, respectively, while Figures 2b and 2d show the corresponding HIRESprocessed images after 120 iterations. The stars in Figures 1 and 2, from west to east, denote the positions of L1448 IRS 2, L1448N, and L1448C, respectively. The IRAS PSC source known as L1448 IRS3 is actually associated with two ammonia emission peaks known as L1448N and C. These are the two extended emission peaks to the east in Figures 1 and 2. On closer scrutiny,

FIG. 2.È60 and 100 km IRAS images of the L1448 cloud core. (a) shows the FRESCO image (equivalent to one iteration of HIRES) at 60 km. The contours go from 12 to 44 MJy sr~1 in steps of 4.0 MJy sr~1. (b) shows the HIRES image (after 120 iterations) at 60 km. Plotted contour levels are at 30, 60, 90, 120, 150, 180, 240, 300, 360, and 400 MJy sr~1. (c) shows the FRESCO image (equivalent to one iteration of HIRES) at 100 km. Contour levels start at 45 MJy sr~1 and are incremented by 3.0 MJy sr~1. (d) shows the HIRES image (after 120 iterations) at 100 km. Because of dynamic range problems in contouring, L1448 IRS 2 is contoured at 30, 34, 38, and 42 MJy sr~1, whereas the lowest contour for the structure encompassing L1448N and L1448C starts at 70 MJy sr~1 and continues at 80, 90, 100, 100, 130, 150, and 170 MJy sr~1. The 1 p noise levels are at D3.2 and 7.5 MJy sr~1 in the 60 and 100 km HIRES maps, respectively. The positions of L1448 IRS 2, L1448N, and L1448C are indicated.

FIG. 1.È12 and 25km IRAS images of the L1448 cloud core. (a) shows the FRESCO image (equivalent to one iteration of HIRES) at 12 km. Contour levels start at 0.3 MJy sr~1 and are incremented by 0.1 MJy sr~1. (b) shows the HIRES image (after 40 iterations) at 12 km. Contour levels start at 4.0 MJy sr~1 and are subsequently spaced by 2.0 MJy sr~1. (c) shows the FRESCO image (equivalent to one iteration of HIRES) at 25 km. Contour levels start at 1.0 MJy sr~1 and are subsequently spaced by 1.0 MJy sr~1. (d) shows the HIRES image (after 40 iterations) at 25 km. Contour levels are plotted at 3, 4, 5, 10, 30, 50, 70, 110, 150, and 190 MJy sr~1, with the peak value in the map (at L1448N) at 220 MJy sr~1. The 1 p noise levels are at D1.3 and 2.0 MJy sr~1 in the 12 and 25 km HIRES maps, respectively. The positions of L1448 IRS 2, L1448N, and L1448C are indicated.

L1448N has been found to contain two embedded Class 0 objects, L1448N(A), L1448N(B), and the dense condensation L1448 NW, whose nature remains to be determined (see Barsony et al. 1998 and references therein). Although two smaller peaks are associated with the L1448N position in the HIRES images shown in Figures 1b, 1d, and 2b, the peak to the southwest marked by the star represents the location of both L1448N(A) and (B), which are members of a binary system, with a separation distance of only D7A (Barsony et al. 1998 and references therein). The HIRES e†ective beams are on the order of 32A at 12 and 25 km and 36A at 60 km, so it is not possible to resolve these objects. The nature of the other source, which appears as an emission ““ Ðnger ÏÏ to the northeast of the binary at both 12 and 25 km and dominates the L1448N contour map as a distinct peak at 60 km, remains unknown. The southeastern source in Figures 1 and 2 is the famous Class 0 protostar, L1448C. L1448 IRS 2 remains undetected to a 3 p level of 0.14 Jy at 12 km and makes its Ðrst appearance at 25 km (see Figs.

700

OÏLINGER ET AL.

1b and 1d). By 60 km, L1448 IRS 2 signals its presence even in the low-resolution FRESCO image, as the westward extension of the lowest contour levels (see Fig. 2a). However, the great power of HIRES processing is showcased in the 100 km images (Figs. 2c and 2d). Whereas L1448 IRS 2 is indistinguishable in the low-resolution FRESCO 100 km image (Fig. 2c), it is clearly resolved as a separate source in the 120 iteration HIRES image (Fig. 2d). This is the Ðrst time that L1448 IRS 2 has been identiÐed as a distinct 100 km source. Our derived Ñuxes and Ñux upper limits for L1448 IRS 2 from these data are listed in Table 1, as are the corresponding IRAS PSC values. 3.1.2. SCUBA Images

The calibrated SCUBA images of L1448 IRS 2 are shown in Figures 3a (450 km) and 3b (850 km). The peak of the submillimeter continuum emission at each wavelength coincides with the position of L1448 IRS 2 (within the errors) as determined from the IRAS PSC. In Table 2 we present a new determination for the position of L1448 IRS 2 derived from elliptical Gaussian Ðts to the peak emission at 450 km and 850 km. Unlike L1448C, L1448 IRS 2 is found to be embedded in an extended, well-resolved dust continuum structure reminiscent of L1448N that harbors multiple sources (see Barsony et al. 1998). For this reason, we list submillimeter Ñuxes for L1448 IRS 2 in several di†erent apertures (see Table 1). For purposes of determining the sourceÏs SED (see ° 3.2 below), we use a 20A aperture centered on L1448 IRS 2 for the Ñux determinations. Furthermore, we note that L1448 IRS 2 is not centered in the extended emission from which the majority of the observed submillimeter Ñux is found to originate (see Table 1). Intriguingly, much of this extended emission originates from a position B40A due east of L1448 IRS 2, which is in the general direction of L1448N.

Vol. 515 TABLE 1 FLUXES FOR L1448 IRS 2

j (km)

Beam

Flux (Jy)

Previous Flux (Jy)

12a . . . . . . . 25a . . . . . . . 60a . . . . . . . 100a . . . . . . 450 . . . . . . . 450 . . . . . . . 450 . . . . . . . 850 . . . . . . . 850 . . . . . . . 850 . . . . . . . 1300 . . . . . .

35@@ ] 28@@ 35@@ ] 28@@ 35@@ ] 37@@ 45@@ ] 40@@ 20A aperture 40A aperture Entire mapped area 20A aperture 40A aperture Entire mapped area 27A FWHM

¹0.14 0.59 ^ 0.12 14.9 ^ 3.1 40.0 ^ 8.5 12.7 ^ 2.5 19.4 ^ 3.9 34.2 ^ 6.8 1.86 ^ 0.4 3.01 ^ 0.6 8.84 ^ 1.7 0.67 ^ 0.2

¹0.3 0.65 ^ 0.08 14 ^ 2 ¹167 ... ... ... ... ... ... ...

a Bachiller & Cernicharo 1986b.

For reference, the position angle of the L1448 IRS 2 molecular outÑow (see ° 3.3 below) is 133¡. The outÑow orientation, therefore, is in a completely di†erent direction from the extended continuum emission found to lie to the east of L1448 IRS 2, which may therefore be signaling the presence of a neighboring prestellar condensation. 3.2. SED and Derived Source Properties of L 1448 IRS 2 The SED of a source is required both for classiÐcation purposes and for estimation of physical parameters of interest. Therefore, we present the SED for L1448 IRS 2 in Figure 4, along with a single-temperature, modiÐed blackbody Ðt of the form (1) S \ B (T )(1 [ e~ql )d) , l l d assuming a 20A source diameter and a l1.5 wavelength dependence of the dust optical depth (Andre et al. 1993).

FIG. 3.ÈSCUBA 450 and 850 km images of L1448 IRS 2. (a) (450 km ; left panel) and (b) (850 km ; right panel) show the SCUBA images of L1448 IRS 2. The (0, 0) map position is at a \ 03h22m18s, d \ 30¡ 34@35A. In (a), contours start at 1.0 Jy per 7A beam, incremented by 1 Jy beam~1, with the 1 p noise 1950 1950 level at 0.22 Jy beam~1. In (b), the lowest contour is 200 mJy per 14A beam, incremented by 200 mJy beam~1 contours, with the 1 p noise level at 30 mJy beam~1.

No. 2, 1999

L1448 IRS 2 : HIRES-IDENTIFIED PROTOSTAR

701

TABLE 2 L1448 IRS 2 SOURCE AND OUTFLOW PROPERTIES Parameter

Symbol and Units

Value

Source L1448 IRS2a RA(1950) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . L1448 IRS2a Dec(1950) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Distanceb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ModiÐed blackbody Ðt temperature . . . . . . . . . . . . . . . Fit optical depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fit source diameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bolometric luminosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

D (pc) T (K) d q 250 km d) (arcsec) L (L ) bol _ L /L submm bol L /(103 ] L ) bol 1.3mm M (M ) env _ M *

Circumstellar envelope mass . . . . . . . . . . . . . . . . . . . . . . . . Hydrostatic core mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

03h22m17s. 9 30¡34@41A 300 22 0.032 20 5.2 0.03 6.6 0.86 0.05c

OutÑow Assumed CO abundance . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characteristic velocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unconfused Ñow region radial extentd . . . . . . . . . . . . . Unconfused Ñow dynamical timescaled . . . . . . . . . . . . Unconfused outÑow massd . . . . . . . . . . . . . . . . . . . . . . . . . . Unconfused outÑow mechanical luminosityd . . . . . . Total Ñow extent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total Ñow mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OutÑow force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Opening angle (1/2 angle of cone) . . . . . . . . . . . . . . . . . . OutÑow position angle (east of north) . . . . . . . . . . . . .

CO/H 2 V (km s~1) char R (pc) min t \ R /V (103 yr) D min char (M ) _ L \ E/t (L ) mech D _ R (pc) flow M (M ) flow _ F \ P/t (M km s~1 yr~1) obs D _ / (deg) P.A. (deg)

1.1 ] 10~4 15.0 0.09 5.8 0.09 0.3 2.6 0.7 2.4 ] 10~4 13.5 133

a Location of 450 and 850 km continuum emission peak. b Bachiller & Cernicharo 1986a. c Assuming an infall rate of 10~5 M yr~1 onto a 3 R hydrostatic core. _ _ d These outÑow parameters were derived from a small region near the source where the emission could be unambiguously disentangled from high-velocity gas powered by the other Class 0 sources in the vicinity (see Barsony et al. 1998 ; Wolf-Chase et al. 1998b).

The values of the dust temperature, T , and the 250 km d listed in Table 2. optical depth, q , derived from this Ðt are 250 The source properties that can be derived from the model Ðt to the SED of L1448 IRS 2 as plotted in Figure 4, such as the source bolometric luminosity, L , the circumstellar bol mass, M , for an mass, M , and the inferred central source c * were assumed infall rate, M0 , are also listed in Table 2 and derived in a manner analogous to that described for other

Class 0 sources in Perseus (Barsony et al. 1998). The tabulated bolometric luminosity was derived by numerical integration under the Ðtted curve plotted in Figure 4. Class 0 protostars have ratios of L /L º 5 ] 10~3 submm bol ratios for and L /(103 ] L ) ¹ 20. The values of these bol 1.3 L1448 IRS 2 are 0.03 and 6.6, respectively (see Table 2), falling well within the Class 0 protostellar range (Andre et al. 1993 ; Barsony 1994 ; Ward-Thompson 1996).

FIG. 4.ÈSED of L1448 IRS 2 : Ðlled triangles represent new IRAS photometry (this work), Ðlled circles represent Ñuxes in 20A diameter apertures from SCUBA (this work), and the cross represents the 1.3 mm point from NRAO 12 m continuum photometry (this work). The plotted Ñuxes and apertures are listed in Table 1. Parameters of the plotted Ðt are listed in Table 2.

3.3. T he NRAO 12 m OT F Map of the L 1448 IRS 2 CO J \ 1 ] 0 OutÑow L1448 IRS 2 was Ðrst identiÐed in a search for IRAS sources associated with ammonia cores in L1448 (Bachiller & Cernicharo 1986b). Until now, no data on the actual source properties of L1448 IRS 2, apart from its IRAS PSC position and Ñuxes, had been available. The FIR, submillimeter, and millimeter maps and photometry presented above are consistent with the identiÐcation of L1448 IRS 2 as a Class 0 protostellar candidate. The last remaining criterion to be met by this source in order to merit Class 0 designation is that it drive its own powerful bipolar molecular outÑow (e.g., Bontemps et al. 1996 ; Barsony et al. 1998 ; Wolf-Chase et al. 1998a). This motivation led us to map the vicinity of L1448 IRS 2 in the CO J \ 1 ] 0 transition in order to search for the presence of high-velocity molecular gas. The previous evidence hinting at the presence of outÑow activity powered by L1448 IRS 2 is shown in Figure 5, in which the gray scale shows the K@ image in the immediate

702

OÏLINGER ET AL.

Vol. 515

FIG. 5.ÈK@ image of L1448 IRS 2 detail of an NIR image (from Hodapp 1994) showing the faint reÑection nebulosity and shocked H emission knots emanating from L1448 IRS 2. Crosses mark the positions of the optically visible Herbig-Haro knots HH 195 A[E (Bally et al. 1997).2 The three lines intersecting the source, L1448 IRS 2, demarcate the symmetry axis and the opening angle of the blueshifted lobe of the large-scale bipolar CO J \ 1 ] 0 outÑow driven by L1448 IRS 2. The Ðve-pointed stars indicate the IRAS PSC positions of L1448 IRS 1 (the brightest source in the Ðeld, associated with the wide opening angle, fan-shaped nebula) and of L1448 IRS 2 (invisible at K@).

vicinity of L1448 IRS 2 (Hodapp 1994). The brightest source in the Ðeld is L1448 IRS 1, with its extended, large opening angle reÑection nebulosity. The IRAS PSC position of L1448 IRS 1 is indicated by the open Ðve-pointed star, which is slightly o†set from the bright NIR point source. The IRAS PSC position of L1448 IRS 2 is indicated by the Ðve-pointed star at the intersection of the indicated lines. Note the complete lack of NIR emission at the position of L1448 IRS 2 (to K@ D 16.5), as would be expected for a Class 0 protostar. A faint, fan-shaped NIR reÑection nebulosity centered on L1448 IRS 2 is present, however. The full opening angle (27¡) of the conical reÑection nebulosity centered on L1448 IRS 2 and its symmetry axis, at P.A. \ 133¡, are indicated by the intersecting diagonal lines in Figure 5. The K@ Ðlter (j \ 2.11, *j \ 0.18 km) includes 0 S(1) emission lines. Various the Brc and the H 2.12 km 2 knots of shock-excited molecular hydrogen emission appear as the nonstellar emission features, both along the indicated symmetry axis and along the lines indicating the opening angle of the reÑection nebulosity. The positions of a recently discovered string of optically visible Herbig-Haro knots (HH 195) are indicated by pluses (Bally et al. 1997). These knots lie along the symmetry axis of the NIR nebula and point back to L1448 IRS 2. Our CO J \ 1 ] 0 outÑow map centered on L1448 IRS 2 is presented in Figure 6. The mapped area covers D47@ ] 7@ (corresponding to B4 ] 0.6 pc at the source) at a position angle of 135¡ (measured east from north) and is outlined by the zigzagged boundary lines. This map could not have

been produced without the spectral line OTF mapping technique, which is a relatively new capability at the NRAO 12 m telescope. This observing mode allows the mapping of large regions of sky at unprecedented sensitivity in relatively small amounts of observing time. The solid contours of Figure 6 represent redshifted emission integrated over the velocity interval 8.1 ¹ V ¹ 17.8 LSR km s~1, whereas the dashed contours represent blueshifted emission integrated over [12.1 ¹ V ¹ [1 km s~1. These velocity intervals were chosen LSR from inspection of individual spectra as well as channel maps (maps produced from each velocity bin), which helped delineate outÑow emission from the ambient cloud emission over the mapped region. The cloud rest velocity in the L1448 ammonia core is at V \ 4.5 ^ 0.25 km s~1 (Bachiller & Cernicharo 1986b). LSR Progressing from west to east, the solid Ðve-pointed stars in Figure 6 demarcate the positions of L1448 IRS 1, L1448 IRS 2, L1448N, and L1448C, respectively. The protostellar binary within L1448N, L1448N(A) ] L1448N(B), cannot be resolved on this scale. The relative size of the 55A beam is indicated in the lower right-hand corner. Pluses indicate the positions of Herbig-Haro objects (Bally et al. 1997) excited by the blueshifted outÑow. The L1448 IRS 2 outÑow symmetry axis (the same as in Fig. 5) is indicated by the line running the length of the map in Figure 6, and the opening angle, deÐned by the Hodapp nebulosity in Figure 5, is indicated by the lines crossing through the central protostellar position.

No. 2, 1999

L1448 IRS 2 : HIRES-IDENTIFIED PROTOSTAR

703

FIG. 6.ÈCO line wing map of the L1448 IRS 2 outÑow map of the high-velocity CO J \ 1 ] 0 emission centered on L1448 IRS 2 : solid contours indicate redshifted gas, and dashed contours indicate blueshifted gas. The 47@ ] 7@ map extent is indicated by the large zigzagged box. The high-velocity redshifted emission has been integrated over the velocity range ]8.1 ¹ V ¹ ]17.8 km s~1, whereas the high-velocity blueshifted emission has been integrated over LSRfor both blueshifted and redshifted emission begin at 2 K km s~1 (B3 p) and increase in the velocity range [12.1 ¹ V ¹ [1 km s~1. Contour levels LSR intervals of 1.5 K km s~1. Proceeding from west to east, the positions of L1448 IRS 1, L1448 IRS 2, L1448N, and L1448C are indicated by the solid stars. The 55A beam size is indicated in the lower right corner. The position angles (solid lines) for the symmetry axis and opening angle of the L1448 IRS 2 molecular outÑow are indicated. Pluses mark the positions of Herbig-Haro objects HH 195 AÈE (close to L1448 IRS 2), and HH 193 AÈC (at the end of the blueshifted outÑow cavity).

Among the L1448 CO outÑows, the one powered by L1448 IRS 2 is unique in maintaining its initial opening angle out to great distances, as shown in Figure 6 by the lines drawn at P.A. \ 120¡ and 147¡. Such an outÑow morphology is needed to explain the presence of both blue- and redshifted high-velocity gas observed at areas in the map that are not intersected by the highly collimated outÑows from L1448C, L1448N(B), and L1448N(A) (Wolf-Chase, Barsony, & OÏLinger 1998b). The spatially invariant angle of the L1448 IRS 2 outÑow cavity also accounts for the presence of several enhanced CO emission regions and Herbig-Haro objects, in particular the emission knots of HH 193, which lie within the relatively transparent inter-

clump medium outside of the L1448 13CO cloud (Bally et al. 1997). The HH 193 emission knots are unusual in that two of the knots (HH 193 A and C) exhibit relatively low radial blueshifted velocities ([18 and [10 km s~1), while the third knot (HH 193B) has a low radial redshifted velocity (]10 km s~1). This is consistent with the overlapping blue- and redshifted CO emission seen in Figure 6 at that position. If the ““ Ðnger ÏÏ of CO emission toward HH 193 outlines the outÑow cavity wall, this implies the HH emission originates in shocked gas at the periphery of the outÑow rather than along the outÑow axis. The alternate picture of the outÑow being driven by a precessing jet does not appear to be consistent with the V-shaped morphology

704

OÏLINGER ET AL.

of the blueshifted outÑow gas nor with the solid bridge of emission that connects IRS 2 with HH 193. Table 2 lists the derived properties of the L1448 IRS 2 CO outÑow. Because of the large-scale confusion caused by colliding molecular outÑows in L1448 (Wolf-Chase et al. 1998b), we estimate the outÑow mass and energetics in two ways. First, we conservatively use only the closest, unresolved, blue- and redshifted emission peaks directly to the northwest and southeast of IRS 2 to determine a lower limit to the mass of the IRS 2 outÑow. Second, we use the blue- and redshifted integrated intensity over the entire region where highvelocity emission is detected in order to estimate the total mass swept up by the L1448 outÑows. In order to derive the outÑow parameters listed in Table 2, we have performed an LTE analysis using a procedure similar to that described in the Appendix of Wolf, Lada, & Bally (1990). To estimate the optical depth of the gas in the 12CO J \ 1 ] 0 line wings, we obtained a 13CO J \ 1 ] 0 spectrum at the blue peak closest to the source, L1448 IRS 2, in the blueshifted outÑow lobe. Figure 7 shows the 12CO and 13CO J \ 1 ] 0 line proÐles toward this position. Near IRS 2, the 12CO line is self-absorbed, as is evidenced by the dip in the 12CO spectrum in the top panel of Figure 7, which is aligned in velocity with the peak of the optically thinner 13CO line, plotted in the bottom panel. Because of the presence of the 13CO emission feature at V \ 0 km LSR velocs~1, which we associate with a local cloud, the lowest ity material in the blueshifted 12CO line wings is excluded from our mass calculation. Since the outÑow is not detected in the 13CO line wings, we assume that emission in the 12CO line wings is optically thin and derive a strict lower

FIG. 7.ÈIsotopic, 12CO and 13CO J \ 1 ] 0 spectra near the peak of the blueshifted outÑow emission closest to IRS 2 (a \ 03h22m12s. 0, d \ ]30¡35@03A. 0). The vertical dashed line at V 1950 \ 4.2 km s~1 in 1950 panels indicates the velocity of the ambient cloud LSR at the position of both IRS 2.

Vol. 515

limit for the mass and energetic parameters of the L1448 IRS 2 outÑow. We have further assumed T \ 15 K (Bally ex et al. 1997), a beam-Ðlling factor of 0.5 for the gas in the line wings (Margulis, Lada, & Snell 1988 and references therein), and a 12CO/H abundance ratio of 1.1 ] 104 (Langer 1977 ; 2 Wannier 1989 ; Langer & Penzias 1990). We have adopted a characteristic outÑow velocity that is approximately equal to the largest observed outÑow velocity with respect to the ambient cloud velocity. This is a reasonable choice for highly inclined outÑows (Cabrit & Bertout 1992). The good spatial separation between blueshifted and redshifted gas in the plane of the sky combined with the small amount of redshifted emission in the blueshifted lobes, suggests that the IRS 2 outÑow is highly inclined, similar to its neighboring outÑow associated with L1448C, which has an inclination angle of about 70¡ (Bachiller et al. 1995). We make no correction for outÑow gas that is masked by the broad CO line core, although much of the outÑowing gas mass may indeed lie within the core velocities. With these caveats, we derive the total mass of high-velocity gas over a 2.6 pc region to be B0.7 M and _ redthe mass associated just with the unresolved blue- and shifted peaks nearest IRS 2, corresponding to an area within a radial extent of 0.09 pc, to be 0.09 M . However, _ we emphasize that this is probably a gross underestimate, since no attempt was made to include the line-core contribution, which is seen to be substantial (Wolf-Chase et al. 1998b). Additionally, we assumed an optically thin highvelocity CO emission. The actual swept-up mass in our mapped area could easily be underestimated by a factor of 10 or more. Most signiÐcant is that our calculated ““ outÑow force ÏÏ (outÑow momentum divided by the outÑow timescale) for the outÑow associated with IRS 2 is D10~4 M km s~1 yr~1. This large derived outÑow momentum Ñux_is consistent with the high values found in the earliest protostellar stage (Bontemps et al. 1996 ; Wolf-Chase et al. 1998a). We note that this estimate is based on the short dynamical timescale calculated by unrealistically assuming that the IRS 2 Ñow is conÐned to the closest peaks in the blue- and redshifted emission. However, it is interesting to note that in doing the same calculation for the much larger, total outÑow region, the mass increases by roughly the same magnitude as the outÑow timescale, thus the outÑow force remains essentially unchanged. Several isolated, single-contoured features appear along the blueshifted portion of the symmetry axis of the IRS 2 outÑow and at the southeast end of the redshifted emission in our outÑow maps. These features have masses of D0.001 M , which is roughly consistent with masses found for CO _ ““ bullets ÏÏ such as those associated with the L1448 C outÑow (Bachiller et al. 1990). Such emission features are typically associated with the highest velocity gas, which is expected to lie along the central axis, with lower velocity gas along the cavity walls. Although our data indicate highvelocity gas out to great distances from the central source, our spatial resolution (55A beam) and detection limits (T * B 0.1 K) preclude observation of the expected narrow R therefore beam-diluted), low-level (¹100 mK) emis(and sion from the highest velocity gas, which may represent the molecular component of the underlying wind that drives the outÑow (e.g., Koo 1989 ; Masson, Mundy, & Keene 1990 ; Lada & Fich 1996 ; Wolf-Chase & Davidson 1997). The features along the symmetry axis of the L1448 IRS 2

No. 2, 1999

L1448 IRS 2 : HIRES-IDENTIFIED PROTOSTAR

outÑow, taken together with the total extent of the highvelocity CO outÑow cavities as delineated by the opening angles drawn in Figure 6, denote the presence of a bipolar molecular outÑow that is 2.6 pc in total extent. 4.

SUMMARY

1. We have used the MCM algorithm imageconstruction software as implemented in HIRES processing at IPAC to resolve L1448 IRS 2 at 25, 60, and 100 km for the Ðrst time. 2. We have imaged L1448 IRS 2 with SCUBA at the JCMT at 450 and 850km and measured its millimeter continuum Ñux with the NRAO 12 m telescope. 3. From the SED Ðt to our data, we have determined that L1448 IRS 2 is a 5.2 L Class 0 protostar. The corre_ sponding central mass is 0.05 (0.5) M for an assumed 10~5 (10~6) M yr~1 infall rate onto a 3_ R hydrostatic core _ _ (Stahler, Shu, & Taam 1980). The derived envelope mass is 0.86 M . _ present a large-area OTF map (47@ ] 7@, corre4. We sponding to 4 ] 0.6 pc) of the high-velocity CO 1 ] 0 emission of the bipolar molecular outÑow centered on L1448 IRS 2. 5. The L1448 IRS 2 outÑow force is determined to be of order 10~4 M km s~1 yr~1, which is similar to the high _ for other Class 0 protostars. values being found 6. We Ðnd the total spatial extent of the bipolar molecular outÑow powered by L1448 IRS 2 to be D2.6 pc.

705

7. We have shown that HIRES processing of IRAS data can be a useful tool in the search for Class 0 protostars. Much information remains to be extracted by the appropriate techniques from the archival IRAS data base. We are currently using various HIRES-processing methods on data from other star-forming regions (including point-source modeling ; see Hurt & Barsony 1996 ; OÏLinger 1997 ; Barsony et al. 1998) in the hope of identifying more of these rare objects. We thank Darrel Emerson, Eric Greisen, and Je† Mangum of NRAO for the development, implementation, and improvement of the spectral-line OTF mapping capability of the 12 m telescope and the NRAO support sta† for assistance with the data acquisition. We also thank Philip Jewell, currently of the JCMT, for advice on continuum observing at the NRAO 12-meter telescope ; Diane Engler and John Fowler for helpful discussions on the subject of HIRES ; and our anonymous referee for many valuable comments and suggestions. J. O., G. W. C., and M. B. gratefully acknowledge partial Ðnancial support from NSF grant AST-95-01788 for this work. This work was performed while G. W. C. held a PresidentÏs Fellowship from the University of California. M. B.Ïs NSF POWRE Visiting Professorship at Harvey Mudd College, NSF AST97-9753229, provided the necessary time to bring this work to completion.

REFERENCES Andre, Ph., Martin-Pintado, J., Despois, D., & Montmerle, T. 1990, A&A, Mardones, D., Myers, P. C., Tafalla, M., Wilner, D. J., Bachiller, R., & 236, 180 Garay, G. 1997, ApJ, 489, 719  Andre, Ph., & Montmerle, Th. 1994, ApJ, 406, 122 Margulis, M., Lada, C. J., & Snell, R. L. 1988, ApJ, 333, 316  Andre, Ph., Ward-Thompson, D., & Barsony, M. 1993, ApJ, 406, 122 Masson, C. R., Mundy, L. G., & Keene, J. 1990, ApJ, 357, L25 Anglada, G., Rodr• guez, L. F., Torrelles, J. M., Estalella, R., Ho, P. T. P., Moriarty-Schieven, G. H., Wannier, P. G., Mangum, J. G., Tamura, M., & Canto, J., Lopez, R., & Verdes-Montenegro, L. 1989, ApJ, 31, 208 Olmsted, V. K. 1995, ApJ, 455, 190 Aumann, H. H., Fowler, J. W., & Melnyk, M. 1990, AJ, 99, 1674 Motte, F., Andre, P., & Neri, R. 1998, A&A, in press Bachiller, R., & Cernicharo, J. 1986a, A&A, 166, 283 Myers, P. C., Bachiller, R., Caselli, P., Fuller, G. A., Mardones, D., Tafalla, ÈÈÈ. 1986b, A&A, 168, 262 M., & Wilner, D. J. 1995, ApJ, 449, L65 Bachiller, R., Cernicharo, J., Martin-Pintado, J., Tafalla, M., & Lazare†, B. Myers, P. C., Mardones, D., Tafalla, M., Williams, J. P., & Wilner, D. J. 1990, A&A, 231, 174 1996, ApJ, 465, L133 Bachiller, R., Guilloteau, S., Dutrey, A., Planesas, P., & Martin-Pintado, J. OÏLinger, J. 1997, M.Sc. Thesis, Univ. California, Riverside 1995, A&A, 299, 857 Orton, G. S., Griffin, M. J., Ade, P. A. R., Nolt, I. G., Radostitz, J., Robson, Bally, J., Devine, D., Alten, V., & Sutherland, R. S. 1997, ApJ, 478, 603 E. I., & Gear, W. K. 1986, Icarus, 67, 289 Barsony, M. 1994, in ASP Conf. Ser. 65, Clouds, Cores, and Low Mass Persi, P., Palagi, F., & Felli, M. 1994, A&A, 291, 577 Stars, ed. D. P. Clemens & R. Barvainis (San Francisco : ASP), 197 Sandell, G. 1994, MNRAS, 271, 75 Barsony, M., Kenyon, S. J., Lada, E. A., & Teuben, P. 1997, ApJS, 112, 109 Sanders, D., & Willner, S. P. 1985, ApJ, 293, L39 Barsony, M., Ward-Thompson, D., Andre, P., & OÏLinger, J. 1998, ApJ, Stahler, S. W., Shu, F. H., & Taam, R. E. 1980, ApJ, 241, 637 509, 733 Surace, J. A., Mazzarella, J., Soifer, B. T., & Wehrle, A. E. 1993, AJ, 105, Bontemps, S., Andre, P., Terebey, S., & Cabrit, S. 1996, A&A, 311, 858 864 Cabrit, S., & Bertout, C. 1992, A&A, 261, 274 Terebey, S., & Mazzarella, J. (eds.) 1994, Science with High Spatial Cao, Y., Prince, T. A., Terebey, S., & Beichman, C. A. 1996, PASP, 108, 535 Resolution Far-Infrared Data (Pasadena : JPL 94-11) ÈÈÈ. 1997, ApJS, 111, 387 Testi, L., & Sargent, A. I. 1998, in preparation Casali, M. M., & Eiroa, C. 1992, A&A, 262, 468 Walker, C. K., Adams, F. C., & Lada, C. J. 1990, ApJ, 349, 515 Casali, M. M., Eiroa, C., & Duncan, W. D. 1993, A&A, 275, 195 Walker, C. K., Lada, C. J., Young, E. T., Maloney, P. R., & Wilking, B. A. Davis, C. J., Matthews, H. E., Dent, W. R. F., & Richer, J. S. 1998, in 1986, ApJ, 309, L47 preparation Wannier, P. G. 1989, in IAU Symp. 136, Center of the Galaxy, ed. M. Giovanetti, P., Caux, E., Nadeau, D., & Monin, J.-L. 1998, A&A, 330, 990 Morris (Dordrecht : Kluwer), 107 Gregersen, E. M., Evans, N. J., II, Zhou, S., & Choi, M. 1997, ApJ, 484, 256 Ward-Thompson, D. 1996, Ap&SS, 239, 151 Griffin, M. J., Ade, P. A. R., Orton, G. S., Robson, E. I., Gear, W. K., Nolt, Ward-Thompson, D., Buckley, H. D., Greaves, J. S., Holland, W. S., & I. G., & Radostitz, J. V. 1986, Icarus, 65, 244 Andre, Ph. 1996, MNRAS, 281, L53 Griffin, M. J., & Orton, G. S. 1993, Icarus, 105, 537 Ward-Thompson, D., Eiroa, C., & Casali, M. M. 1995, MNRAS, 273, L25 Hodapp, K.-W. 1994, ApJS, 94, 615 Wolf, G. A., Lada, C. J., & Bally, J. 1990, AJ, 100, 1892 Holland, W. S., Cunningham, C. R., Gear, W. K., Jenness, T., Laidlaw, K., Wolf-Chase, G. A., Barsony, M., & OÏLinger, J. 1998b, in preparation Lightfoot, J. F., & Robson, E. I. 1999, Proc. SPIE 3357, in press Wolf-Chase, G. A., Barsony, M., Wootten, H. A., Ward-Thompson, D., Hurt, R. L., & Barsony, M. 1996, ApJ, 460, L45 Lowrance, P. J., Kastner, J. H., & McMullin, J. P. 1998a, ApJ, 501, L193 Hurt, R. L., Barsony, M., & Wootten, A. H. 1996, ApJ, 456, 686 Wolf-Chase, G. A., & Davidson, J. A. 1997, in Proc. ESA Symp. : The Far Koo, B.-C. 1989, ApJ, 337, 318 Infrared and Submillimetre Universe, ed. A. Wilson (Noordwijk : ESA), Lada, C. J., & Fich, M. 1996, ApJ, 459, 638 339 Langer, W. D. 1977, ApJ, 212, L39 Zhou, S. 1995, ApJ, 442, 685 Langer, W. D., & Penzias, A. A. 1990, ApJ, 357, 477 Zhou, S., Evans, N. J., II, Koempe, C., & Walmsley, C. M. 1993, ApJ, 404, Lehtinen, K. 1997, A&A, 317, L5 232 Levine, D. M., & Surace, J. 1993, in IPAC UserÏs Guide, 5th ed. (Pasadena : IPAC), 6-1