Hubble Space Telescope and VLA Observations of Two Optical ...

THE ASTRONOMICAL JOURNAL, 117 : 1143È1150, 1999 March ( 1999. The American Astronomical Society. All rights reserved. Printed in U.S.A.

HUBBL E SPACE T EL ESCOPE AND VLA OBSERVATIONS OF TWO OPTICAL CONTINUUM KNOTS IN THE JET OF 3C 380 CHRISTOPHER P. OÏDEA, WILLEM DE VRIES,1 JOHN A. BIRETTA, AND STEFI A. BAUM Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218 ; odea=stsci.edu, devries=stsci.edu, biretta=stsci.edu, sbaum=stsci.edu Received 1998 July 7 ; accepted 1998 November 30

ABSTRACT We present Hubble Space T elescope Wide Field Planetary Camera 2 broadband red and linear ramp Ðlter (isolating redshifted [O II] j3727) observations and subarcsecond-resolution 15, 22, and 43 GHz VLA observations of the radio-loud quasar 3C 380. We conÐrm the report of de Vries et al. that there is good correspondence between the locations of two optical and radio knots/hot spots in the jet. We show that the optical knots are continuum rather than line emission. The radio-optical spectrum can be Ðtted either by a normal radio spectrum (a D [0.8) with a break somewhere in the infrared or by a steeper single power law (a D [1) (where S P la). We suggest that the two knots are radiating optical synchrotron emission. This would make thel knots in 3C 380 the most luminous optical knots/hot spots currently known. Key words : galaxies : active È galaxies : jets È quasars : individual (3C 380) 1.

INTRODUCTION

De Vries et al. (1997) compared radio and HST Wide Field Planetary Camera 2 (WFPC2) R-band morphologies for a subset of the 3CR that are in the sample of CSS sources constructed by Fanti et al. (1990). De Vries et al. noted the coincidence of two optical and radio knots in the jet in the quasar 3C 380. We present new multifrequency VLA observations and HST WFPC2 R-band and linear ramp Ðlter (LRF) observations of redshifted [O II] j3727 images. We discuss the properties of the optical/radio knots in the jet in 3C 380.

Optical synchrotron emission in radio jets provides an important probe of particle acceleration regions and the physics of energy transport, and it constrains particle acceleration mechanisms and magnetic Ðeld strength (see, e.g., Meisenheimer, RoŽser, & SchloŽtelburg 1996). The Hubble Space T elescope (HST ) has detected several new optical jets and has allowed previously known jets to be studied in unprecedented detail (e.g., PKS 0521[36, Macchetto et al. 1991a ; 3C 66B, Macchetto et al. 1991b, Jackson et al. 1993 ; 3C 78, Sparks et al. 1995 ; 3C 264, Crane et al. 1993, Baum et al. 1997 ; 3C 15, Martel et al. 1998). However, there are still only a small number of detections of optical synchrotron sources, and each new source has the potential to provide important information on this phenomenon. The radio-loud quasar 3C 3802 is listed as a compact steep-spectrum (CSS) radio source in the list compiled by Fanti et al. (1990) because early radio measurements suggested its size was less than the 20 kpc upper limit for CSS sources. On the parsec scale, 3C 380 exhibits a twisted, one-sided radio jet (Wilkinson et al. 1985 ; Pearson & Readhead 1988 ; Cawthorne et al. 1993 ; Polatidis & Wilkinson 1998). The jet shows apparent superluminal motion of components that accelerate away from the core out to a scale of D100 pc (Polatidis & Wilkinson 1998). On the arcsecond scale, the radio source contains a one-sided jet that appears to be a continuation of the jet seen at the parsec scale, which contains two bright knots (Wilkinson et al. 1984, 1991 ; Flatters 1987 ; Pearson, Perley, & Readhead 1985 ; Simon et al. 1990 ; Akujor et al. 1991). The source is embedded in a di†use halo with a total extent of D14A ] 9A (D85 ] 55 kpc2) (Wilkinson et al. 1991 ; van Breugel et al. 1992), implying that 3C 380 is not really a CSS source but is more likely to be a classical double oriented close to our line of sight.

2.

OBSERVATIONS

2.1. V L A Observations We obtained VLA observations at 22 and 43 GHz in order to study the high-frequency radio properties of the knots. We also used archival VLA observations at 15 GHz originally taken by Wilkinson et al. (1991). The parameters of the VLA observations are given in Table 1. Since 3C 380 is dominated by a strong Ñat-spectrum point source (D2 Jy), we were able to use it as its own phase calibrator. The Ñux density scale was determined using observations of 3C 286 matched in elevation. The amplitude stability was checked using observations of the nonvariable gigahertz-peaked spectrum source J0713]438. The data were calibrated and reduced using NRAOÏs AIPS software. The images were deconvolved using the CLEAN algorithm (Clark 1980 ; Cornwell & Braun 1989). Self-calibration was applied Ðrst to the phases and then to both phases and amplitudes, starting with a point-source model and then including clean components as appropriate. However, the Ðnal images are still limited in dynamic range (probably because of the sparse u-v coverage). We combined the visibility data from all the independent observing runs at each wavelength after correcting for variability of the core when necessary. In order to construct the radio-optical spectrum, we constructed VLA images at each frequency, convolved to a resolution of 80 mas (which is slightly superresolved at 15 GHz). We determined the Ñux density of the knots within an aperture of 300 mas in diameter (the same aperture used

ÈÈÈÈÈÈÈÈÈÈÈÈÈÈÈ 1 Also Kapteyn Astronomical Institute, University of Groningen. 2 We adopt a Hubble constant of H \ 75 km s~1 Mpc~1 and a deceleration parameter of q \ 0.1 ; at the 0redshift of 3C 380 (z \ 0.692), the scale is 6.1 kpc arcsec~1.0

1143

1144

OÏDEA ET AL.

Vol. 117

TABLE 1 VLA OBSERVATIONS

Date 1983 1996 1996 1996 1996 1997

ConÐguration

Frequency (GHz)

Bandwidth (MHz)

Time (minutes)

A A A A A B

14.96 22.48 22.48 43.31 43.31 43.31

50.0 50.0 50.0 50.0 50.0 50.0

69 10 10 60 60 50

Oct 29 . . . . . . Dec 14 . . . . . . Dec 16 . . . . . . Dec 14 . . . . . . Dec 16 . . . . . . Feb 15 . . . . . .

NOTE.ÈThe 15 GHz data were originally obtained by Wilkinson et al. 1991 and were taken from the VLA archive.

in the HST optical photometry). The two knots are detected with good signal-to-noise ratio (S/N) in the radio images. The uncertainties in the Ñux densities are due to (1) the di†ering contribution from the underlying di†use lobe emission at each frequency and (2) uncertainties in the absolute Ñux density calibration, which are due to the atmosphere at these relatively high radio frequencies and are estimated to be roughly 5%, 10%, and 20%, respectively, at 15, 22, and 43 GHz. 2.2. HST Observations We obtained HST observations in the Planetary Camera (PC) of WFPC2 (Trauger et al. 1994) through the F702W broad red Ðlter and the linear ramp Ðlter centered at a wavelength of 6303 Ó, which is the redshifted wavelength of the [O II] j3727 emission line (see Table 2). TABLE 2 HST WFPC2 OBSERVATIONS

Date

Filtera

Mean Wavelengthb (Ó)

1994 Sep 14 . . . . . . 1995 Sep 14 . . . . . .

F702W FR680N

6867.8 6302.4

Bandwidthc (Ó)

Timed (s)

1382 82

140 300

a WFPC2 Ðlter used. F702W is a broad red Ðlter similar to Cousins R. b Central wavelength. c FWHM of the Ðlter plus system response. d Exposure time for a single exposure. There were two exposures taken through each Ðlter to allow cosmic rays to be removed. TABLE 3 FLUX DENSITIES Band 2 cm . . . . . . . . 1.3 cm . . . . . . 0.7 cm . . . . . . 6900 Ó . . . . . .

Core 1.96 2.34 2.57 0.63

Jy Jy Jy mJy

Knot 1 204.5 141.3 85.7 9.7

mJy mJy mJy kJy

Knot 2 201.0 137.8 81.1 1.6

mJy mJy mJy kJy

NOTES.ÈThe core is variable at radio wavelengths. The Ñux density in mJy for knot 1 and for knot 2 was determined within a 0A. 3 diameter aperture. Note that this will include some of the di†use ““ lobe ÏÏ emission in the radio 15 and 22 GHz images. The uncertainties in the radio Ñux densities are estimated to be roughly 5%, 10%, and 20%, respectively, at 15, 22, and 43 GHz ; errors in the 6900 Ó Ñux densities are estimated to be roughly 10% and 20%, respectively, for knots 1 and 2.

The HST observations were obtained as part of a snapshot survey of the 3CR sample by Sparks and collaborators (e.g., de Ko† et al. 1996 ; McCarthy et al. 1997). The F702W data were reduced as described by de Vries et al. (1997). The calibration pipeline was rerun using updated reference Ðles. Cosmic rays were removed by combining both images with the CRREJ task in the STSDAS package in IRAF. The F702W image has a point-spread function (PSF) FWHM of D65 mas. In the F702W image, we subtracted a model PSF in order to remove contamination from the bright quasar nucleus. Because of the poor sampling of the WFPC2 PSF, some residuals remain near the quasar nucleus. We performed photometry on the two knots using the task PHOT in IRAF with a 3 pixel radius aperture. We applied corrections for the Ðnite aperture (Table 2a of Holtzman et al. 1995) and for charge transfer efficiency following Whitmore & Heyer (1997). We applied a 15% correction for extinction [using E(B[V ) \ 0.065]. We determined the conversion from counts per second in the image to janskys using the SYNPHOT package in STSDAS. The Ñux densities of the knots are given in Table 3. Knot 1 is detected with good S/N ; however, knot 2 is much fainter and its properties are harder to determine in the HST image. The LRF CCD reduction was also done in the standard way (see de Vries et al. 1999 for details). We used the SYNPHOT package in IRAF to calculate expected count rates for our sources in F702W and the LRF. These count rates were used to normalize both images to the same Ñux scale. We assumed a combination of a Ñat (in F ) continuum j plus [O II] j3727 emission line (cf. spectra in Gelderman & Whittle 1994) for the source input spectrum. After normalizing the images, the images were subtracted to produce continuum-free LRF images and line-free F702W images (see Fig. 6 below).

3.

RESULTS

3.1. Morphology In the 15 GHz radio image, at a resolution of 120 ] 110 mas, a faint hint of a jet extends from the core toward the northwest along a position angle of [50¡ (Fig. 1). About 0A. 73 arcsec from the core a bright knot is seen (here called knot 1). This knot is the brightest feature in the ““ largescale ÏÏ structure. About 0A. 4 beyond knot 1 is knot 2. The

No. 3, 1999

JET OF 3C 380

1145

1.5 1.0

0.8

1.0

ARC SEC

ARC SEC

0.6

0.5

0.4

0.2

0.0

0.0

-0.2 0.2

-0.5

0.5

0.0

-0.5

0.0

-0.2

-0.4 -0.6 ARC SEC

-0.8

-1.0

FIG. 3.ÈVLA image at 43 GHz with a resolution of 80 mas

-1.0

ARC SEC

FIG. 1.ÈVLA image at 15 GHz with a resolution of 120 ] 110 mas at position angle 22¡.0. The scale is 6.1 kpc arcsec~1.

two knots are superposed on di†use, lobelike emission that extends primarily to the northeast and east of the jet axis. This is only the ““ tip of the iceberg,ÏÏ and these structures are embedded inside a di†use structure with a total extent of D14A seen at 1.6 GHz (Wilkinson et al. 1984, 1991 ; van Breugel et al. 1992). The details of the subarcsecond structure are seen more clearly in our 80 mas resolution radio image at 22 GHz (Fig. 2). Knot 1 was found to be roughly 40 mas in size and edge-brightened, with the bright edge nearest the nucleus (Wilkinson et al. 1984 ; Simon et al. 1990), and is unresolved at our 80 mas resolution. However, knot 2 is clearly resolved across the jet (see also Simon et al. 1990). A Gaussian Ðt using JMFIT in AIPS produces a deconvolved FWHM of

174 ] 31 mas at a position angle of 39¡.1. Thus, knot 2 is oriented perpendicular to the jet axis and could be the ““ termination shock ÏÏ at the end of the jet. In this image, there is only faint di†use radio emission beyond knot 2, which is most likely due to the projected lobe (or possibly the remnant of an older jet ?). The 43 GHz image at 80 mas resolution shows only the core and the two knots (Fig. 3). The PSF-subtracted HST optical image shows only the two knots and not any of the ““ lobe ÏÏ emission (Fig. 4). Knot 1 is well detected in the HST image and appears to be unresolved, consistent with the VLA image. Knot 2 is much fainter than knot 1 in the HST image, and its size and shape are not well determined. Deeper HST images are needed to better determine the properties of knot 2 and to search for additional extended emission.

1.2

1.0

ARC SEC

0.8

0.6

0.4

0.2

0.0

-0.2

0.6

0.4

0.2

0.0

-0.2 -0.4 ARC SEC

-0.6

-0.8

-1.0

FIG. 2.ÈVLA image at 22 GHz with a resolution of 80 mas

FIG. 4.ÈHST PC F702W image with model PSF subtracted. Pixel size is 0A. 0455. Residuals are present near the nucleus because of imperfect matching of the PSF and poor sampling by the PC pixels.

1146

OÏDEA ET AL.

The overlay of the optical and radio images is shown in Figure 5. Note the good one-to-one correspondence between the radio and optical shown by the new VLA observations. There may be a slight o†set between the optical and radio centroids of knot 2, or possibly, the optical emission is associated with just a portion of the radio knot. Deeper HST observations are needed to conÐrm this. 3.2. Nature of the Optical Emission We did not detect the two knots in the [O II] emissionline image (Fig. 6). The upper limit on the [O II] line emission in the LRF passband is D10%. Thus, the two knots are dominated by optical continuum light instead of emission lines. Without optical polarization measurements, we cannot deÐnitively distinguish between di†erent mechanisms for the optical continuum light. However, because of the one-to-one radio-optical morphological correspondence, and the continuity in the spectral shape (see below), we prefer synchrotron emission. Since all optical jets for which

Vol. 117

the emission mechanism has been determined are synchrotron, we will adopt it for the purposes of the discussion here. 3.3. Spectrum The radio to optical spectrum of the knots is shown in Figure 7. Between 15 and 43 GHz, knots 1 and 2 have spectral indices of [0.8 and [0.85, respectively. The spectral index between 1.6 and 15 GHz is about [0.6 (P. N. Wilkinson 1998, private communication), indicating that the spectrum is steepening over the radio band. The highfrequency radio spectra extrapolated to the optical overpredict the optical Ñux density by factors of D5 and D20 for knots 1 and 2, respectively. This suggests that (1) the spectrum continues to steepen or breaks between the radio and optical and (2) the steepening is much stronger in knot 2 than in knot 1 ; i.e., if there is a cuto†, it occurs at lower wavelength in knot 2. However, it is possible to Ðt the radio and optical data with a single (steeper) spectral component of [0.97 and [1.14, respectively, for knots 1 and 2. This is as steep as the optical jet in 3C 15 (Martel et al. 1998). If this is the case, then a break or steepening of the spectrum is not

FIG. 5.ÈOverlay of HST PC F702W image (gray scale) and VLA 22 GHz image (contours)

No. 3, 1999

JET OF 3C 380

1147

FIG. 6.ÈHST images of 3C 380. L eft, F702W ; middle, continuum-subtracted [O II] emission-line image in LRF ; right, line emissionÈsubtracted, continuum-only image in F702W.

required between the radio and the optical. Further observations in the near-IR and submillimeter are required to distinguish between these possibilities. 3.4. Properties of the Knots The properties of the knots, assuming incoherent synchrotron radiation for the radio-optical spectrum, are given in Table 4. Standard assumptions are made (see, e.g., Miley 1980). Note that we assume that the Ñux density is not Doppler boosted. However, given that the jet is one-sided and is superluminal on the parsec scale (Polatidis & Wilkinson 1998), at least modest Doppler boosting may be present in the kiloparsec-scale structure. These properties are similar to those estimated for knots or hot spots in other sources (e.g., Meisenheimer et al. 1989 ; Biretta, Stern, & Harris 1991). The electron radiative lifetimes are very short, D103 and D10 yr, respectively, at 22 GHz and 6900 Ó. The electrons radiating in the optical must have been generated within a few parsecs of their current location (unless they are transported out from the nucleus in a loss-free manner, as suggested for M87 by Owen, Hardee, & Cornwell 1989 ;

Biretta 1993). However, Sparks, Biretta, & Macchetto (1996) have pointed out that the optical and UV emission in M87 is concentrated toward the center of the jet (rather than the edges), suggesting that the radiating electrons Ðll the jet rather than being conÐned to a surface layer. As also noted by Heinz & Begelman (1997), this argues against the ““ loss-free pipe ÏÏ model. Thus, in the remainder of this paper we assume that the existence of optical synchrotron emission indicates the presence of particle acceleration. We note that 3C 380 contains the most luminous optical knots/hot spots currently known. Table 5 lists the known optical jets/hot spots in order of increasing P , the 650 nm estimated optical power at 650 nm wavelength. The optical jet in 3C 380 is a factor D5 brighter than the next most luminous jets, those in 3C 212 and 3C 245, and is D2 times brighter than the hot spot in 3C 2 (all recently discovered with HST by Ridgway & Stockton 1997). Since the optical spectral index of 3C 380 is not known, we have conservatively assumed a value a \ [1.2. The high luminosity opt favorable Doppler boosting or could be attributed to either high intrinsic luminosity. In the former case one might

FIG. 7.ÈRadio to optical spectra of knot 1 (left) and knot 2 (right). The solid line is the radio spectrum extrapolated to the optical band. The dotted line is a least-squares Ðt to the radio and optical data.

1148

OÏDEA ET AL.

Vol. 117

TABLE 4 KNOT PROPERTIES Parameter

Knot 1

Knot 2

Spectral index : Radio, a ........................................ [0.81 [0.85 radio Radio-optical, a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . [0.97 [1.14 ro Projected distance from core, D (kpc) . . . . . . . . . . . . . 4.9 7.3 1.4 GHz power, log P (W Hz~1) . . . . . . . . . . . . . . . . 27.1 27.2 1.4 Luminosity (radio-optical), log L (ergs s~1) . . . . . . 45.0 45.0 ro Minimum pressure, P (dyne cm~2) . . . . . . . . . . . . . 3.9 ] 10~7 1.5 ] 10~7 min Magnetic Ðeld, B (G) . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.0 ] 10~3 1.3 ] 10~3 min P Electron energy, c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 ] 105 3.1 ] 105 opt Electron lifetime (yr) : Optical, q . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 20 opt 22 GHz, q ....................................... 1.4 ] 103 3.0 ] 103 rad NOTES.ÈFor the purposes of our calculations, we make the following simplifying assumptions : We assume the radio spectrum (Ðrst row of the table) extends from 107 to 1014 Hz. We assume the Ðlling factor and ratio of electron to proton energy are both unity. We assume that the Ñux density is not Doppler boosted. Note that Doppler boosting would reduce the estimated power, luminosity, minimum pressure, and magnetic Ðeld and increase the electron lifetime.

expect some evidence of the jet pointing more nearly toward the observer, such as a foreshortening of the jet or strong curvature due to projection of small intrinsic bends, but the 3C 380 optical jet is neither remarkably short nor strongly bent compared with other optical jets. It seems more likely that the high luminosity can be attributed to intrinsic properties, such as high kinetic luminosity in the jet, or

stronger interaction with a denser/higher pressure external medium (Wilkinson et al. 1984 ; Fraix-Burnet 1992). 4.

IMPLICATIONS FOR THE RADIO SOURCE

Wilkinson et al. (1984) suggested that the distortions in the jet and lobes in 3C 380 are due to interaction with dense gas. However, we note that (1) Flatters (1987) did not Ðnd

TABLE 5 OPTICAL SYNCHROTRON JETS AND HOT SPOTS

Object

z

Lengtha (arcsec)

Lengtha (kpc)

S (l )b l 0 (kJy)

j c 0 (nm)

3C 111f . . . . . . . . . . . . . . . . . . . 3C 66B . . . . . . . . . . . . . . . . . . . 3C 264 . . . . . . . . . . . . . . . . . . . 3C 33f,g . . . . . . . . . . . . . . . . . . . 3C 78 . . . . . . . . . . . . . . . . . . . . . 3C 15 . . . . . . . . . . . . . . . . . . . . . 3C 274g,i . . . . . . . . . . . . . . . . . 3C 123f . . . . . . . . . . . . . . . . . . . PKS 0518[458f,g,j . . . . . . 3C 20f,g . . . . . . . . . . . . . . . . . . . PKS 0521[36 . . . . . . . . . . 3C 346 . . . . . . . . . . . . . . . . . . . 3C 273g . . . . . . . . . . . . . . . . . . . 3C 245 . . . . . . . . . . . . . . . . . . . 3C 212 . . . . . . . . . . . . . . . . . . . 3C 2f . . . . . . . . . . . . . . . . . . . . . 3C 380 . . . . . . . . . . . . . . . . . . .

0.049 0.021 0.022 0.059 0.029 0.073 0.004 0.218 0.035 0.174 0.055 0.162 0.158 1.029 1.049 1.037 0.692

118.0 8.0 2.2 120.0 1.5 4.2 25.0 8.1 248.0 24.0 6.5 3.6 23.0 1.5 2.2 0.9 1.4

103.1 3.2 0.9 126.6 0.8 5.3 2.1 25.3 159.6 63.3 6.4 9.0 56.2 10.3 15.2 6.2 8.5

1.1 9.0 32.0 1.3 33.0 7.0 4200.0 \1 68.0 0.8 38.0 24.7 35.7 0.7 0.6 1.5 11.3

440 440 702 440 691 702 665 440 440 440 440 645 648 670 670 670 690

d opt [2.9 [2.4 [1.4 [2.8 [1.2h [1.2h [1.2 [1.5 [1.6 [3.7 [1.4 [1.8 [1.3 [1.2h [1.4 [1.4 [1.2h a

log P e 650 nm (W Hz~1)

References

19.23 19.32 19.42 19.49 19.70 19.85 20.16 \20.34 20.50 20.57 20.61 21.21 21.31 21.50 21.52 21.89 22.25

1, 2 3, 4 5, 6 1, 2, 7 8 9 10, 11 1, 2 1, 2 1, 2, 12 13, 14 1 15, 16 17 17 17 18

a Length of optical jet, or distance of hot spot from nucleus. b Flux used to estimate P . 650isnm c Wavelength at which Ñux speciÐed. d Optical spectral index, deÐned in sense that S P laopt . l e Estimated power at 650 nm. f Hot spot. g High optical polarization seen, conÐrming optical synchrotron emission. h Assumed value of a . opt i Virgo A, M87. j Pictor A. REFERENCES.È(1) Dey & van Breugel 1994 ; (2) Meisenheimer et al. 1989 ; (3) Macchetto et al. 1991b ; (4) Butcher, van Breugel, & Miley 1980 ; (5) Baum et al. 1997 ; (6) Crane et al. 1993 ; (7) Meisenheimer & RoŽser 1986 ; (8) Sparks et al. 1995 ; (9) Martel et al. 1998 ; (10) Biretta et al. 1991 ; (11) Perlman et al. 1998 ; (12) Hiltner et al. 1994 ; (13) Macchetto et al. 1991a ; (14) Keel 1988 ; (15) RoŽser & Meisenheimer 1991 ; (16) Thomson, Mackay, & Wright 1993 ; (17) Ridgway & Stockton 1997 ; (18) this paper.

No. 3, 1999

JET OF 3C 380

evidence for high Faraday rotation measures, and (2) we do not Ðnd evidence for extended [O II] j3727 emission in our LRF image. Thus, the observations do not provide support for the interaction hypothesis, though they do not rule it out. Given the edge-brightening of knot 1 seen in VLBI data (Wilkinson et al. 1984 ; Simon et al. 1990) and the presence of optical synchrotron emission, we suggest that knot 1 is the location of a shock in the jet where particles are accelerated. If the knot were simply a brightness enhancement produced by a twisting relativistic jet turning closer to the line of sight, it would be unlikely to be accompanied by optical synchrotron. As discussed above, the electrons radiating optical synchrotron have very short lifetimes and must have been created within a few parsecs of their current location. The presence of a strong shock in the jet at knot 1 with a projected distance of 5 kpc from the core (and likely much larger deprojected) suggests that the jet has continued to propagate supersonically for the entire ([5 kpc) distance. It seems unlikely that the jet could propagate this far if the twisting of the jet on the parsec scale were due to a growing Kelvin-Helmholtz instability (see, e.g., Hardee, Clarke, & Howell 1995 ; Norman 1993). Some alternatives are that either (1) the jet is able to stabilize itself against disruption by the instability, or (2) the twisted parsec-scale jet structure represents quasi-helical structures wrapped around a more stable jet Ñow (e.g., as suggested for M87 by Owen et al. 1989). The very organized magnetic Ðeld structure along the parsec-scale jet (Cawthorne et al. 1993) supports the latter hypothesis. Deeper VLBI images may help to settle this question. Knot 2 is likely to be the terminal shock at the true end of the jet. We note that knot 2 is a factor of D6 weaker in the optical than knot 1 and is also a factor of D4 broader (in the radio). This is consistent with a picture in which the jet Mach number and collimation are reduced in the shock at knot 1. The jet then produces a weaker and broader terminal shock at knot 2. 5.

SUMMARY

We have presented high-resolution VLA images at 15, 22, and 43 GHz and HST WFPC2 R-band and linear ramp

1149

Ðlter [O II] j3727 images of two optical hot spots in the radio jet of 3C 380. 1. We Ðnd good one-to-one correspondence between the optical and radio hot spots in the jet, conÐrming the report of de Vries et al. (1997). 2. The optical Ñux densities of the knots are below an extrapolation of the radio spectrum (with a D [0.8), implying that the spectrum breaks or cuts o† between the radio and optical. However, it is possible to Ðt the radio and optical data with a single (steeper a D [1) spectrum. 3. The optical knots are not detected in the [O II] j3727 image and therefore must be optical continuum. 4. The electron lifetimes are very short in the optical (D10[20 yr). The electrons must have been accelerated within a few parsecs of their current location. Thus, the optical emission probes the Ðne-scale structure of the particle acceleration region. 5. The optical knots/hot spots in 3C 380 are the most luminous currently known. As the extreme example of this phenomenon, they are worthy of further study.

We are grateful to Bill Sparks and Duccio Macchetto for helpful discussions and encouragement. We thank Peter Wilkinson and an anonymous referee for comments on the paper. C. P. O. thanks the Institute of Geophysics and Planetary Physics, Lawrence Livermore National Laboratory, for hospitality during which this project was begun. The VLA is operated by the National Radio Astronomy Observatory, which is a facility of the National Science Foundation, operated under cooperative agreement by Associated Universities, Inc. This work is based on observations made with the NASA/ESA Hubble Space T elescope, obtained from the data archive at the Space Telescope Science Institute. STScI is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. This research made use of the NASA/IPAC Extragalactic Database (NED), which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration, and NASAÏs Astrophysics Data System Abstract Service.

REFERENCES Akujor, C. E., Spencer, R. E., Zhang, F. J., Davis, R. J., Browne, I. W. A., & Hardee, P. E., Clarke, D. A., & Howell, D. A. 1995, ApJ, 441, 644 Fanti, C. 1991, MNRAS, 250, 215 Heinz, S., & Begelman, M. C. 1997, ApJ, 490, 653 Baum, S. A., et al. 1997, ApJ, 483, 178 Hiltner, P. R., Meisenheimer, K., RoŽser, H.-J., Laing, R. A., & Perley, R. A. Biretta, J. A. 1993, in Astrophysical Jets, ed. D. Burgarella, M. Livio, & 1994, A&A, 286, 25 C. P. OÏDea (Cambridge : Cambridge Univ. Press), 263 Holtzman, J. A., et al. 1995, PASP, 107, 156 Biretta, J. A., Stern, C. P., & Harris, D. E. 1991, AJ, 101, 1632 Jackson, N., Sparks, W. B., Miley, G. K., & Macchetto, F. 1993, A&A, 269, Butcher, H. R., van Breugel, W., & Miley, G. K. 1980, ApJ, 235, 749 128 Cawthorne, T. V., Wardle, J. F. C., Roberts, D. H., Gabuzda, D. C., & Keel, W. C. 1988, ApJ, 329, 532 Brown, L. F. 1993, ApJ, 416, 496 Macchetto, F., et al. 1991a, ApJ, 369, L55 Clark, B. G. 1980, A&A, 89, 377 ÈÈÈ. 1991b, ApJ, 373, L55 Cornwell, T., & Braun, R. 1989, in ASP Conf. Ser. 6, Synthesis Imaging in Martel, A. R., et al. 1998, ApJ, 496, 203 Radio Astronomy, ed. R. A. Perley, F. R. Schwab, & A. H. Bridle (San McCarthy, P. J., Miley, G. K., de Ko†, S., Baum, S. A., Sparks, W. B., Francisco : ASP), 167 Golombek, D., Biretta, J., & Macchetto, F. 1997, ApJS, 112, 415 Crane, P., et al. 1993, ApJ, 402, L37 Meisenheimer, K., & RoŽser, H.-J. 1986, Nature, 319, 459 de Ko†, S., Baum, S. A., Sparks, W. B., Biretta, J., Golombek, D., Meisenheimer, K., RoŽser, H.-J., Hiltner, P. R., Yates, M. G., Longair, M. S., Macchetto, F., McCarthy, P., & Miley, G. K. 1996, ApJS, 107, 621 Chini, R., & Perley, R. A. 1989, A&A, 219, 63 de Vries, W. H., OÏDea, C. P., Baum, S. A., & Barthel, P. D. 1999, ApJ, Meisenheimer, K., RoŽser, H.-J., & SchloŽtelburg, M. 1996, A&A, 307, 61 submitted Miley, G. 1980, ARA&A, 18, 165 de Vries, W. H., et al. 1997, ApJS, 110, 191 Norman, M. L. 1993, in Astrophysical Jets, ed. D. Burgarella, M. Livio, & Dey, A., & van Breugel, W. J. M. 1994, AJ, 107, 1977 C. P. OÏDea (Cambridge : Cambridge Univ. Press), 211 Fanti, R., Fanti, C., Schilizzi, R. T., Rendong, N., Parma, P., van Breugel, Owen, F. N., Hardee, P. E., & Cornwell, T. J. 1989, ApJ, 340, 698 W. J. M., & Venturi, T. 1990, A&A, 231, 333 Pearson, T. J., Perley, R. A., & Readhead, A. C. S. 1985, AJ, 90, 738 Flatters, C. 1987, Nature, 326, 683 Pearson, T. J., & Readhead, A. C. S. 1988, ApJ, 328, 114 Fraix-Burnet, D. 1992, A&A, 259, 445 Perlman, E. S., Biretta, J. A., Zhou, F., Sparks, W. B., & Macchetto, F. D. Gelderman, R., & Whittle, M. 1994, ApJS, 91, 491 1999, AJ, in press

1150

OÏDEA ET AL.

Polatidis, A. G., & Wilkinson, P. N. 1998, MNRAS, 294, 327 Ridgway, S. E., & Stockton, A. 1997, AJ, 114, 511 RoŽser, H.-J., & Meisenheimer, K. 1991, A&A, 252, 458 Simon, R. S., Readhead, A. C. S., Mo†et, A. T., Wilkinson, P. N., Booth, R., Allen, B., & Burke, B. F. 1990, ApJ, 354, 140 Sparks, W. B., Biretta, J. A., & Macchetto, F. 1994, ApJS, 90, 909 ÈÈÈ. 1996, ApJ, 473, 254 Sparks, W. B., Fraix-Burnet, D., Macchetto, F., & Owen, F. N. 1992, Nature, 355, 804 Sparks, W. B., Golombek, D., Baum, S. A., Biretta, J., de Ko†, S., Macchetto, F., McCarthy, P., & Miley, G. K. 1995, ApJ, 450, L55 Thomson, R. C., Mackay, C. D., & Wright, A. E. 1993, Nature, 365, 133 Trauger, J. T., et al. 1994, ApJ, 435, L3

van Breugel, W. J. M., Fanti, C., Fanti, R., Stanghellini, C., Schilizzi, R. T., & Spencer, R. E. 1992, A&A, 256, 56 Whitmore, B., & Heyer, I. 1997, New Results on Charge Transfer Efficiency and Constraints on Flat-Field Accuracy (Instrum. Sci. Rep. WFPC2 97-08) (Baltimore : STScI) Wilkinson, P. N., Akujor, C. E., Cornwell, T. J., & Saikia, D. J. 1991, MNRAS, 248, 86 Wilkinson, P. N., Booth, R. S., Cornwell, T. J., & Clark, R. R. 1984, Nature, 308, 619 Wilkinson, P. N., Cornwell, T. J., Kus, A. J., Readhead, A. C. S., & Pearson, T. J. 1985, in Physics of Energy Transport in Extragalactic Radio Sources, ed. A. H. Bridle & J. A. Eilek (Green Bank : NRAO), 76