MULTIWAVELENGTH INSIGHTS INTO MIXED ... - IOPscience

The Astronomical Journal, 125:555–571, 2003 February # 2003. The American Astronomical Society. All rights reserved. Printed in U.S.A.

MULTIWAVELENGTH INSIGHTS INTO MIXED-MORPHOLOGY BINARY GALAXIES. I. ISOCAM, ISOPHOT, AND H IMAGING1 Donovan L. Domingue2 Department of Chemistry and Physics, CBX 082, Georgia College and State University, Milledgeville, GA 31061; [email protected]

Jack W. Sulentic3 Department of Physics and Astronomy, University of Alabama, Box 870324, Tuscaloosa, AL 35487; [email protected]

Cong Xu, Joseph Mazzarella, and Yu Gao4 Infrared Processing and Analysis Center, Jet Propulsion Laboratory, Mail Stop 100-22, California Institute of Technology, Pasadena, CA 91125; [email protected]; [email protected], [email protected]

and Roberto Rampazzo Osservatorio Astronomica di Padova, INAF, vicolo dell’Osservatorio 5, I-35122 Padova, Italy; [email protected] Received 2002 July 3; accepted 2002 October 18

ABSTRACT We present H and ISO mid- and far-IR observations for a sample of mixed-morphology galaxy pairs that reveal both the stellar and nonstellar signatures of the interaction process. A mixed-morphology pair is perhaps the simplest form of galaxy-galaxy interaction because it is expected to involve only a single rapidly rotating gas-rich component paired with a gas-poor elliptical or lenticular galaxy. A primary assumption that we address is whether spirals are the only IR emitter in these mixed (E+S) pairs. Our observations reveal that many of the early-type galaxies exhibit weak (low equivalent width) emission, as often observed in field elliptical galaxies. These are the classical mixed-morphology pairs. However, some of the early-type components, especially the lenticular galaxies, show evidence for significant star formation, with H equivalent widths and 15 lm luminosities comparable to or exceeding those of their often much larger spiral companions. Our sample contains five Seyfert 2 nuclei, of which three can be described as companions on the end of a spiral arm. The Seyfert nucleus is often accompanied by a starburst region, while other such companions currently show only the starburst component. These pairs are among the best candidates for direct interaction fuelling of both starbursts and active galactic nuclei. Key words: galaxies: elliptical and lenticular, cD — galaxies: ISM — galaxies: photometry — galaxies: spiral

Investigating the multiwavelength signatures of interaction-induced secular evolution and star formation in spiralspiral pairs is complicated because both components of the system are gas-rich and rapid rotators. Models suggest that the relative orientation of angular momentum vectors adds an additional variable to the interaction equation in such pairs. The spatial resolution of IRAS at far-infrared wavelengths typically allowed only a global measure of emission properties for most galaxy pairs. Mixed pairs involving a spiral or irregular galaxy paired with an elliptical or lenticular galaxy offer a twofold simplification because they contain a single gas-rich component and a single component with high specific angular momentum. The former allows us to assign all the nonstellar material and associated star formation activity to a single component, at least as an initial assumption. The latter minimizes the role of relative orientation of component angular momentum vectors from the set of variables thought to influence interaction-induced effects. Such pairs are also an ideal sample to search for evidence of component cross fueling, because any large quantity of gas observed in the E/S0 components is prima facie evidence for gas transferred from the late-type spiral companion. The Catalogue of Isolated Pairs of Galaxies (CPG, Karachentsev 1972) is an ideal sample for binary galaxy studies because of its size, completeness, and relatively unbiased selection (see, e.g., Sulentic 1989). Morphological

1. INTRODUCTION

Binary galaxies constitute about 10% of the noncluster population (Xu & Sulentic 1991), and disturbed morphologies are a common property of a large fraction of such pairs (e.g., Arp 1966; Karachentsev 1972). Early numerical simulations (Alladin 1965; Toomre & Toomre 1972) showed that these peculiarities were a manifestation of galaxy-galaxy interactions. Both interactions and mergers are believed to play an important role in galaxy evolution over cosmological timescales. Galaxies in bound orbits, as well as those involved in chance encounters, may evolve differently from isolated galaxies. Both gravitational and hydrodynamic effects can redistribute gas within the galaxies and enhance the rate of star formation (Keel et al. 1985; Kennicutt et al. 1987; Xu & Sulentic 1991).

1 Based on observations with the Infrared Space Observatory (ISO), an ESA project with instruments funded by ESA Member States (especially the PI countries: France, Germany, the Netherlands and the United Kingdom) with the participation of ISAS and NASA. 2 This research was carried out as a part of D. L. D.’s Ph.D. thesis in the Department of Physics and Astronomy at the University of Alabama. 3 Visiting Astronomer, German-Spanish Astronomical Centre, Calar Alto, operated by the Max Planck Institute for Astronomy, Heidelberg, jointly with the Spanish National Commission for Astronomy. 4 Also Department of Astronomy, LGRT-B 619E, University of Massachusetts, 710 North Pleasant Street, Amherst, MA 01003-9305.

555

556

DOMINGUE ET AL.

classification of pairs in the CPG suggests that about 25% of CPG pairs consist of a spiral/irregular paired with an elliptical/lenticular galaxy. This is consistent with the expected fraction of such pairs, assuming that they form from random combinations of field galaxies (Sulentic 1991). However, the observed number of pairs is far in excess of that expected under the assumption that pairs form by random capture (Chaterjee 1987). Cross-correlation of the CPG with the IRAS Point Source Catalog (PSC) revealed that 54% of the mixed pairs are FIR sources (Sulentic 1989). An FIR luminosity excess is observed from mixed pairs, albeit weaker than that found for spiral-spiral pairs (Xu & Sulentic 1991; Hernańdez Toledo et al. 2001). So far MIR/ FIR studies of pairs have been resolution limited (but see Surace et al. 1993; Bushouse, Telesco, & Werner 1998), which means that we know little about (1) the relative emission contribution of pair components and (2) the distribution of emission within the component galaxies. There is evidence that some early-type components of mixed pairs show blue nuclear regions (Rampazzo et al. 1995) and young stellar populations have also been found through population synthesis (de Mello et al. 1996). This effect, if real, may arise from recent bursts of star formation in ambient gas, component cross-fueling, or past merger events. Holmberg (1958) found that galaxies in pairs tend to have similar colors, and this ‘‘ Holmberg effect ’’ was also observed in mixed pairs (Demin et al. 1984; Reduzzi & Rampazzo 1995). If real, it suggests that environment plays a strong role in the secular evolution of the pairs or that the pairs induce parallel evolution, at least as manifested by the star formation history (Kennicutt et al. 1987). If pair components form in a common environment (as opposed to formation by random capture) and if initial environmental conditions determine galaxy morphology, then why do we observe 25% mixed pairs in the CPG? If 25% mixed pairs are expected from random pair formation, then either (1) pair formation by random capture is much more common than we assume, (2) most pairs formed in the early universe but after galaxy morphologies were determined by their environments, or (3) most E+S pairs are a product of their secular evolution as pairs. Neither of the first two alternatives are very appealing, so we are motivated to look for evidence supporting the latter hypothesis. A ‘‘ nurture ’’ scenario places more importance on interaction-induced effects such as accretion, merging, and tidal stripping. This scenario allows for morphology change and greater type discordance in pairs. Rampazzo & Sulentic (1992) proposed four scenarios to account for the large number of mixed-morphology pairs. They included two scenarios that involve classification refinement and two that describe possible physical evolution processes. The scenarios involve some apparent E+S pairs as (1) ‘‘ early-type ’’ systems involving an E/S0 paired with an S0 misclassified as a spiral, (2) ‘‘ disky ’’ systems with a spiral and an S0 misclassified as an elliptical, (3) systems with early-type components showing induced spiral arms, and (4) systems that might be the late stage in the coalescence of higher order multiplets (triplets or compact groups), with the elliptical component as the merged portion of an originally spiral-rich system. An examination of large mixed-pair samples, keeping these possibilities in mind, can lead to a better defined subset of true E+S pairs, as well as more useful subsamples for studying galaxy interaction and evolution.

Vol. 125 1.1. Infrared Observations of Mixed Pairs

Most galaxy pairs were not resolved in the IRAS survey, which covered 96% of the sky in four wavelength bands centered at 12, 25, 60, and 100 lm. Most previous studies of binary galaxies using the IRAS database assumed that the mixed-morphology pairs contained a single IR bright component, but tests of this assumption require higher resolution data. Study of the IR spectral energy distribution of pairs would also benefit from broader wavelength coverage (IRAS detected many fewer pairs at 12/25 lm than at 60/ 100 lm). The Infrared Space Observatory (ISO, Kessler et al. 1986) offered the possibility to extend the IRAS results along these lines, using ISOCAM and ISOPHT for MIR and FIR imaging, respectively. While it is not a survey telescope, pointed observations with ISO allowed us to map the MIR and FIR emission in nine and 10 pairs with these respective instruments. In the next section we discuss our selection criteria, which resulted in a sample of likely interacting and IR bright mixed pairs. In x 3 we also describe the reduction procedures for our ground-based and ISO observations. Section 4 presents our results about the R-band and H emission distributions as well as a reevaluation of the morphological properties of the sample. The ISOCAM and ISOPHT images and comparison with H are presented in xx 5 and 6. Finally, we examine the results in the context of published data for individual pairs. A quantitative comparison of IR emission between this sample and isolated control samples will be presented in a future paper.

2. SAMPLE SELECTION

The CPG is likely the most complete local sample of visually selected isolated galaxy pairs. It includes 603 binary systems and is estimated to be complete to mZw = 15.0 and out to 104 km s1 (Xu & Sulentic 1991). The CPG is based on a visual search of the Palomar Sky Survey north of 3 , with a magnitude limit of mZw 15.7. The CPG employed an isolation criterion that requires xji ai ; j ¼ 1; 2 ð1Þ x12 aj and aj ai aj ; j ¼ 1; 2 ;

ð2Þ

where xij is the separation between the jth component and the nearest neighbor galaxy, x12 is the separation of the pair components, and a is the major-axis diameter of the respective galaxies. The isolation criterion identifies the least isolated pairs with coefficients: ¼ 5 ; ¼ 0:5 ; ¼ 4 :

ð3Þ

This means that identically sized galaxies in a pair cannot have neighbors closer than 5 times the distance between the components of the pair. The isolation criterion increases the probability that the observed pairs are physical binaries, rather than chance projections. Radial velocity measures exist for all CPG pairs, which allow the identification of discordant pairs (10% of the CPG). The CPG provides a well-defined sample without obvious morphological or luminosity biases.

No. 2, 2003

MIXED-MORPHOLOGY BINARY GALAXIES. I.

557

TABLE 1 Observations Galaxy Pair

ISOCAM

ISOPHOT

CPG

NGC

R

H

LW8

LW3

60 (lm)

100 (lm)

160 (lm)

79.................. 83.................. 202 ................ 234 ................ 394 ................ 414 ................ 439 ................ 460 ................ 468 ................ 536 ................ 542 ................ 547 ................ 548 ................ 552 ................ 572 ................ 579 ................ 591 ................

IC 258/9 1141/2 2872/4 3226/7 5296/7 ... ... 5909/12 5953/4 6636 ... ... 6962/4 ... 7374 ... 7752/3

x x x x x x x x x x x x x x x x x

x x x x x x x x x x x x x x x x x

x ... ... ... ... ... x x x ... x ... ... x x x x

x ... ... ... ... ... x x x ... x ... ... x x x x

... ... ... ... ... ... x x x ... x x x x x x x

... ... ... ... ... ... x x x ... x x x x x x x

... ... ... ... x ... x x ... ... x ... x ... x x ...

A high-priority subsample of mixed pairs was chosen from the CPG with (1) IRAS f (60 lm) 1 Jy, (2) component separation 0 11, is included in our sample. Both of

558

DOMINGUE ET AL.

Vol. 125

TABLE 2 Basic Data for the Pairs

Pair CPG (1)

(2)

(3)

mHo (mag) (4)

Vrad (km s1) (5)

a (arcmin) (6)

Type Rev. (7)

Class Tifft (8)

Sep. (arcmin) (9)

Class Inter (10)

79w ............... 79e ................ 83w ............... 83e ................ 202w ............. 202e .............. 234w ............. 234e .............. 394w ............. 394e .............. 414w ............. 414e .............. 439w ............. 439e .............. 460w ............. 460e .............. 468w ............. 468e .............. 536w ............. 536e .............. 542w ............. 542e .............. 547w ............. 547e .............. 548w ............. 548e .............. 552w ............. 552e .............. 572w ............. 572e .............. 579w ............. 579e .............. 591w ............. 591e ..............

02 49 40.9 02 49 46.1 02 55 09.7 02 55 12.2 09 25 42.5 09 25 47.3 10 23 27.0 10 23 30.6 13 46 18.6 13 46 23.5 14 04 49.9 14 04 54.6 14 51 27.9 14 51 29.4 15 11 28.0 15 11 41.3 15 34 32.3 15 34 35.0 18 22 03.5 18 22 05.7 19 31 08.2 19 31 10.4 20 34 42.3 20 34 46.6 20 47 19.0 20 47 23.9 21 07 41.4 21 07 45.8 22 45 59.2 22 46 01.0 23 17 20.9 23 17 21.7 23 46 58.5 23 47 04.7

41 03 18 41 03 06 00 10 40 00 11 01 11 25 55 11 25 27 19 53 54 19 51 54 43 51 04 43 52 20 12 43 40 12 42 17 09 19 19 09 20 05 75 23 02 75 23 05 15 11 38 15 12 02 66 36 59 66 37 19 54 06 07 54 05 33 07 59 22 07 59 19 00 19 17 00 18 02 03 52 23 03 52 40 10 52 05 10 51 15 05 39 39 05 38 41 29 27 32 29 29 01

15.32 15.49 13.45 13.98 12.57 13.05 13.06 12.06 14.87 12.21 14.78 14.76 15.93 14.78 14.63 14.41 13.18 13.65 14.58 15.39 14.40 15.81 14.64 15.86 12.99 13.81 14.44 15.35 15.27 14.55 14.69 15.00 14.24 13.00

5547 6339 8514 8714 2885 3575 1168 1063 2314 2755 4118 4158 8743 8889 7268 7456 2072 2101 4259 4478 4111 3980 7856 7887 4419 4032 8068 8196 7546 7301 10160 10008 5125 5328

0.6 0.9 0.8 0.7 1.6 2.4 2.5 3.7 0.9 5.4 1.1 0.5 0.3 0.6 1.1 1.0 1.7 1.5 1.6 0.5 1.7 0.8 0.6 0.9 2.8 1.5 1.0 0.6 0.5 0.8 0.6 0.7 0.9 2.5

S0 S0 E S E Sb E Sb S0 Sb Sm S0 Sa S0 Sb E S0/a Sc Sc Comp. Sb E E Sc Sa E Sa Comp. E Sb Sc S0 I0 Sc

ABS WEK ABS MID ABS ABS ABS STG MID ABS MID STG STG STG ABS ABS MID STG MID STG MID ABS ABS WEK ABS ABS STG STG ABS MID MID WEK STG MID

1.02 1.02 0.71 0.71 1.29 1.29 2.35 2.35 1.57 1.57 1.10 1.10 0.86 0.86 0.79 0.79 0.77 0.77 0.42 0.42 0.68 0.68 0.67 0.67 1.76 1.76 1.23 1.23 0.92 0.92 1.00 1.00 2.00 2.00

ATM ATM ATM ATM DIS DIS LIN LIN DIS DIS DIS DIS LIN LIN DIS DIS LIN LIN ATM ATM ATM ATM LIN LIN DIS DIS LIN LIN DIS DIS DIS DIS LIN LIN

Nucleus (11) ... ... ... Sy2 ... ... Liner Sy1.5 ... ... ... ... ... ... ... ... Sy2 ... ... Sy ... ... ... ... ... ... ... Sy ... ... ... ... ... ...

Notes.—Col. (1): CPG pair member. Cols. (2)–(3): right ascension and declination (units of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcminutes, and arcseconds; J2000.0). Col. (4): optical magnitude in the standard Holmberg system from Karachentsev 1987 as originally outlined in Paturel 1976, 1979. Col. (5): radial velocity in km s1. Col. (6): component major-axis diameter in arcminutes. Col. (7): revised Hubble type based on unpublished CCD images. Col. (8): Tifft 1982 spectroscopic class based on emission-line strength (ABS, WEK, MID, STG) as determined from photographic spectra. Col. (9): angular separation of pair components in arcminutes. Col. (10): CPG interaction class defined in text (ATM, DIS, or LIN). Col. (11): nuclear activity as published.

those pairs contain a Seyfert galaxy. No luminous infrared galaxies [LIRGS; log (LFIR/L) e 11.3] or ultraluminous infrared galaxies [ULIRGs; log (LFIR/L) e 12] are found in the CPG mixed-pair sample. LIRGs and ULIRGs are often interpreted as merging systems. Such objects may be selected against in the CPG for a number of reasons. The onset of extremely luminous and warm far-infrared dust emission is often observed in objects that are apparently late in the merger phase, when the pair no longer shows discrete components. Another possibility is that the production of a LIRG or ULIRG may require the merger of two gas-rich disk galaxies; such systems are clearly selected against in a mixed-morphology sample. 3. OBSERVATIONS AND DATA REDUCTION

the Infrared Space Observatory (ISO, Kessler et al. 1996). Both the LW3 (12–18 lm) and LW8 (10.7–12 lm) filters were used to sample different regions of the mid-IR spectrum. Each pair was observed in a 4 4 step raster mode with 600 pixels. Steps of 600 were used between pointings, which involved multiple integration times of 2 s. The final images have 35 35 pixels, resulting in a 362. In both cases a broadband ˚ ) Johnson R filter centered near 6580 A ˚ (FWHM 1380 A (Johnson 1965) was obtained. Narrowband observations ˚ ) tuned were made with interference filters (FWHM 100 A to H+[N ii] at the appropriate redshifts. Table 5 lists the transmission properties of the narrowband filters. The Rband images permit subtraction of continuum light from the narrowband images. The R-band frames were normalized to the narrowband frames by comparative photometry of stellar sources in each frame. This permits removal of the continuum, because stellar sources do not exhibit H emission lines. The scaling and subsequent subtraction introduce the largest uncertainty in our photometric measurements. The standard stars HD 84937 and Landolt standard 104334 were used for Palomar photometric calibrations and BD +28 4211 was observed for Calar Alto photometry. Two of the four nights at Calar Alto and one of the three nights at Palomar were photometric. Table 6 lists derived absolute photometry for only the photometric nights. 4. OPTICAL RESULTS

The Balmer lines of hydrogen, in emission, arise primarily from H ii regions in the disk or nuclear regions of component galaxies. Sometimes more diffuse emission due to shocks is observed. In the latter case the contribution of [N ii] 6548, 6583 emission to the observations can even exceed that of H. The greatest chance of this occurring involves galactic nuclei with various levels of activity. The ratio of H/[N ii] 3.0 is observed in extranuclear H ii regions, with the inverse often seen in nuclear regions. Regions of active star formation are associated with H ii regions, which explains the strong correlations observed between H and MIR/FIR emission in galaxies (Kennicutt et al. 1987). H is a tracer of star formation because hydrogen is ionized by stellar radiation primarily from young OB TABLE 5 Transmission Properties of the Narrowband Filters center ˚) (A

Weff ˚) (A

Tpeak

6600 ............... 6607 ............... 6667 ............... 6700 ............... 6737 ............... 6737 ............... 6800 ...............

110.6 50.45 76.39 113.9 70.51 187.96 126.8

0.65 0.69 0.77 0.58 0.73 0.83 0.46

Vol. 125 TABLE 6 H Data

Galaxy Pair CPG (1)

log fH (ergs s1 cm2) (2)

EW ˚) (A (3)

LH (107 L) (4)

SFR (M yr1) (5)

fn/ft (6)

79w ................... 79e .................... 83w ................... 83e .................... 202w ................. 202e .................. 234w ................. 234e .................. 394w ................. 394e .................. 414w ................. 414e .................. 439w ................. 439e .................. 460w ................. 460e .................. 468w ................. 468e .................. 536w ................. 536e .................. 542w ................. 542e .................. 547w ................. 547e .................. 548w ................. 548e .................. 552w ................. 552e .................. 572w ................. 572e .................. 579w ................. 579e .................. 591w ................. 591e ..................

... ... 13.1 11.9 11.8 12.1 ... 11.9a ... ... ... ... ... ... 12.9 13.3 11.8a 11.9a 12.4 12.6 12.1 13.5 15.4 13.2 12.2 13.2 12.6 12.6 13.4 12.6 ... ... 12.4a 12.6a

1 5 8 48 12 15 45 58 36 57 52 69 31 20 10 3 51 58 34 70 34 3 1 14 7 2 14 60 9 36 60 44 82 23

... ... 3.1 51.4 7.1 5.5 ... 0.8 ... ... ... ... ... ... 3.6 1.5 3.7 3.0 2.4 4.3 7.2 0.3 0.01 2.1 6.6 0.6 8.8 9.1 1.2 7.2 ... ... 5.6 3.8

... ... 0.87 14.39 1.99 1.54 ... 0.24 ... ... ... ... ... ... 1.01 0.42 1.04 0.84 0.67 1.20 2.02 0.08 0.002 0.59 1.85 0.17 2.46 2.55 0.34 2.02 ... ... 1.57 1.06

... 1.00 1.00 0.08 1.00 0.11 1.00 0.22 0.18 0.03 0.03 1.00 1.00 0.93 0.10 1.00 0.20 0.12 0.05 0.47 0.13 1.00 1.00 0.10 0.12 1.00 0.11 1.00 1.00 0.08 0.07 0.62 0.41 0.04

Notes.—Col. (1): pair ID. Col. (2): logarithm of the H+[N ii] flux for sources where photometric observations were available. Col. (3): equivalent width of H+[N ii]. Col. (4): H luminosity (H = 75 km s1 Mpc1). Col. (5): star formation rate as determined from H+[N ii] flux without correction for extinction or [N ii]/H ratio. Col. (6): fractional contribution of the nuclear region to the total H+[N ii] emission from each galaxy. a Values as published in Kennicutt et al. 1987.

stars. R-band continuum and H+[N ii] interference filter images are presented in Figure 2 for sources with matching ISOCAM data. Figure 3 presents the Corresponding data for sources without ISOCAM observations. Table 6 lists the observed and derived H properties of the sample. 4.1. H Morphologies The H emission–line images reveal the presence and location of star-forming regions within pair components. The morphology is relevant to classifying normal disk versus interaction-induced star formation in tidal features and excess nuclear emission. Since MIR/FIR emissions tend to follow H, these images are of great help in interpreting IR data with lower resolution and signal-tonoise ratio (S/N). One can identify several characteristic distributions of H emission in spiral components (see, e.g., Combes et al. 1994): (1) knotty emission distributed

No. 2, 2003


561

Fig. 2.—R-band (left), continuum subtracted H+[N ii] images (second from left), H+[N ii] images convolved to match the PSF of ISOCAM (second from right), and ISOCAM LW3 (15 lm) images for the pairs of our sample that have both H and ISOCAM imaging available. The orientation of the images is such that north is up and east is left.

throughout the disk with/without strong nuclear emission, (2) emission concentrated in the nuclear region, plus a partial or complete ring near the edge of the disk (cartwheel-like), and (3) strong and irregular central emis-

sion sometimes accompanied by radial filaments. Another possibility involves (4) diffuse emission throughout the disk, which might indicate dominance of forbidden [N ii]. We are relatively insensitive to the latter possibility

562

DOMINGUE ET AL.

Vol. 125

Fig. 2.—Continued

because we use a broadband continuum template. Various permutations of possibility 3 have also been proposed to describe the emission morphology in early-type galaxies (Macchetto et al. 1996). A final possibility involves (5) emission associated with tidal features.

Table 7 lists pair components along with simplified H emission morphology classes. They are divided into galaxies with mostly nuclear emission, emission that extends beyond the central concentration but shows no clear structure, galaxies with both nucleus and disk, and finally disk galaxies

No. 2, 2003


563

Fig. 3.—R-band, continuum-subtracted H+[N ii] images for pairs of our sample that have no matching ISOCAM images. The orientation of the images is such that north is up and east is left.

that have nuclear emission that contributes less than 5% of the total emission. All of the early-type components in our sample show signs of nuclear emission. Those in CPG 394, 468, 536, 579, and 591 show extended emission outside of the nucleus (but not with an obvious disk morphology). Both disk and nuclear emission are prevalent in the spiral components, with 11 of 17 pairs showing both. Four of the spirals display only knotty disk emission, while the S0 in CPG 79 and the spiral in CPG 439 show only nuclear emission. The contribution of the nuclear emission in the spirals is generally less than 0.15 of their total integrated H flux (see Table 6 for ratios fn/ft). The five Seyfert nuclei in the H sample contribute more than 20% of their integrated flux except for CPG 83e, which displays a starburst in its

disk. Based on the sample nuclear contribution properties, CPG 439e fits the description of a Seyfert galaxy, though no known spectroscopic classification exists for this relatively high central surface brightness galaxy. The median nuclear flux for the sample spirals is 10% of their total flux. This is similar to the fraction reported for an independent sample of interacting galaxies (Kennicutt et al. 1987) and approximately 2.5 times the fraction reported for isolated spirals (Kennicutt & Kent 1983). This may indicate that the spirals in our mixed pairs have tidally induced nuclear star formation activity. The source of uncertainty involves the relative contribution of [N ii] emission to the nuclear measures, as well as the uncertainty in scaling the broadband R images for continuum subtraction.

564

DOMINGUE ET AL.

Vol. 125

cal results for SFRs and H equivalent widths are presented in Table 6 with the derived fluxes.

TABLE 7 Emission Morphologies Galaxy CPG

H

Mid-IR

Far-IR

79w ........... 79e ............ 83w ........... 83e ............ 202w ......... 202e .......... 234w ......... 234e .......... 394w ......... 394e .......... 414w ......... 414e .......... 439w ......... 439e .......... 460w ......... 460e .......... 468w ......... 468e .......... 536w ......... 536e .......... 542w ......... 542e .......... 547w ......... 547e .......... 548w ......... 548e .......... 552w ......... 552e .......... 572w ......... 572e .......... 579w ......... 579e .......... 591w ......... 591e ..........

Nuclear Nuclear Nuclear Disk+nuclear Nuclear Disk+nuclear Nuclear Disk+nuclear Extended Disk Disk Nuclear Nuclear Nuclear Disk+nuclear Nuclear Extended Disk+nuclear Disk Extended Disk+nuclear Nuclear Nuclear Disk+nuclear Disk+nuclear Nuclear Disk+nuclear Nuclear Nuclear Disk+nuclear Disk+nuclear Extended Extended Disk

No detection Point source ... ... ... ... ... ... ... ... ... ... Point source Point source Extended Point source Extended Extended ... ... Extended No detection ... ... ... ... Extended Extended No detection Extended Extended Extended Extended Disk+nuclear

... ... ... ... ... ... ... ... ... ... ... ... Detected Detected Detected No detection Detected Detected ... ... Detected No detection No detection Detected Detected No detection Detected Detected No detection Detected Detected Detected Detected Detected

5. ISOCAM IMAGING

4.2. Optical Measurement of Star Formation H emission from paired and isolated galaxies has been well studied (Kennicutt & Kent 1983; Kennicutt 1983; Bushouse 1987; Kennicutt et al. 1987; Pogge & Eskridge 1993). The star formation rate (SFR) can be estimated from the H luminosities of galaxies, using models of the photoionization properties (Kennicutt 1983) of various possible initial mass functions (IMFs). An alternate measure involves using the H equivalent width (EW) as a continuum-normalized measure of the current SFR. The narrowband filters used for this work usually include the nearby forbidden [N ii] 6548, 6583 lines along with H. Fluxes quoted here represent those of H+[N ii], as corrections for [N ii] were not applied. Ratios of H to [N ii] are calculated and applied in Kennicutt & Kent (1983) and were found to be less than 25%. The normal value for the H/[N ii] ratio is 3.0 in H ii regions. It frequently shows a reversal in galactic nuclei. Shocked emission in galactic disks might be undetectable using broad R band as a continuum model, although this assumption is less certain when strong interactions are common. Spectroscopic coverage of the pairs in this sample ranges from detailed to nonexistent. The five known Seyfert nuclei in our sample typically show [N ii]6583/H > 1.0. CPG 394e (Rampazzo et al. 1995) and 591e (Keel et al. 1985) also show this reversal. Numeri-

ISO provided improved resolution and sensitivity in the 3–25 lm (MIR) range. We used the LW8 filter to measure 11.3 lm emission, which is assumed to be a polycyclic aromatic hydrocarbons (PAHs) contribution. Longer wavelength continuum emission from warm grains was sampled with the LW3 (15 lm) filter. LW8 and LW3 MIR images for nine mixed pairs are presented in Figure 2. H images for the pairs were convolved with the point-spread functions (PSFs) of the LW3 and LW8 ISOCAM filters in order to better compare with the MIR images and test the hypothesis that mid-IR emission primarily originates from star-forming regions and therefore follows H closely. The H images were degraded to the pixel size of 600 pixel1 and convolved with the observed ISOCAM PSF in order to facilitate comparison with ISOCAM maps. The PSF was taken from CIA and applied using the IRAF task STSDAS. ANALYSIS.FOURIER.FCONVOLVE.5 The convolved images are also presented in Figure 2 for useful comparison. The nine ISOCAM images can be classified by the relative strength of MIR emission in a similar manner to the H data. Measured flux (Table 8) serves as a first-order emission strength indicator because of the sample’s well-defined redshift distribution. CPG 79 and 460 are composed each of two weak mid-IR emitters ( f < 30 mJy). Only two pairs, CPG 542 and 572, show one strongly emitting galaxy at MIR wavelengths plus an undetected companion, corresponding most closely to our initial expectation for mixed pairs. Pairs with two strong emitters include CPG 439, 468, 552, 579, and 591. Seyfert nuclei contribute significantly to 5 IRAF is distributed by the National Optical Astronomy Observatory, which is operated by the Association of Universities for Research in Astronomy Inc., under cooperative agreement with the National Science Foundation.

TABLE 8 Mid-Infrared Data Galaxy CPG

FLW8 (mJy)

FLW3 (mJy)

79w ................... 79e .................... 439w ................. 439e .................. 460w ................. 460e .................. 468w ................. 468e .................. 542w ................. 542e .................. 552w ................. 552e .................. 572w ................. 572e .................. 579w ................. 579e .................. 591w ................. 591e ..................

0.6 32 55 108 16 4 492 222 211 2 85 66 0.7 60 102 47 127 137

0.4 24 37 106 12 4 347 164 171 1 100 70 0.5 52 64 44 100 83

No. 2, 2003


the emission from CPG 468 and 552. All of the MIR descriptions match those based upon the H images. The EW as a measure of H emission strength is therefore verified as a good predictor for the mid-infrared brightness of mixed-pair components. Determining whether our mid-infrared detections are point sources or extended emission requires comparisons with the known point-spread function (PSF) of the ISOCAM filters. The LW3 and LW8 filter PSFs have measured ratios of the mean peak emission in the central pixel to the total integrated flux of a point source. The peak emission to total flux ratios are 0.37 and 0.42, respectively, for LW3 and LW8. At 11.3 and 15 lm, all detected galaxies, except CPG 79e, CPG 439e&w, CPG 460e, likely contain extended mid-infrared emission. Here CPG 439e exhibits point-source properties that are consistent with its H nuclear concentration and our proposal that this is a Seyfert galaxy. These findings are summarized in Table 7. The photodissociation region environments of galaxies have been studied with MIR ISO spectra and imaging of emission features originally known as the unidentified infrared bands. These bands have been associated with PAHs, which have the predicted emission peaks near observed features at 3.3, 6.2, 7.7, 8.6, and 11.3 lm. Although the identification of the bands is controversial, they may be caused by the bending and vibrational modes of PAH molecules (Puget & Le´ger 1989). The relative ratios of these bands, sometimes referred to as aromatic features in emission, have been shown to be consistent with the 6.2, 7.7, and 8.6 lm features in normal galaxies (generally having a 1 : 2 : 1 ratio, Helou et al. 2000). These features are nearly independent of heating indicators such as FIR colors, which can be taken as evidence that PAH features are emitted from grains that are not in thermal equilibrium but, rather, undergoing transient heating by individual photons (Helou et al. 2000). The 11.3 lm feature is more dependent on FIR color, as it has been shown to be more dominant in comparison with the other PAH bands in galaxies with colder dust temperatures (Lu et al. 1999). A possible interpretation is that, since the 11.3 lm feature is linked with neutral PAHs instead of ionized PAHs, these galaxies have a larger contribution of luminosity from regions with weaker radiation fields (Helou 2000). The 11.3 lm lineemission maps for our sample have the same general structure as the 15 lm continuum maps, although line emission dominates in almost all of the measured galaxies at a factor of 1.3. This indicates that PAHs, or more generally aromatic grains, are likely to be strong contributors to the MIR spectra of the galaxies in our sample. The mid-infrared continuum in the 15 lm band increases when excess heating of the small grains occurs due to starburst activity. The star-forming knots in the Antennae galaxies manifest this phenomenon (Vigroux et al. 1996) as a shift in the thermal emission peak toward shorter wavelengths. We find no evidence for correlations between global 11.3/15 lm colors and normalized MIR luminosity, galaxy component separation (SEP), FIR colors, or Hubble type. Such correlations would be surprising in view of the small SEP range. The pair with the highest 60/100 lm color, CPG 552, contains the galaxies with the lowest 11.3/15 lm colors [ f(LW8)/f(LW3) d 1], possibly as a result of increased heating of the small grains emitting at 15 lm.

565

Comparison of the distribution of MIR and H emission serves as a diagnostic of both the dust heating sources and physical constituents of the interstellar medium. Figure 4 displays cross-sectional cuts through the pair images (along the major axis) at MIR and (convolved) H wavelengths. In general, the structure is similar in the two imaging bands for all of the pairs, and a relatively constant ratio is found across the components in all but three of the eight pairs for which they were computed. Both components of CPG 552 and the spiral member of CPG 591 show a mid-infrared excess in the nuclei. Recent optical spectra of the CPG 552 components (Kewley et al. 2001) are classified as Seyfert 2(e) and H ii(w). There appears to be a nuclear mid-infrared deficit in the Seyfert component (NGC 5953) of CPG 468 because the MIR to H ratio drops at the nucleus and produces a mid-infrared excess ring in a ratio image (Fig. 5). It is possible that this ring is an artifact of the convolution process, but there is no evidence for this in any of the other ratio images. This may represent evidence for a dusty ring that increases H extinction and small-grain emission. Detailed spectroscopy of this galaxy (Gonza´lez Delgado & Perez 1996) shows evidence for a starburst ring surrounding the Seyfert nucleus. The presence of a strong [N ii]/H ratio could also enhance the relative nuclear component flux with respect to the circumnuclear H ii regions. The MIR data for the spiral member of CPG 591 (NGC 7753) had the highest effective resolution because of its large apparent size. Figure 6 shows a mid-IR–to–H ratio map that reveals a relative mid-IR excess in the nucleus, which may be the result of excess nuclear extinction of H. The Figure 2 images allow a full two-dimensional comparison of FIR and H, which show emission knots in CPG 460 (edge of spiral disk), 542 and 591, where H is anomalously strong relative to MIR. These seem unlikely places for an [N ii]/H reversal and may indicate emission regions with unusual excitation conditions. 6. ISOPHOT IMAGING

In normal galaxies the FIR is dominated by thermal emission from large dust grains. The IRAS survey mapped the sky at 60 and 100 lm and detected a large fraction of galaxy pairs brighter than V 15.0. Beam resolution in the IRAS survey allowed resolution of only the widest pairs, although this depends somewhat on the orientation of the pair’s major axis relative to the IRAS scan direction. The use of ISOPHOT enabled improved resolution maps (as good as the best IRAS deconvolution processed data; Surace et al. 1993) and higher sensitivity. Figure 7 shows 60 and 100 lm ISOPHOT contours overlayed on R-band images for the 10 mixed pairs with new FIR data. ISOPHOT detections can be grouped into pairs with emission from one or both components. The double-detection pairs include CPG 439, 468, 552, 579, and 591. The FIR emission from CPG 439 is not resolved into separate components, but the emission extension toward the southwest companion is seen in both bands, leading us to infer detection of the companion. A long tidal tail extends beyond the companion toward the southwest, so we cannot rule out that the emission is associated with this feature. Both components in CPG 468, 552, 579, and 591 are unambiguously resolved. The Seyfert/ starburst CPG 468w (NGC 5953) dominates the 60 lm emission by a factor of 3. The compact member of CPG 552 dominates its spiral companion by factors of 4 and 2 at

566

DOMINGUE ET AL.

Vol. 125

Fig. 4.—Cross-sectional plots of the ISOCAM pair images most likely to have extended mid-infrared emission. The solid line is the mid-IR (LW3 and LW8 average) and the dashed line represents the H cross-section. The diamonds are the ISOCAM/H ratio at each point in the cross-section. The vertical scale is the same for the ratio values. In pairs where only one galaxy was detected, cross-sections are along the major axis of that component.

60 and 100 lm, respectively. CPG 579 contains a detected S0, but its spiral dominates the pair at both 60 and 100 lm by a factor of 2. CPG 591w (NGC 7752) is apparently only a (relatively compact) starburst, which shares the pair’s 60 lm emission equally with the spiral, while the considerably larger spiral dominates by a factor 2 the cirrus-dominated emission at 100 lm. Pairs CPG 460, 542, 547, 548 and 572 are dominated by FIR emission from the spiral components with all early-type components undetected below the 3 level. These are the classic mixed-morphology pairs. The weakest (observed at 100 lm only) spiral detection in the sample involves CPG 548w, which is also the spiral component with lowest H equivalent width (see Table 6). This reflects the classic FIR-H correlation. The unusually strong detections of the elliptical/compact companion on the end of a tidal arm in CPG 552 is likely due to its Seyfert nucleus. These classifications indicate that five of the seven ISOPHOT imaged pairs with H and MIR classifications of one or no strong emitters also match the FIR description. All double component emitters at MIR and H wavelengths that were also imaged with ISOPHOT show corresponding double detections at 60 and 100 lm. 7. COMMENTS ON INDIVIDUAL PAIRS

CPG 79: A remarkable early-type pair assigned to the least-isolated category in the CPG. The POSS1-based Sb+E

classification is corrected to SB0+S0 using CCD images. The pair is kinematically assigned to the weak cluster AWM 7, where it lies on the southwest outskirts (Koranyi & Geller 2000). This makes the rather large component velocity difference (largest of our sample) and early-type morphologies less surprising. The assigned ATM interaction class should be viewed with suspicion. The low levels of H and MIR emission are consistent with these properties. All significant H, ISOCAM, IRAS FIR (Hernańdez Toledo 1998), and radio continuum (NVSS) emission appears to coincide with the nucleus of the SB0 component (CPG 79e). CPG 83: Well-studied, unambiguously interacting and possible coalescing pair involving a disrupted Sb (Seyfert 2) spiral (83e) and an elliptical (83w) with distorted outer envelope. It was assigned an ATM interaction type because of isophotal overlap. The pair shows kinematic (Bransford et al. 1999) and morphological similarities with CPG 29, although in the latter case the component velocity difference is 3 times larger (Marziani et al. 1994). A giant extranuclear H ii region located in a distorted ring of emission regions dominates the H luminosity (Fig. 3). The H ii complex contributes 12% of the total H flux, compared with 8% from the nucleus. Much CO emission is associated with the emission regions west of the nucleus (the galaxy is more CO luminous than Arp 220, Gao et al. 1997). The elliptical com˚ ), centrally conponent NGC 1143 shows weak (EW = 8 A centrated H emission. A dwarf galaxy at the northeast

No. 2, 2003


Fig. 5.—Mid-infrared to H ratio image of CPG 468. The west component, NGC 5953, exhibits a relative mid-IR excess in a partial ring.

edge of the pair and the larger neighbor 2 pair diameters northwest show concordant redshifts. Unfortunately, no ISO pointings except for Short Wavelength Spectrometer spectra (Prieto & Viegas 2000). CPG 202: A classic mixed SBb+E pair, which may form a triplet with a fainter neighbor to the northeast. A DIS interaction type reflects asymmetries in the spiral disk. The component velocity difference is quite large, 700 km s1. NGC 2872, the much-studied elliptical, shows centrally concen˚ . The NVSS database trated H emission with EW = 12 A shows that both components are continuum radio sources. No ISO pointings. CPG 234: Very well studied classic mixed SBb+E pair classified as interaction type LIN because of the spatial

Fig. 6.—Mid-infrared–to–H ratio image of CPG 591 eastern spiral galaxy (NGC 7753). The nucleus exhibits a mid-infrared excess possibly due to extinction of H.

567

proximity of the components and apparent linkage through one of the spiral arms. Star formation regions defined by the H emission trace out a well-defined ‘‘ integral sign ’’ pattern (Fig. 3). The elliptical component displays a large H ˚ , probably related to the LINER nucleus. PreEW = 45 A vious studies show the central region of the spiral to have two sources of excitation, including a Seyfert 2 and a starburst (Gonza´lez Delgado & Perez 1997). The same study finds no overall disk star formation enhancement. A KAO 100 lm observation shows 100 lm emission only from the spiral component (Bushouse et al. 1998). There are no known ISO pointings. CPG 394: The most hierarchical mixed (SBb+S0) pair in our sample. The DIS interaction class reflects the distorted outer envelope of the dwarf early-type companion and the possibility that the outer spiral arms may have been tidally generated. The spiral component exhibits a LINER nucleus and diskwide star formation, while the S0 shows evidence for kinematic decoupling in the core (Rampazzo et al. 1995). The S0 also shows an extranuclear emission region offset toward the southeast. There are no ISO pointings. CPG 414: A complex late-type galaxy paired with an SB0. The late-type shows chaotic barred and spiral structure, suggesting that it may be a spiral galaxy that was disrupted in an extreme encounter/collision. The DIS interaction class also extends to the SB0, where faint outer extensions are seen on broadband images. The bar of the late-type member exhibits no H emission consistent with the interpretation of this galaxy as a spiral, rather than an irregular of some kind. H regions of modest intensity are spread throughout the spiral arms, with the strongest emission from the region between the two components. The SB0 shows a strong H ii–class emission spectrum (Contini, Considere, & Davoust 1998). There are no ISO pointings. CPG 439: The members of this pair are difficult, and a best guess would be S0/Sa+S0. A high surface brightness arc or obscured ring partially surrounds the nucleus of the primary component on unpublished good-seeing BV images. CPG 439 merits interaction class LIN because of a long tidal tail that extends from the secondary in a direction opposite the primary. A faint and diffuse bridge connecting the components is also suspected and may be confirmed on the H image. The tail shows no distinct regions of H emission. Moderately strong radio emission appears to coincide with the primary. BV colors for the primary and secondary (0.66 and 0.62) are similar to measures of the countertidal tail (BV = 0.71; Schombert, Wallin, & Struck-Marcell 1990). All of the evidence is consistent with the aftermath of a strong interaction event. The northern galaxy is dominant in our ISO MIR and FIR imaging, although both components are detected. The spiral nucleus is a good candidate for Seyfert 2 activity, based upon the available data for the five known Seyferts in our sample. CPG 460: A possible classic Sb+E pair, although no discrete emission regions are seen in the highly inclined candidate spiral component. The DIS interaction class reflects obvious isophotal asymmetries in both galaxies. The H image of the spiral component shows a string of emission regions on its southern edge, possibly part of a ring. This is reflected in the ISOCAM images, which show a similar intensity distribution, except for the anomalously bright H knot at the southwest end of the disk. The spiral is also detected in the 60 and 100 lm ISOPHOT images. The ellip˚) tical component shows weak H emission (EW = 3 A

Fig. 7.—Contours of ISOPHOT 60 and 100 lm images for 10 galaxy pairs from our sample overlaid on their Digitized Sky Survey images. Contours begin at 3–4 of the noise levels with 1–3 intervals. The orientation of the images is such that north is up and east is left. Each image is the same scale as the CPG 439 image.

MIXED-MORPHOLOGY BINARY GALAXIES. I. comparable within the errors to normal ellipticals in the Kennicutt & Kent (1983) sample. And it is detected at the S 4 mJy level in the 11.3 and 15 lm maps, which are among the weakest detection in our sample. CPG 468: At first glance an SBc spiral with an elliptical or compact galaxy on the end of a tidal arm, meriting an LIN interaction classification. However the companion is the dominant MIR/FIR source and harbors a Seyfert nucleus. The pair component velocity difference is consistent with zero. The spiral component of this much-studied pair shows a distorted disk filled with emission regions knots, as can be seen in H and a LINER (Dahari 1985) nucleus. It has already been argued (Jenkins 1984) that the companion may be accreting gas from the spiral component. A crossfuelled Seyfert2+starburst? HST archival images reveal a strong flocculent spiral pattern in the companion. Perhaps interactions can both create and destroy spiral galaxies? Additional multiwavelength ISO data can be found in Clavel et al. (2000). CPG 536: This pair was not included in the CPG E+S sample. Although classified an S+S pair it appears to be an edge-on spiral paired with a compact galaxy. The pair was assigned an ATM interaction class. Our H image shows chaotic emission patches surrounding the compact companion, which appears to confirm that the galaxies are interact˚ , which ing. The compact galaxy has a large H EW = 70 A is one of the largest in the sample and reflects its Seyfert 2 classification (Keel 1996). It is one of three Seyfert 2s in our sample (with CPG 468 and 552) that are found in reasonably compact companion galaxies. CPG 542: A hierarchical pair composed of a large Sc spiral and a dwarf elliptical. The pair was assigned an ATM interaction class, but perhaps DIS is more appropriate, considering the asymmetries visible in the spiral component, which are reinforced by our H image. The side of the disk nearest the elliptical companion shows a prominent string of emission regions without counterparts on the far side. Rotation curve data also shows evidence for interaction (Rampazzo et al. 1995). Two bright H regions in the southwest arm were resolved on the MIR images. The spiral is also detected at both 60 and 100 lm in the ISOPHOT maps. The elliptical shows ˚ ) and faint corresponding weak H emission (EW = 3 A MIR emission, which are consistent with the lack of gas reported by Rampazzo et al. (1995). CPG 547: A classic (edge-on) S+E pair. The LIN interaction class reflects the small separation of the pair with a possible luminous bridge. Only the spiral is detected on our ISOPHT images. A bright emission region is detected near the edge of the spiral in the overlap region on our H image. The spiral and elliptical components show H EW = 14 ˚ , respectively, which are typical values for ‘‘ unenand 1 A hanced ’’ galaxies. Keel (1993) reports that the spiral of the pair displays a disturbed rotation curve. CPG 548: A hierarchical Sab+E pair that is strikingly similar to CPG 542. It is less clear that a DIS interaction class is warranted in this case. H emission from the spiral is mostly concentrated in an outer ring, as well as a near nuclear complex on the northwest side, with a total EW = 7 ˚ . The center of the spiral component is weakly detected at A 60 and 100 lm in our ISOPHOT images. The elliptical ˚ ), again comparable within the shows weak H (EW = 2 A errors to normal isolated ellipticals (Kennicutt & Kent 1983). Only weak kinematic evidence (Rampazzo et al.

569

1995) to support the hypothesis that the galaxies are interacting. CPG 552: The pair CPG 552 is composed of a spiral with a compact (elliptical?) companion on the end of one of the apparently tidally generated arms. This motivated a LIN interaction class. The spiral shows an asymmetric distribution of H ii regions around its nucleus (the remnants of a disk or perhaps a bar?). The western tidal arm shows H emission and is not simply stellar debris from the interaction. The inner part of the opposite arm is also detected, but the actual bridge must be predominantly stellar. The arm is also weakly detected at 15 lm, indicating that it contains both dust and gas. As with H, the central region of the spiral and the compact companion dominate the MIR emission. FIR maps show that the emission is more concentrated in the compact companion. This is less surprising when the surprising Seyfert 2 nature of the companion is known. Another candidate cross-fuelled Seyfert galaxy. CPG 572: CPG 572 is a classic Sb+E pair with a DIS interaction class motivated by the somewhat chaotic spiral structure in the primary component. The spiral is a strong MIR source, with S15 = 60 mJy. The spiral arms are resolved in the H, along with significant emission from its inner disk. The spiral is also detected at both 60 and 100 lm in our ISOPHOT maps. The elliptical component shows an ˚ , but, surprisingly, no corresponding MIR/ H EW = 9 A FIR detections (3 detection upper limits 0.5 mJy). CPG 579: CPG 579 may be an Sbc+S0 pair, but unpublished CCD images show evidence for faint outer spiral arms in the S0 component. Our MIR/FIR and H data support the latter interpretation. Part of the confusion may be a reflection of the relatively large recession velocity (10,000 km s1) of this pair. A DIS interaction classification reflects structural asymmetries in both components. The northern galaxy is an asymmetric spiral with considerable nonnuclear H emission on the north and west sides, as well as a bright region south of the nucleus in the direction of the companion. This spiral is detected in all MIR/ FIR images. The southern component shows a significant amount of centrally concentrated line emission, with ˚ , and is also detected in all MIR/FIR pointings, EW = 44 A although it is very weak at 100 lm. CPG 591: One of four hierarchical (component majoraxis ratio 3–5) pairs where a large spiral is accompanied by a dwarf early-type galaxy. At first glance this wellstudied pair would be classified Sc+S0 and would be assigned interaction class LIN, because the companion lies on the end of an apparently tidally extended arm. The dwarf spheriodal classification is called into question by the strong H emission and MIR/FIR detections. The spiral displays weak nuclear H emission, as well as diskwide sources. H emission is also detected the full length of the arm that connects with the companion, making this one of the better candidates for cross-fuelling activity. The gradient in velocity along the arm is consistent with this interpretation (Marcelin et al. 1987). The two largest H ii complexes in the southern arm appear as strong sources in the MIR ISOCAM maps along with an MIR-emitting H ii region on the northeast side. The northern arm is detected in the ISOCAM images. Nuclear MIR emission, while not fully resolved, is comparable to that of the H emission. This does not agree with the H nuclear/disk emission ratio as the unresolved nuclear MIR is a larger fraction of the total emission than in H. The companion, shows a larger 15 lm flux than the

570

DOMINGUE ET AL.

large spiral. It also dominates the 60 lm emission of the pair in the ISOPHOT images indicating a warmer color and high star formation rate. Seyfert activity cannot be far behind, if other pairs in this sample are any indication. At present both components show H ii–class spectral line ratios (Mazzarella & Boroson 1993). The companion shows a large velocity gradient (almost 500 km s1, Marcelin et al. 1987) within 2500 and the largest H EW in our sample. ISO has independently detected FIR emission from cold dust in this system (Siebenmorgen, Krugel, & Chini 1999). 8. SUMMARY AND DISCUSSION

Analysis of new H and ISO images for mixed pairs leads us to reject the hypothesis that they always show signatures of active star formation (or active galactic nuclei) that are dominated by the spiral component. New R-band images indicate that nine of the pairs in this sample are true E+S pairs. The new H images for these classical E+S pairs reveal that many of the early-type galaxies are indeed systems that show little or no IR or optical signatures of star formation. However, some of the early-type components do show significant levels of IR and H emission that indicate both thermal (star formation) and nonthermal (AGN) activity. This activity appears to favor disky early-type systems (CPG 552 is a notable exception), such as dwarf lenticular or spheroidal systems. Some are so gas-rich, however, that they are difficult to classify at all. Not surprisingly, optical/IR enhancements are also seen in the spiral components. The spirals show an increase in relative nuclear emission compared with field spirals (including two known Seyfert2 nuclei), and all spirals with nuclear emission have corresponding knotty disks except for CPG 439. ISO has given new insight into the mid-infrared behavior of the pairs, especially at MIR bands where many were undetected by IRAS (especially at 12 lm). Convolution of the H images to match ISOCAM resolution shows that H and MIR emission follow each other quite closely, with only a few exceptions. This represents a two-dimensional confirmation of the strong one-dimensional correlation between global H and MIR/FIR that has been known for some time (Kennicutt et al. 1987). The two-dimensional images show that this spatial correlation extends across a wide range of emission levels and optical morphologies. ISOCAM images show that galaxies with significant H emission are also luminous 15 lm sources, which can be taken as an indication that the detected galaxies contain small dust grains as well as PAH molecules, since the 11.3 lm maps reveal both similar structure and comparable flux levels to the 15 lm maps. ISOPHOT imaging has enabled the determination of pair component FIR flux ratios for 10 of the pairs. Five of the ISOPHOT maps reveal a far-IR contribution from galaxies other than the spiral. These five galaxies are typed as (1) lenticular S0 galaxies in three pairs, (2) as an I0/S0 in the fourth case, and (3) as a elliptical/compact in the remaining source. The questions posed regarding the origin of mixed pairs may be addressed here, beginning with the issue of whether or not observed misclassifications can eliminate the problem. We confirm earlier conclusions (Rampazzo & Sulentic 1992) that a significant mixed-pair population exists. The existence of mixed pairs is perhaps a problem only if one regards them as (near) primordial systems. Perhaps the sol-

Vol. 125

ution lies in looking at compact groups where one also sees a significant morphological mix in strongly interacting systems (Sulentic 1991). If both compact groups and pairs form continuously out of loose groups, then mixed pairs are expected some of the time. It seems likely that a true excess of early-type members are created by the interactions in compact groups (see Stephan’s Quintet for a good example, Sulentic et al. 2001). Statistical evidence suggests that mixed pairs are found in about the right fraction if they form in loose groups (Sulentic 1989, 1991). A smaller interactionproduced early-type population in pairs may be offset by cross-fuelling activity that turns some of them (back?) into gas-rich systems. At least seven of the true ellipticals appear to exhibit no unusual H or IR characteristics. They are members of classical E+S pairs. At first glance one of the spiral components CPG 552w might be seen as a candidate S0 with tidally generated arms (see Rampazzo & Sulentic 1992), but detection of the arms in both H and MIR shows that they have a significant nonstellar component and were likely pulled from a spiral disk. The most ambiguous classifications involve the S0 galaxies, many of which have strong signs of interaction and star formation. If these galaxies represent a class of stripped spirals, then they might represent a morphological transition stage. Whether or not they show secular evolution, many of them now contain considerable nonstellar material (also true for the compact elliptical in CPG 552). A surprising number in this sample show starburst and Seyfert activity. We did not directly select these pairs for such activity, although an IRAS-bright requirement was imposed and can be said to represent an enhanced probability for such activity. There is no single hypothesis that could explain the existence of these diverse pairs. The observed H and MIR/ FIR emission from the S0’s and from a few ellipticals may be the result of (1) spirals misclassified as S0’s, (2) pairs as a stage in the coalescence of compact groups, with the earlytype as a merger remnant, (3) transfer of nonstellar matter from the gas-rich to the gas-poor member, (4) spiral galaxy ‘‘ harassment ’’ resulting in gas stripping and formation of lenticular systems, and (5) a cluster-like harassment involving a gravitational response that heats the disk and sends 90% of available gas to the center leading to nuclear starbursts and AGNs. Hypothesis 1 probably applies to CPG 579, where both components are likely to be spirals. CPG 439 is a second, less certain, possibility in which the morphology of the second galaxy shows no spiral characteristics, except for a long counter tidal filament. Hypothesis 2 is strongly disfavored because (1) the earlytype components are rarely first-ranked and are often dwarfish, (2) no early-type in this sample shows a diameter (or luminosity) DE > 1.5 DS, and (3) none of them show H ˚ or show LIRG/ULIRG emission. EW > 100 A Hypothesis 3 involves cross-fueling, which appears to be a possibility, especially for CPG 234, 468, 536, 552, and 591. All except CPG 536 show apparent connections between the companion and an arm of the gas-rich primary. Three of the companions are Seyfert2 or composite Seyfert-starburst, while the other two are H ii–starburst systems. Numerical simulation for CPG 591 supports a cross-fueling scenario. While CPG 468 shows an emission bridge between the galaxies, simulations have argued against cross-fueling for this pair.

No. 2, 2003


Since most of the ellipticals show normal emission, we are left with the need to explain two high H EW lenticular components of CPG 394 and 414. Both of these objects appear as featureless S0’s, and the S0 of CPG 414 is a Markarian object with strong ultraviolet excess. The former is likely a low-level victim of its much larger and more massive spiral neighbor, while the latter most likely victimized at a high level its spiral neighbor. The multitude of interaction signs suggest that most of these pairs have been interacting for a long time. This is especially true if the lack of an H i deficit in the spiral components of mixed pairs (Zasov & Sulentic 1994) can be reconciled with the observed average star formation enhancement (Xu & Sulentic 1991). That conclusion was reached by using component specific angular momentum as a normalization factor. New work (Hernańdez Toledo et al. 2001) suggests that a possible systematic dissipation of angu-

571

lar momentum in the spiral components. Such a specific angular momentum deficit has masked the H i deficit. J. S. and D. D. acknowledge support under NASA-JPL contract 961557. ISOCAM and ISOPHOT data were reduced using CIA and PIA, respectively, which were joint developments of the ESA Astrophysics Division and the ISOPHOT consortium. D. D. is grateful for ISO reduction training received Les Houches summer school on Infrared Space Astronomy Today and Tomorrow. D. D. acknowledges W. C. Keel for many useful discussions. C. X., J. M., and Y. G. are supported by the Jet Propulsion Laboratory, California Institute of Technology. This research has made use of the NASA/IPAC Extragalactic Database (NED), which is operated by JPL, Caltech, under contract with the NASA. This work has received partial financial support from the Italian Space Agency (ASI contract I/R/037/01).

REFERENCES Koranyi, D. M., & Geller, M. J. 2000, AJ, 119, 44 Alladin, S. M. 1965, ApJ, 141, 768 Klaas, U., Laureijs, R. J., Radovich, M., & Schulz, B. 1998, ISOPHOT Arp, H. C. 1966, Atlas of Peculiar Galaxies (Pasadena: Caltech) Calibration Accuracies, SAI/98-092/Dc Bransford, M. A., Appleton, P. N., McCain, C. F., & Freeman, K. C. 1999, Laureijs, R. J., Klaas, U., Richards, P. J., & Schulz, B. 1999, ISOPHOT ApJ, 525, 153 Data Users Manual, SAI/95-220/Dc Bushouse, H. A. 1987, ApJ, 320, 49 Lemke, D., et al. 1996, A&A, 315, L64 Bushouse, H. A., Telesco, C. M., & Werner, M. W. 1998, AJ, 115, 938 Lu, N., et al. 1999, in The Universe as Seen by ISO, ed. P. Cox & M. F. Cesarsky, C., et al. 1996, A&A, 315, L32 Kessler, SP-47 (Noordwijk: ESA), 929 Chaterjee, T. 1987, Ap&SS, 135, 131 Macchetto, F., Pastoriza, M., Caon, N., Sparks, W. B., Giavalisco, M., Clavel, J., et al. 2000, A&A, 357, 839 Bender, R., & Capaccioli, M. 1996, A&AS, 120, 463 Combes, F., Prugniel, P., Rampazzo, R., & Sulentic, J. W. 1994, A&A, 281, Marcelin, M., Lecoarer, E., Boulesteix, J., Georgelin, Y., & Monnet, G. 725 1987, A&A, 179, 101 Contini, T., Considere, S., & Davoust, E. 1998, A&AS, 130, 285 Markarian, B. M., Lipovetskii, V. A., Stepanian, J. A., Erastova, L. K., & Dahari, O. 1985, ApJS, 57, 643 Shapovalova, A. I. 1989, Soobshchen. Spezial. Astrof. Obs., 62, 5 Demin, V., Zasov, A., Dibai, E., & Tomov, A. 1984, Soviet Astron., 28, 367 Marziani, P., Keel, W. C., Dultzin-Hacyan, D., & Sulentic, J. W. 1994, de Mello, D. F., Sulentic, J. W., de Souza, R. E., Reduzzi, L., & Rampazzo, ApJ, 435, 668 R. 1996, A&A, 308, 387 Mazzarella, J. M., & Boroson, T. A. 1993, ApJS, 85, 27 Gabriel, C. et al. 1997, in ASP Conf. Ser. 125, Astronomical Data Analysis Paturel, G. 1976, A&A, 53, 305 Software and Systems VI, ed. G. Hunt & H. E. Payne (San Francisco: ———. 1979, Bull. Inf. CDS, 17, 14 ASP), 108 Pogge, R. W., Eskridge, P. B. 1993, AJ, 106, 1405 Gao, Y., Solomon, P. M., Downes, D., & Radford, S. J. E. 1997, ApJ, 481, Prieto, M. A., & Viegas, S. M. 2000, ApJ, 532, 238 L35 Puget, J. L., & Le´ger, A. 1989, ARA&A, 27, 161 Gonza´lez Delgado, R. M., & Perez, E. 1996, MNRAS, 281, 781 Rampazzo, R., Reduzzi, L., Sulentic, J. W., & Madejsky, R. 1995, A&AS, ———. 1997, MNRAS, 284, 931 110, 131 Helou, G. 2000, in Infrared Space Astronomy, Today and Tomorrow, ed. Rampazzo, R., & Sulentic, J. W. 1992, A&A, 259, 43 F. Casoli, J. Lequeux, & F. David (New York: Springer), 339 Reduzzi, L., & Rampazzo, R. 1995, ApL&C, 30, 1 Helou, G., Khan, I., Malek, L., & Boehmer, L. 1988, ApJS, 68, 151 Schombert, J. M., Wallin, J. F., & Struck-Marcell, C. 1990, AJ, 99, 497 Helou, G., Lu, N. Y., Werner, M. W., Malhotra, S., & Silbermann, N. Siebenmorgen, R., Krugel, E., & Chini , R. 1999, A&A, 351, 495 2000, ApJ, 532, L21 Siebenmorgen, R., Starck, J. L., Cesarsky, D. A., Guest, S., & Sauvage, M. Hernańdez Toledo, H. M. 1998, Ph.D. thesis, Univ. Nac. Autonoma 1996, ISOCAM Data Users Manual, SAI/95-222/Dc Me´xico Sulentic, J. W. 1989, AJ, 98, 2066 Hernańdez Toledo, H. M., Dultzin-Hacyan, D., & Sulentic, J. W. 2001, AJ, ———. 1991, in Morphological and Physical Classification of Galaxies, ed. 121, 1319 G. Longo, M. Capaccioli, G. Busarello (Dordrecht: Kluwer), 293 ———. 1958, Medd. Lunds Astron. Obs, Ser. 2, 136, 1 Sulentic, J. W., Rosado, M., Dultzin-Hacyan, D., Verdes-Montenegro, L., Jenkins, C. R. 1984, ApJ, 277, 501 Trinchieri, G., Xu, C., & Pietsch, W. 2001, AJ, 122, 2993 Johnson, H. L. 1965, ApJ, 141, 923 Surace, J. A., Mazzarella, J., Soifer, B. T., & Wehrle, A. E. 1993, AJ, 105, Karachentsev, I. D. 1972, Comm. Spec. Astrophys. Obs. USSR, 7, 1 864 ———. 1987, Double Galaxies (Moscow: Nauka) Tifft, W. 1982, ApJS, 50, 319 Karachentseva, V. 1973, Comm. Spec. Astrophys. Obs. USSR, 8, 1 Tuffs, R. & Gabriel, C. 2003, in ASP Conf. Ser., Galactic Star Formation Keel, W. C. 1993, AJ, 106, 1771 across the Stellar Mass Function, ed. J. M. De Buizer (San Francisco, ———. 1996, AJ, 111, 696 CA: ASP), in press Keel, W. C., Kennicutt, R. C., Hummel, E., & van der Hulst, J. M. 1985, Toomre, A., & Toomre, J. 1972, ApJ, 178, 623 AJ, 90, 708 Vigroux, L., et al. 1996, A&A, 315, L93 Kennicutt, R. C., Jr. 1983, ApJ, 272, 54 Vorontsov-Velyaminov, B. A. 1959, Atlas and Catalogue of Interacting Kennicutt, R. C., Jr., & Kent, S. M. 1983, AJ, 88, 1094 Galaxies (Moscow: Sternberg Inst.) Kennicutt, R. C., Jr., Roettiger, K. A., Keel, W. C., van der Hulst, J. M., & Xu, C., Gao, Y., Mazzarella, J., Lu, N., Sulentic, J. W., & Domingue, D. L. Hummel, E. 1987, AJ, 93, 1011 2000, ApJ, 541, 644 Kessler, M. F., et al. 1996, A&A, 315, L27 Xu, C., & Sulentic, J. W. 1991, ApJ, 374, 407 Kewley, L. J., Heisler, C. A., Dopita, M. A., & Lumsden, S. 2001, ApJS, Zasov, A. V., & Sulentic, J. W. 1994, ApJ, 430, 179 132, 37