2003ApJ...585..281H

We have conducted a deep search for neutral hydrogen gas associated with the faint optical ``jets'' of NGC 1097 using the Very Large Array. Measurable H I would have been expected if the jets were tidal in origin given their moderately blue optical and near-infrared colors. The jets are free of H I emission to a limiting surface density (Σ_HI_) of 0.06 M_☉pc^–2 (3 σ) over a 1102 km/s velocity range. We also rule out extended H I emission down to 0.02 M_☉pc^–2 (3 σ, ΔV=45 km/s) within a 4' FWHM aperture centered on the right-angle turn in jet R1. We have detected an H I source [M_HI_=(5.1±1.0)x10⁶ M_☉] coincident with a small edge-on spiral or irregular galaxy (NGC 1097B) 12' southwest of NGC 1097, situated between two jets. Two other ∼10⁶ M_☉H I point sources in the field are considered marginal detections. Neither are associated with the optical jets.

The jets' radio-X-ray spectral energy distribution is most consistent with starlight. However, from their morphology, optical/near-infrared colors, and lack of H I, we argue that the jets are not tidal tails drawn out of NGC 1097's disk or stars stripped from the elliptical companion NGC 1097A. We also reject in situ star formation in ancient radio jets as this requires essentially 100% conversion of gas into stars on large scales. Instead, we conclude that the jets represent the captured remains of a disrupted dwarf galaxy that passed through the inner few kiloparsecs of NGC 1097's disk.

We present N-body simulations of such an encounter that reproduce the essential features of NGC 1097's jets: A long and narrow ``X''-shaped morphology centered near the spiral's nucleus, right-angle bends, and no discernible dwarf galaxy remnant. A series of jetlike distributions are formed, with the earliest appearing ∼1.4 Gyr after impact. Well-defined X shapes form only when the more massive galaxy has a strong disk component. Ram-pressure stripping of the dwarf's interstellar medium would be expected to occur while passing through NGC 1097's disk, accounting for the jets' lack of H I and H II. The remnants' (B-V) color would still agree with observations even after ∼3 Gyr of passive evolution, provided the cannibalized dwarf was low-metallicity and dominated by young stars at impact.