X-RAY-EMISSION, OPTICAL NEBULOSITY AND DUST IN EARLY-TYPE GALAXIES - I - THE DUSTY NEBULAR FILAMENTS IN NGC 5846

We present new optical imagery of the X-ray bright elliptical galaxy N GC 5846, the dominant galaxy of a small group of galaxies. A filamenta ry dust lane with a dust mass of similar to 7 10(3) M. is detected in the central few kpc of NGC 5846. The optical extinction properties of the dust features are consistent with those of dust in our Galaxy. The morphology of the dust features are strikingly similar to that observ ed for the optical nebulosity and the X-ray emission. A physical conne ction between the different phases of the interstellar medium therefor e seems likely. We discuss three different options for the origin of t he dusty nebular filaments: Condensation out of a cooling flow, mass l oss of late-type stars within NGC 5846, and material donated by a smal l neighbouring galaxy. We conclude that the dust as well as the optica l nebulosity are most likely products of a recent interaction with a s mall, relatively gas-rich galaxy, probably of Magellanic type. Dust gr ains in the dust lane are destroyed by sputtering in the hot, X-ray-em itting gas in less than or similar to 10(7) yr, which is shorter than the crossing time of a (small) galaxy through the central 5 kpc of NGC 5846. This indicates that the dust must be replenished to be consiste nt with the observed dust mass, at a rate of similar to 10(-3) M. yr(- 1). We argue that this replenishment can be achieved by evaporation of cool, dense gas cloudlets that were brought in during the interaction . The evaporation rate of cool gas in NGC 5846 is consistent with the ''mass deposition rate'' derived from X-ray measurements.The energy lo st by the hot gas through heating of dust grains and evaporation of co ol gas clouds in the central few kpc of NGC 5846 is adequately balance d by heat sources: transport of heat by electron conduction into the c ore of the X-ray-emitting gas and loss of kinetic energy of the in-fal ling galaxy. There does not seem to be a need to invoke the presence o f a ''cooling flow'' to explain the X-ray and optical data.