|Figure 26: The left panel of this figure shows an adaptively smoothed image of NGC 4261 as observed with the ACIS-S3 CCD on-board Chandra. The arm-like distribution of the X-ray sources is clearly seen, especially in the Southern part of the galaxy. The white ellipse is the same as the white ellipse in the optical image on the right. The white rectangle shows the field of view of the observation (which was taken in 1/2 subarray mode). The green circles show the discrete sources in the central saturated region of the image. The right panel shows an R-band DSS2 image of NGC 4261 with the X-ray sources shown as red ellipses. The white ellipse shows the D25 ellipse of the galaxy (the 25mag optical isophote), as in the Chandra image. (Figure from Zezas et al. 2003, ApJ, 599, L73.)
Simulations of galaxy mergers show that the colliding galaxies launch tidal tails consisting of gas and stars. (A spectacular example is the Antennae galaxies.) As the collision evolves and the two galaxies merge into a single object, the tidal tails disperse and become weaker. A few Gyr after the completion of the merger there is very little evidence for the violent past of the remnant. At the same time, gas and stars in the tidal tails may form clumps resembling dwarf galaxies, which continue to orbit the merger remnant. However, because of their very weak optical emission it is difficult to detect them against the much brighter light of the main galaxy. It is also possible that long after the completion of the merger there is still some gas left in the tidal tails, but because of its very low surface brightness is again very difficult to detect.
We believe that in the case of NGC 4261, localized starformation events were triggered either by a shock propagating along the tidal tails or by some of the "tidal dwarf galaxies" falling back on the merger remnant. The star-light from these star-forming regions is outshined by the light of the main galaxy, making their detection in optical images very difficult. However, it is possible to trace these star-formation events from the young X-ray binaries associated with them. High mass X-ray binaries (HMXBs) form ~ 10-100 times more efficiently than Low mass X-ray binaries (LMXBs) mainly because they are less susceptible to effects such as supernova kicks or commonenvelope phases which destroy most LMXBs before they reach their X-ray emitting phase. The difference in the production rates of these two classes of X-ray binaries can result in an anisotropic distribution of X-ray sources in an early type galaxy, where regions of recent star-formation are associated with an excess of X-ray sources. However, this excess of sources is visible only for as along as the lifetimes of the HMXBs (a few hundred Myrs). We can use this diagnostic to infer the violent past of some of the most serene-looking elliptical galaxies in only a few cases. However, because these star-formation events are expected to occur more frequently shortly after the merger, we would expect to find that anisotropic X-ray source distributions are more common among elliptical galaxies with more recent merger histories. We are currently investigating this.