There are many other interesting aspects of interaction induced starbursts, and in the remaining space I simply mention a few.
Star Formation Knots: When viewed with sufficient resolution the star-forming regions in Arp 299 and NGC 4038/9 are found to break up into many distinct knots along with a diffuse component. Figure 5 shows an HST FOC image of NGC 3690 [50, 51]. These knots have typical diameters of less than 10pc and may evolve into globular clusters [50, 52]. However these knots are not unique to mergers [50, 53, 54, 55], and it is not clear if this mode of star formation is increased in such interactions and if it is related to the enhanced SFE. It may just be an important mode of star formation in general. A detailed comparison of the luminosity functions of such knots from different types of environments would help resolve this question.
 |
Figure 5. HST FOC 2200 Å observation of the B-C-C' complex in NGC 3690 (Arp 299 West) from Vacca 1995 [51]. These observations show that many of the massive stars in this system are confined to very bright, compact knots. |
The Return of Tidal Debris: Due to the strong tides
experienced during merging encounters, appreciable amounts of stars and
gas are lifted high above the merging systems into tidal features.
These features frequently exhibit significant substructure, the largest
having observational properties typical of dwarf galaxies
[56,
57,
58,
59].
Much of this material remains bound to the
remnant on long-period orbits, and will take several Gyr to fall-back
[60,
61].
Tidal clumps that are far out along the tails may be
able to avoid tidal stripping when they fall back towards the remnant
and should become long-lived dwarf companions
[61],
while the more tightly bound material will fall back into the
remnant. The stellar component of these tails will wrap coherently in
the central potential, giving rise to fine structure features
[60]
while the gas may
feed a prolonged period of low-level star formation. The largest clumps
have as much as a few 108
M
of H I
[58,
59],
and their return may give rise to smaller bursts. The overall star
formation history in merger events should be similar to that illustrated
by Worthy (this volume).
Because of the timescales and amounts of gas involved, we therefore do not expect two merging spirals to turn quickly into an elliptical, but rather for there to be a series of transitions, e.g. to an S0pec, to an S0, to a dust lane elliptical, etc. [59]. By the time the more obvious signs of its merger origin have faded and the remnant has evolved into a bona fide elliptical, the stars formed during the merger induced starburst will have aged 2-5 Gyr, leaving very little indication of a merger origin in the remnants broad band colors. This picture is quite similar to the one emerging from studies of cluster populations at high redshift (see Dressler, these proceedings).