There are many signs of recent or ongoing gravitational interactions in the Local Group, including the warped disks of the Milky Way, M 31, and M 33, the Magellanic Stream, and the integral-sign distortion of NGC 205, companion to M 31. However, the details of these interactions are often difficult to establish, and the cumulative effect of interactions not directly leading to mergers remains largely unknown.
Fortunately, there is now - in the Milky Way - some good, detailed evidence for interactions leading to accretions. Three pieces of evidence stand out as particularly reliable among the many that have been claimed.
First and most impressive is the Sagittarius dwarf galaxy, hidden from us
behind the Milky-Way bulge until its recent discovery
by Ibata et
al. (1994).
Located at a distance of 16 kpc from the
galactic center, this dwarf appears very elongated in a direction
approximately perpendicular to the galactic plane and is thought
to move in a nearly polar orbit with current peri- and apogalactic
distances of ~ 20 kpc and ~ 60 kpc, respectively
(Ibata & Lewis
1998).
Although it may have started out with a mass of as much as
1011 M
or as little as ~ 109
M
(Jiang & Binney
2000),
the dwarf is estimated to currently have a mass of
2 x 108-109
M and an
orbital period of about 0.7-1 Gyr.
It will probably disrupt completely over the next few orbits and will
then deliver its four globular clusters, one of which appears to be its
nucleus (e.g.
Da Costa &
Armandroff 1995),
to the halo of the Milky Way.
As Searle & Zinn (1978) conjectured already, similar accretions of gas fragments and dwarfs may have built this halo over a prolonged period. A second piece of evidence strongly supporting this view is the observed retrograde mean motion of certain subsystems of globular clusters (Rodgers & Paltoglou 1984; Zinn 1993). How could a monolithic collapse possibly have led to a 15% minority of slightly younger halo globulars orbiting in the opposite sense from the majority of old globulars and the disk itself? Accretions from different directions provide a natural explanation.
Most accretions into the halo must have occurred in the first 25%-30%
of the age of our Galaxy. Colors and inferred minimum ages of halo stars
suggest that by 10 Gyr ago such accretions had diminished to a trickle
and since then 
6
Sagittarius-like dwarfs can have been accreted
(Unavane et
al. 1996).
Hence, the ongoing accretion of Sgr Dwarf is
by now a relatively rare event.
However, a much more massive accretion may still lie in the future. This is suggested by the Magellanic Stream, the third piece of good evidence for a relatively strong interaction involving the Milky Way. This stream of H I extends over 120° in the sky, arching from the Magellanic Clouds through the south galactic pole to declination -30°, where it was first discovered (Dieter 1965; Mathewson et al. 1974). After a long and tortuous history of interpretations, modern models based on a past gravitational interaction between the LMC-SMC system and the Milky Way are now reasonably successful at explaining the observed morphology of the stream, the high approach velocities near its end, and the existence of a counter-stream on the other side of the Clouds (e.g. Gardiner & Noguchi 1996). According to such models, the stream and counter-stream represent a tidal tail and bridge drawn from the outer gas disk of the SMC during a close passage to the Milky Way about 1-1.5 Gyr ago. The prediction is that the LMC-SMC binary will soon break up and the more massive LMC will be the first to merge with the Milky Way in about 7-8 Gyr (Lin et al. 1995).
The LMC's mass is about 4% of that of the Milky Way, and its visual luminosity twice that of the entire halo. Hence, this future accretion will be a major event, at least an order of magnitude more massive and spectacular than the ongoing Sgr Dwarf accretion. Our descendants can expect significant halo growth, induced star formation, and probably also a thickening of the present thin disk of the Milky Way.
The main message from the above evidence is that - even though most
accretions in galaxies outside the Local Group are difficult to
detect - they must have occurred primarily early (z
2) and
must have contributed significantly to the growth and perhaps even
morphology of many disk galaxies similar to ours and M 31.