Until the early 1970's, the exact nature of the tail- and bridge-like
features observed in many peculiar galaxies was the cause of much
debate. Indeed, analysis of the discussions at the 15th IAU Symposium,
``Problems of Extragalactic Research'' (1961) illustrates the then uncertain
nature of the problem. Vorontsov-Vel'yaminov, well acquainted with peculiars as
compiler of The Atlas and Catalogue of Interacting Galaxies,
called for new physics to explain the bridges and
tails he observed in his subjects; Thomas Gold advocated magnetic
forces, although he noted that they would be unable to restrain matter
once it was in stellar form. Fritz Zwicky confirmed that the bridges
are dynamically young but contain old stars, while the Lindblads
postulated that ``streams of matter'' drawn by tidal forces may be responsible.
Thus, whilst being a subject that stimulated much debate, there was
little agreement over the cause of these bridges and tails.
At this time, P'fleiderer and Siedentopf were examining whether spiral
patterns in galaxies might be excited by encounters with other passing
galaxies. Although their results indicated that insufficient time had
passed since galaxies were formed for the numbers of spirals observed
to have been generated by such encounters, their research - using
computers to model the encounters - would have important consequences
for the problem of bridges and tails.
In 1972, at a time when much research was directed towards
understanding how these characteristic features of interacting
galaxies could have formed, two brothers, Alar and Juri Toomre
published a paper that would revolutionise our understanding of how
interacting galaxies form bridges and tails. In Galactic Bridges and Tails, the Toomres presented
the a basic recipe for bridge and tail building based on the tidal
interaction of two galaxies. They examined 4 well known systems with
well defined bridges and tails. Although highly idealised - a galaxy was modeled as a
central mass surrounded a disk of massless test particles - the model
encounters were able to reproduce with varying degrees of success
the characteristic features of the real systems. The results for the
``Antennae'', one of the closest interacting systems sporting a
symmetric pair of crossed tails, gave them particular pleasure.
Since then, Barnes (1988) and Bothun, Mihos and Richstone (1993) have engaged
the problem of reproducing the characteristic features
``Antennae''. This system has been called the ``Rosetta Stone'' of
interacting galaxies; it would be fair to say that the general feeling
is that if NGC 4038/9 can be successfully modeled, then a
comprehensive theory of this class of galaxies is close at hand.
If such a stage is to be reached, however, then the results of modeled
encounters must match those of observations. In a recent review,
Barnes (1998) outlined four key features of the ``Antennae'' which have thus
far eluded attempts to model them. It should be clear that model
matching, and in parallel, the visualisation of the results of simulations
are of vital importance if a clear understanding of galaxy
interactions is ever to be achieved. In this report, I shall outline
the fruits of my Summer's work in this particular area.
In section 2 I present a brief account of how our understanding of
collisions between galaxies has evolved over the course of this
century. From the earliest postulations of Lundmark and Lindblad to
the ingenuity of Holmberg and the wild theorizing of Zwicky to the
breakthrough computer experiments of the Toomre brothers, the history
of the subject is both colourful and invigorating, and in my opinion, worthy
of some discussion.
Section 3 deals with previous attempts to model NGC 4038/9
numerically; in particular, I consider the 1972 effort of Toomre and
Toomre which relies on a restricted 3 body code, and Barnes' 1988
attempt which utilises a 3 dimensional fully self consistent N-Body code.
I also briefly discuss a recent paper by Mihos, Richstone and Bothun
(1993) on star formation rates in NGC 4038/9, and indicate areas in
which future models must improve.
Section 4 is concerned with model matching process. For
any physical model of the Antennae to be considered plausible, it must
reproduce the observed features of the system. I
discuss the essential elements of any model matching scheme and part
of my Summer's work.
Visualisation of Data, the topic of section 5, reflects the bulk of my
Summer's work; in particular, it deals with the 3 dimensional representation of
our data. After due consideration, the highly interactive 3D visualisation
package Geomview was chosen as the viewing tool. I briefly
discuss the development of software that processes both observational
and simulation data into a form suitable for 3D viewing; the hopes for
future development of this software; and the related topic of
automating the model matching process in an efficient and flexible way.