So, I didn't get a chance to give, a lot 
of detail about elliptic galaxies or the 
patterns of, mass versus rotation and 
spiral galaxies, we'll get to some of 
that maybe later. 
But, I want to tell you what we know 
about galactic evolution, because it'll 
give us a framework for thinking about 
this in the same way that the history of 
the solar system gave us a framework 
thinking of the solar system. 
galaxies, it turns out, evolve in a very 
different way than stars, stars 
essentially unless they're in a very 
close binary, evolve pretty much 
independently. 
A star, evolves, on it's own evolutionary 
track. 
Galaxies are far more social creatures 
remember that most galaxies are found in 
clusters, many stars are found in 
clusters, but in a cluster of galaxies 
the fraction of the volume taken up by 
the galaxies is much larger. 
Remember that if you count the halos 
Andromeda and the Milky Way essentially 
are touching. 
Even if you don't count the galaxies 
compute the ratio of size to distance 
between say, Andromeda and the Milky Way 
and size to distance between say the sun 
and Proxima Centauri, and you will get 
the idea that while interstellar space is 
vast and empty, intergalactic space is 
far less vast especially in the context 
of a cluster. 
And so interactions between galaxies, 
therefore are going to be far more 
important in the evolution of a single 
galaxy than is the case in stars. 
Now, I talked about spiral galaxies we 
find as part of our hints to 
understanding how galaxies evolve That 
elliptical galaxies are far more common 
in the interior center of the core of the 
densest clusters in the areas where they 
are densest. 
And this is the region where you expect 
galactic interactions to occur more than 
anywhere else, because the galaxies are 
packed tight near the center. 
And galactic interactions, it turns out, 
can destabilize the disc structure. 
This is sort of a reasonable thing. 
imagine remember that the disc maintains 
itself because of conservation of angular 
momentum. 
So you have a bunch of stars orbiting in 
one particular direction, but then they 
collide with another bunch of stars with 
a different direction of angular 
momentum. 
Those collisions will lead to a combinium 
momentum, but the flatten disk structure 
will probably not survive. 
And the result of collisions is likely to 
be since you have collisions with I mean 
collisions of stars are gravitational 
collisions. 
They're scattering events, but the result 
will be that stars will go off in 
with varying orbital parameters and the 
net result is sort of this elliptical 
galaxy where stars orbit with all kinds 
of inclinations. 
And so if you believe that collisions and 
mergers happen in a cluster you would 
expect ellipticals to survive in clusters 
and that is indeed what, in the centers 
of clusters, that is indeed what we find. 
There's another difference between galaxy 
clusters and star clusters or galaxies as 
large clusters of stars, which is that in 
between stars, the interstellar medium is 
there but the interstellar medium, 
remember gas and dust per whole 
interstellar medium in the Milky Way 
comprises the tiny fraction of the mass 
of the Milky Way unless you count the 
extended halo in which case may be it 
counts for about half the mass of the 
Milky Way excluding dark matter. 
In the case of galaxies we find this is 
very different. 
The typical galaxy cluster, galactic 
cluster is full of hot gas whereby hot, 
again I mean million kelvin gasses and 
the mass of this hot gas usually exceeds 
the sum of the masses of all of the 
galaxies. 
So, a majority of the, again excluding 
dark matter, a majority of the mass of a 
galactic cluster is actually spread out 
in between the galaxies in the form of 
intergalactic medium and again we think 
of interactions as the reason for this in 
inter, intergalactic interactions, tidal 
forces can strip a small fraction of a 
galaxy's gas away and ejected out into 
the interstellar medium. 
And then, if we have recurring 
interactions, then over time some 
significant quantity of gas will be 
transferred to the intsteller medium. 
The net result is the observation that in 
galactic clusters, most of the mass is in 
between the galaxies and not in the 
galaxies. 
So what do interactions between galaxies 
look like? Well, galaxies colliding is 
certainly not like cars colliding because 
galaxies are sort of for all that they 
are massive they are extended objects. 
And the main interaction between two 
galaxies is gravitational. 
So, the typical interaction at slow 
relative velocities is a very stately 
thing, it's a gravitational interaction. 
and a typical case to study is sort of 
the motion of a dwarf galaxy through the 
plain of the Milky Way or a globular 
cluster on it's orbit through the plain 
of the Milky Way or of a globular cluster 
from one galaxy through the plain of 
another galaxy, and, it's a little bit 
tricky to make the calculation, but you 
can sort of imagine but if you have a 
bunch of stars scattered around, and 
along comes a globular cluster, pushing 
through moving this way, then in its 
wake, what is going to happen is that the 
stars are going to gravitationally 
attracted to this visitor and so there 
will be a motion of stars towards and 
after it takes a while for these stars to 
move and after the cluster of pass, has 
pass, there will be a wake. 
A region of increased mass density from 
where it is 
attracted star. 
So, an object moving through a field of 
stars always is dragging behind it 
increased mass density which is 
decelerating it. 
And it in turn is of course accelerating 
these stars so momentum is being 
transferred from the moving object to the 
galaxy relative to collection of stars 
relative to which it is moving. 
This is called dynamical friction and 
because it's second order of 
gravitational effect, it's not surprising 
that it increases like GM^2. 
Why GM^2 because the first GM is the 
magnitude of the wake, that the star 
creates and the second GM is the 
magnitude of the force that that wake 
exercises on the star. 
So, 2 GM row is the density of the medium 
of course the denser the medium exerts 
more friction and then this 1 / v^2 can 
be argued either from dimensional grounds 
v being the speed with which this thing 
is moving relative to the stars around it 
or because the, the basic idea is, the 
faster you're moving through the less 
time there is for this wake to establish 
itself and exert a force. 
The net result is that a visitor will be 
slowed down so a globular cluster, as it 
passes through the disk, through the disk 
population, will be slowed down and 
slowly spiral into a smaller and smaller 
orbit and eventually its interaction with 
the discs will become so frequent. 
That tidal and forces will then break it 
up and the globular cluster will not live 
forever. 
It will eventually merge with a disc. 
And, this will happen faster for massive 
clusters. 
so if you started globular clusters with 
all kinds of masses, the smallest one 
would survive the longest. 
Another, sort of, observation to file 
away this is what a slow collison between 
galaxies look like. 
there are faster collisions where the 
relative speeds are bigger and, 
essentially, the 2 galaxies pass through 
each other without time for wakes to get 
organized. 
Without the stars having time to react, v 
is much larger. 
The stars still rearrange each other but 
they rearrange with a delay, this 
liberates gravitational energy that was 
converted while the galaxies were 
accelerating towards each other. 
Some of that is converted into kinetic 
energies of the stars, the result can be 
the explosion. 
A large amount, of kinetic energy of the 
stars in one galaxy and in dramatic 
cases, like the the, something cartwheel 
nebula. 
[LAUGH] Like the cartwheel galaxy, this 
has caused the emission of a lot of gas 
in a circular ring around the galaxy, and 
the stars that we see here, these blue 
stars are of course nothing to do With a. 
They were not ejected from the galaxy. 
The gas was ejected. 
The gas is plowing into the intergalactic 
medium. 
That's creating a high density shock wave 
and that's generating star creation so 
this beautifully decorative ring around 
the galaxy is marking the place where the 
shock wave is progressing through the 
intergalactic medium and this may well 
have resulted From a collision with one 
of these two objects here over on the 
right. 
another case is this elliptical galaxy, 
which has what is called a polar ring, 
likely the result similarly of a 
collision. 
Perhaps the collider was destroyed and 
merged with the elliptical but the 
resulting energy ejected this ring. 
Now, these rings are very useful because 
they do not orbit in the plane of a 
galaxy, they are influenced by the 
gravity of whatever is around there. 
This gives us an opportunity to measure 
by studying their orbit or parameters, 
the dark matter distribution 
perpendicular to the plane. 
And somebody might have argued based on 
the orbit of the sun and the stars and 
the clouds and the Milky Way that dark 
matter could be confined to the disk of 
the Milky Way. 
the reason we are absolutely certain it's 
a spherical distribution is the, that, 
the orbits of these objects, which orbit 
way outside the plane of the galaxy agree 
with the predictions of the dark matter, 
cold dark matter model. 
when the collisions are more delicate or 
in any collision between galaxies, 
remember galaxies are in free fall. 
The gravitational attraction or the 
gravitational interaction of one with the 
other, is a tidal interaction and this 
can lead to these beautiful objects. 
On the left, of course M51 the whirlpool 
galaxy interacting with not quite a 
collision, but obviously an interaction 
with the smaller galaxy in the upper 
right, 
We see that one of the spiral arms has 
grown and is extending towrads this 
galaxy and slightly outside the frame the 
other spiral arm extends very far to the 
left. 
These are both tidal effects gas or 
material and stars are being attracted on 
the one hand and attracted away from the 
galaxy on the other hand in the same way 
that we get high tide on two sides of the 
Earth. 
These are the tidal effects of the 
passing galaxy. 
And an even more brilliant example of 
these sort of tidal tails, are these 
antennae galaxies, as they are called 
here. 
And then a pair of galaxies in collision. 
And we see the effects of tidal forces in 
these large tendrils of gas that have 
been emitted by both of the galaxies. 
in another recent observation of the same 
effect, NGC68672 has recently been, been 
measured to be the largest spiral galaxy 
known. 
the scale here is 100,000 light years so 
this thing is about 2 or 3 times the size 
of the Milky Way, but that's kind of 
cheating. 
One of the reasons these spiral arms are 
so extended is because it seems that this 
has undergone a collision, perhaps over 
here, in the upper left, and so it has 
got these two tightly extended spiral 
arms coming out on two sides. 
And the most exciting thing here is that 
there is some structure in this upper 
left its tidal tail and it seems as 
though some of the tail has become 
sufficiently cold and dense to collapse 
and there's the region that is in circle 
there is a region where astronomers 
suspect that a actual globular cluster or 
a dwarf galaxy might be forming. 
So this is the weird case where rather 
than destroying galaxies and merging 
them, here a collision is actually 
creating a new galaxy, a new wrinkle on 
our theories of galaxy interactions, 
something modelers will indubitably 
incorporate as time goes on, if indeed 
this thing is verified. 
Now an interesting question you might 
ask, is, well. 
So, if what you are telling me is that 
very often, two galaxies, will in fact, 
merge, because they will be slowed down, 
ripped apart by each other's tidal 
effects, and in fact, end up as one 
merged object. 
Well that's very nice for the stars, they 
can orbit whatever they want and the gas. 
Each of these galaxies would have had a 
super massive black hole in its center. 
What do those do? Well that's an 
interesting question. 
And what you would guess, of course, is 
that the two black holes, being so 
massive, would very rapidly find 
themselves in the core, of the combined 
object. 
So you could find yourself, in the 
interesting situation, like NGC6240 which 
is a galaxy whose shape clearly suggest, 
a past in which some interaction happened 
and a x-ray image of the center, shows 2, 
x-ray sources, two black holes. 
This has a binary super massive black 
hole and they are seperated by, this 
doesn't say, but what it meant to say, is 
that these things are separated by an 
order of 3000 light years or a kiloparsec 
an eighth of the distance from the sun to 
the center of the galaxy is the distance 
between these two super massive black 
holes. 
Those are some very interesting highly 
relativistic, orbital dynamics, and over 
time of course friction with 
the dynamical friction, with the medium 
as well as gravitational radiation will 
cause these things to eventually over 
hundreds of millions of years or billions 
of years to move closer to each other, 
eventually merge with each other, and 
there you will get a blaze of 
gravitational radiation, that perhaps 
someone will be able to detect at that, 
to detect at that point. 
So, based on all these observations of 
galactic interactions we can try to 
understand, a model of galaxy formation. 
And the original model of galaxy 
formation was basically the solar nebula 
at large. 
So, remember, the solar nebula formed 
from a cloud that collapsed, a proto 
star, that story. 
Imagine the same thing on a galactic 
scale so this is the this is the cartoon 
version of the ELS model of galaxy 
formation. 
You start with a protogalactic cloud. 
The protogalactic cloud collapses in the 
center where the density gets highest. 
stars starts forming long before the 
densities of the outskirts are high 
enough. 
Eventually of course angular momentum 
will flatten the cloud just as it did the 
protosteller disc but, in the center 
perhaps enough stars formed in clusters 
managed to collapse so there was 
fragmentation and smaller sub clouds 
became critical and collapsed before the 
flattening so then you'd find the old 
globular cluster stars orbiting in orbits 
that are not necessarily in the plane of 
the disk, and then eventually the disk 
flattens out and you get the issue of 
the, the, the shape of the galaxy as we 
know it. 
This is a very reasonable model, and it 
probably is part of the truth. 
There are some issues with it. 
One is, it does, it explains why you 
find, halo stars and halo clusters 
orbiting in inclined orbits, but not why 
so many of them are in retrograde orbits. 
Orbits considering that as the cloud 
collapsed angle of momentum reservation 
would have sped up whatever rotation it 
has. 
It also turns out that this collapse 
occurs very rapidly and it does not have, 
it does not leave for a cloud the size of 
The Milky Way this is the Kelvin Helmoltz 
gravitational collapse time scale. 
It's a rapid timescale and even in the 
case of a cloud the size of the Milky Way 
you don't have 2 billion years in which 
to create halo stars. 
Remember the ages of globular clusters 
were between 11 and 13 billion years. 
This is 1 of the key pieces of evidence 
that this model is no sufficient. 
In addition there's this correlation 
between distance, orbital radius and 
modelocity in globular clusters 
that says that the distant clusters are 
younger. 
This does not match, in this theory you'd 
expect the clusters to be formed in the 
center to be the. 
The oldest ones and, and also the fact 
that the disk components, the thin and 
thick disk differ in age, in this case, 
in this model everything happens fast and 
expect stellar creation to start in the 
disk as soon as densities got high 
enough. 
And there is not, again a billion, 
billions of years of separation that can 
explain all of the phenomena that we see. 
So this is not a sufficient model. 
Another sort of opposing model is what's 
called the heirarchical merger model, in 
this case a galaxy forms more the way we 
said planets formed. 
Form, so, there is a collapse and what 
collapses are basically globular clusters 
or dwarf galaxies, protogalactic 
fragments and they can a mass from a 
million solar masses to maybe bigger than 
a galaxy. 
But in the usual way that fragmentation 
works we know that the most common will 
be the smallest ones. 
So you'll have a lot of, globular cluster 
size. 
Protogalactic fragments, with a mass of 
about a million solar masses. 
And then, as these orbit each other, and 
interact with each other, dynamical 
friction slows them down, and, 
eventually, they are brought into this, 
big spherical, mass distribution. 
And, there, density becomes high enough 
that they start forming stars and 
globular clusters in the center, and 
because these things have formed stars 
individually then in the center of each 
cloud you start star formation then this 
explains why you get different chemical 
histories because different clusters 
independently initiated different 
fragments independently initiated star 
formation and then these things are now 
beginning star formation at the same time 
that they're interacting with each other. 
Collisions and tidal forces will break 
some of them. 
Which ones will break most likely? Well 
the ones with, which experience the most 
dynamical friction namely the mass, most 
massive one. 
So the more massive ones are more likely 
to be slowed down and disrupted because 
of dynamical friction those that have 
been disrupted will produce the halo 
field stars and, some of the others the 
stars and the gas will have been blown 
away, but the globular clusters as more 
massive, will have survived. 
So we'll have bare globular clusters and 
we'll still have some dwarf galaxies. 
Now, in this model about 90% of all the 
globular clusters formed, especially the 
massive ones, as I've said, are going to 
be destroyed. 
The globular clusters we see in the halo 
today will be the 10% that survived and 
with an emphasis on the less massive ones 
which is why we typically find globular 
clusters with masses up to a million 
solar masses. 
And then very early in the dense center 
as this these things are starting to 
merge. 
You form a bulge, so in this case star 
formation proceeds from the inside out, 
in this hierarchical merger model and so 
you'd expect very old stars in the bulge. 
Explaining why there's new stars in the 
bulge, is a detail of the model I won't 
really have time to get into. 
Of course in principal both models 
probably have a role to play. 
We do not have I should say a 
satisfactory theory of galaxy formation, 
this is an area of active research. 
But I think the, the consensus is that, 
the Hierarchical Merger Model in a 
slightly more elaborate version than what 
I've described is 
the more likely, it's more likely to be 
closer to what it is that really goes on. 
Can we see the evidence of this? So for 
stars, we used clusters to test our 
theories of stellar evolution. 
How do you test your theories of galactic 
evolution? we do not have a long main 
sequence understanding so we don't, can't 
time galaxies in the way we timed stars, 
but we can look for very young 
galaxies.Where would you find very young 
galaxies? You'd find very young galaxies 
by looking very, very far so this is the 
very famous 
Hubble extreme dark field. 
basically the Hubble Telescope was aimed 
at a region, the constellation Formax. 
This shows just a region 3 by 3 arc 
minutes in the sky which as far as anyone 
knew when they aimed at it, was 
completely empty of anything. 
There was no known objects in that region 
and then they took a very long exposure 
and in this field they found about 10,000 
formerly unknown galaxies. 
They looked at a region where there was 
nothing in order to estimate, in order to 
be able to focus on very dim and 
therefore very distant objects. 
These galaxies are billions of light 
years away. 
Aha, billions of light years away tells 
us that we are seeing them billions of 
years ago, we'll get into that soon and 
this means that the galaxies we see here 
are likely to be younger than the ones we 
see a million light years away like 
Andromeda or the Milky way which we see 
right now. 
And so, in this Hubble Deep Field indeed, 
we see that the galaxies seem to be bluer 
than the galaxies we see around us, and 
their shapes are far more irregular. 
It is a possibility that's under study 
that what we're seeing here is, in some 
of these cases the proto-galactic 
fragment that will eventually form the 
galaxies that we see today. 
As I said, galaxy formation is still a 
very active field of study and the 
understanding we have of the dynamics of 
the qualitative level we can handle, I 
think is all we're going to be able to 
achieve today.