Does Dark matter exists ?

We hear everyday "dark matter" this, "dark energy" that...but what are those things, are they figment of our imagination, just nice explanatory "device" or a real thing ? I'll vouch for the second one for now. This does not mean there is no kernel of true or in fact the whole true, but I still don't feel 99% sure ;). It does not matter what the final picture will be, I think it is a good idea to embrace what we know and build some understanding so we can poke further in the matter of "dark matter".
The "dark matter" term was invented to explain some anomalies which seemed to imply that the Universal law of gravitation seems to predict motions in the Solar system fairly well, but once we scale up to the size of Galaxies it worked no more.
Before the middle of 19th century Solar system was composed of 7 planets, Uranus beign the last one. But there was something ikky because Uranus seemed to perturbe on its orbit, so the question was do we fiddle with the Newtown graviational law or decide there is nothing wrong with it but instead there is some yet undiscovered planet influencing Uranus. Scientists decided that there is another planet and in 1846 Netuptune was observed based on the calculations.
Neptune was the "dark matter" of the 19th century.
My point of telling you this little story is to show you that we have been here already i.e. questioning the correctnest of the graviatational law.


We have to first start with some understanding of the law of gravitation and what are its implication for the motion of one object around other.
Let's take for example a planet orbiting a star. So we have a planet(`M_p`) orbiting around a star (`M_s`), like the image shown on the right. The planet is moving with velocity `v` and acceleration `a` and the `r` is the distance between the star and the planet.
For the inquiring minds in circular motion acceleration always points toward the center
In real life the planets rotate along ellipse, but as simplification we will pretend that the planet goes around in a circle.
We would make a use of (1) derived from law of Gravitation i.e. `v^2 = (G*M_s)/r` (look at the Kepler law derivation below, if you are interested). What this relation `v ~ 1/r` is telling us is that the further away the planet is the slower the velocity have to be. And yes that seems to be the case with our Solar system, if we do the calculations they tend to very well match observation. Look at the table and the graph below, further away slower the speed.
I didn't saw this before I made the graph, but isn't interesting that Saturn velocity and distance (in A.U.) almost coincide. A.U. btw is astronomical unit and is equal the distance between the Sun and the Earth i.e ~149 mln km. The other interesting coincidence is that the Pluto velocity is almost 1/10'th of Mercury ;) and of course the distance is 1/100th.

  Kepler law derivationclick to toggle>  [+]
In general we know that the force acting on the object in our case planet from the Newton second law is : `F = m*a` but we know also that the gravitational force that act on planet is : `F = G*(M_s*m_p)/r^2` `M_s` - mass of the star `M_p` - mass of the planet so : `F => m_p*a_p = G*(M_s*m_p)/r^2` when we divide both sides to `m_p`, we get : `a_p = (G*M_s)/r^2` We also know that the acceleration for uniform circular motion is : `a = v^2/r` so we do again the same trick : `a_p => v^2/r = (G*M_s)/r^2` which gives us a solution for the velocity : (1) `v^2 = (G*M_s)/r` but we know that T (the period for circular motion) is equal to : `T = (2*pi*r)/v` then we rise to second power : `T^2 = (4*pi^2*r^2)/v^2` so we substitute reciprocial `v^2` using (1): `T^2 = (4*pi^2*r^2)*(r/(G*M_s))` so get finally the Kepler 3rd law : `G*M_s*T^2 = 4*pi^2*r^3` G - gravitational consant `6.67 * 10^11` M - star mass. T - the period it takes the planet to rotate around the star r - the distance between the two bodies
I wrote the law in this way because it is easier to remember...look carefully GMT like Greenwich Mean Time ;)

planet distance | mln km distance | A.U. diameter | km velocity | km/s period | years
Mercury57.9100.3874,80047.90.241
Venus108.2000.72312,100350.615
Earth149.5901.00012,75029.81
Mars227.9401.5246,80024.11.88
Jupiter778.5405.203142,80013.111.9
Saturn1,433.4009.523120,6609.629.5
Uranus2,876.67019.20851,8006.884
Neptune4,503.40030.08749,5005.4165
Pluto5,874.00039.7463,3004.7248

So far so good, let's pretend now that the planet can orbit inside the star. I know it can't but just bear with me for the sake of argument, we will need this analogy for the Galaxy primer. The image on the right shows how such thing may look like.
What is different in this case ? The difference is that only the part of the mass of the star which is enclosed inside the orbit of the planet will have gravitational influence on the velocity of the planet, the outside-mass part will cancel out and will have no effect, so if we use formula (1) again the conclusion will be that the velocity of the planet should tend to be lower as the orbit get closer and closer toward the center. Look at the graph above again, do you see the downward slope from Mercury toward the origin of the graph.
What I want to point out is that the more mass there is inside the orbit of an object the faster it will travel and vice versa.
That is nice... law of Gravitation and Kepler third law are very cool tools, just measuring star velocities allows us to calculate the period of rotation, mass of the objects and so on...

The surprise came when astronomers started to measure the speeds of stars in different galaxies. The planet-relation-curve we just saw was not obeyed, instead the velocities of the stars were for all intends and purposes the same across the expanse of a galaxy (except those close to the center). (Look at the diagram on the right, case A: solar system, case B: galaxy).
So the question was then how to explain all this. One way was to assume that law of Gravitation was not correct on large distances or may be modify it in some way to accommodate for the observed behaviour (MOND theory tries to do exactly that). But scientists were reluctant to that, the law worked so well in everyday classical calculation and also on solar system scale that it would be a total waste. The idea is to simplify the laws of nature, not to make them more complicated if we can. Right ?
What to do ? what to do ?
It was consequence from the Einstein General theory of relativity (not special relativity), which predicted that under big gravitational influence even light will bend and exhibit effect similar to the well known optical lensing. And so the phenomena was called gravitational lensing. The good thing is that such effects are observed by astronomers.
Look at the two images. Left one is in the Abell 2218 galaxy cluster, you can clearly see the lensing.
So if this is true phenomena that happens in the universe, then let's try Einstein proposition. Below is the equation from General relativity that describes the angle of deflection caused by gravitation :

`alpha = (4*G*M)/(c^2*R)`

where : From the relation `alpha ~ M/R` we can deduce that knowing `alpha` we can calculate the mass of a galaxy or galaxy cluster. And when the astronomers did the calculations the mass were many times bigger than the one calculated by the normal methods of counting the average numbers of stars and weighting them or via any other methods...
I'm interested to find details how such calculations are done, so when I have something I will update this document
We can't expect the mass to be concentrated at the center of the galaxy, because if that is the case then the furthest stars speeds will fall-off as the stars are further from the center like in our star-planet example. Remember what we deduced the more mass there is inside the orbit of an object the faster it will have to orbit around. But we said that the observed mass inside the orbits did not account for the observed velocities. So the idea then is that most of the mass should spread into some sort of invisible hallo's (dark matter) trough-out and outside of the visible part of the galaxy. Just to clarify in the case of perfect circle and homogenous density (the one we talked above) the gravitational influence will cancel out, but in real world with the galaxies we observe of course that is not the case.
The distorted image of the distant object can appear in three possible ways depending on the shape of the gravitational lens:
How the astronomers can be sure that the lensed object is the same ? - By measuring the spectrum and twisted shapes of the objects behind.
By measuring the angle of bending, astronomers can calculate the mass causing the gravitational lens effect (the greater the bend, the more massive the lens). Using this method, astronomers have confirmed that galaxies indeed have high masses exceeding those measured by luminous matter. One important thing to mention is that the reason 'dark matter' is invisible is because it does not interact electro-magnetically with normal matter (protons, neutrons, electrons) and light (photons), but has a mass i.e. interact gravitationally. And as we know gravitation is `10^36` weaker than electro-magnetism, so you need alot of mass to have visible effect.

Summation

So the lesson of the story is that the idea of 'dark matter' preserve the consistency of the Universal law of gravitation.
There is also other theories that try to explain all this, afaik not with such a great success as 'dark matter' idea : It seems the consensus is on unobserved yet type of particle which is massive but does not interact very well with normal matter.