As far as my understanding goes, dark matter is nothing but an amount of gravitational force, from yet unresolved/undiscovered source(s), needed to explain some observed attributes of our universe. Then, is it called "Matter" only because of gravity? Because other properties of matter like occupation of some space, mass etc is not found in this case. And, can it be some new kind of force other than gravity?

  • 20
    $\begingroup$ xkcd.com/1758 $\endgroup$
    – AGML
    Nov 18, 2016 at 6:36
  • 4
    $\begingroup$ I guess I have asked for definition of Matter in some weird way. $\endgroup$
    – Gulshan
    Nov 18, 2016 at 8:26
  • 3
    $\begingroup$ Does it really matter what the unknown is called? Does it change the observation or interpretation of it? Call it dark energy if you like. $\endgroup$ Nov 18, 2016 at 9:10
  • 4
    $\begingroup$ Matter isn't a well defined scientific definition, but I think your instincts are correct in questioning whether dark matter is really matter, that said, matter doesn't have a clear definition so the question lacks precision. See related: physics.stackexchange.com/questions/192564/… Dark matter also needs to be discovered and to some degree, examined before it's defined. See also (on matter): physics.about.com/od/glossary/g/Matter.htm physicsforidiots.com/physics/particles-and-forces/… $\endgroup$
    – userLTK
    Nov 18, 2016 at 9:22
  • 11
    $\begingroup$ @Trilarion lets not call dark matter "dark energy". That would be confusing. Call it invisible mass, call it unknown mass, call it transparent galaxy glue, call it that god-damn stuff (little joke) but don't call it dark energy. Dark energy is the name/cool catch-phrase for something completely different. $\endgroup$
    – userLTK
    Nov 18, 2016 at 9:27

3 Answers 3


As the universe expands the density of matter goes down. For example if the volume of some specific region of the universe doubles then the density of the matter in that region halves. More precisely, suppose we take the scale factor of the universe, $a(t)$, to be unity right now and we take the current average density to be $\rho_0$, then at a time $t$ the density will be:

$$ \rho(t) = \frac{\rho_0}{a^3(t)} $$

This should be intuitively obvious. Suppose the universe doubles in size, i.e. the scale factor increases from $1$ to $2$, then the volume increases by a factor of $2^3 = 8$ so the density falls to $\rho_0/8$.

But ...

Even though this seems intuitively obvious it is only true for matter, and strictly speaking it's only true for pressureless matter (though to a good approximation the matter in the universe is pressureless). For photons the density as a function of time is given by:

$$ \rho(t)_\text{photon} = \frac{\rho_{0\text{photon}}}{a^4(t)} $$

Note that this has an $a^4$ dependence not $a^3$. And for dark energy the density is independent of time:

$$ \rho(t)_\text{de} = \rho_{0\text{de}} $$

(assuming that dark energy behaves like a cosmological constant).

The point of all this is that when we say matter we means anything that scales like matter as the universe expands. So the phrase dark matter means something we can't see that scales as $1/a^3$. We don't know what dark matter is made of, but we do know that whatever it is it has to behave like ordinary matter as the universe expands.

  • 5
    $\begingroup$ Would you kindly clarify "pressureless matter". Why it is pressureless. $\endgroup$
    – J. Manuel
    Nov 18, 2016 at 14:28
  • $\begingroup$ @J.Manuel just to clarify. Are you asking why we can treat matter as pressure less at the moment, or are you asking what would happen if matter wasn't pressure less? $\endgroup$ Nov 18, 2016 at 18:35
  • $\begingroup$ @JohnRenie Anything. I just want to be clarified on what is pressureless matter. $\endgroup$
    – J. Manuel
    Nov 19, 2016 at 0:31
  • 2
    $\begingroup$ @RBarryYoung briefly, as space expands photons both dilute and redshift and both effects lower the photon energy. So the photon energy density goes down faster that $1/a^3$. To go into this in more detail would need you to ask a new question on it. $\endgroup$ Nov 20, 2016 at 6:37
  • 2
    $\begingroup$ Nitpick: It's not completely accurate to say that "matter" is something that scales as $a^{-3}$. Physicists often distinguish between "hot dark matter" (light particles that travel near the speed of light, even at the current temperature of 3 K) and "cold dark matter" (heavier particles that travel non-relativistically at a temperature of 3 K.) "Hot dark matter" scales as $a^{-4}$ for the reasons you described above. (Hot dark matter is also pretty much ruled out by experimental constraints, so this distinction is a bit academic.) $\endgroup$ Nov 20, 2016 at 13:36

Too long for a comment, but on the nomenclature, Fritz Zwicky's observations of other galaxies made it apparent to him that galaxies had to have much more mass than could be seen. The visible stars added together were about 100 times too light to explain their stellar orbital velocity. There had to be a lot of unseen mass to hold the galaxies together. He called this unseen mass "dunkle materie", or dark matter.


When Zwicky made this claim, the study of other galaxies was in its infancy and and it wouldn't have been unreasonable at the time to assume that Zwicky's "dark matter" was clouds of dust and gas and maybe dark/extinguished stars, some asteroids here and there. That dark matter was some new kind of undiscovered "particle" wasn't generally agreed to until the 1980s but and by then, it already had its catchy name in place.

Unexplained mass is a more accurate term. Dark Matter is a cooler name. Cool names tend to stick, accurate or not.

  • 2
    $\begingroup$ Perhaps simply Einsteins equations are not well adjusted? ;-) In electrodynamics there is something called self induction. To proof such phenomenon for gravitational processes is not easy knowing only the situation on earth. $\endgroup$ Nov 18, 2016 at 8:25
  • 5
    $\begingroup$ I would say "dark matter" is also accurate, and more accurate than "unexplained mass" would be. It is dark, literally, as it does not interact with light. We have detected it interacts with and causes gravity, and forms something like clouds (for example the famous bullet cluster image), so it's also hard to claim it is something which can't be called "matter". $\endgroup$
    – hyde
    Nov 18, 2016 at 10:30
  • 5
    $\begingroup$ @hyde I disagree that dark matter is "literally dark". In ordinary English, "dark" means that there's not much light around, and a substance that doesn't interact much with light is described as "transparent". $\endgroup$ Nov 18, 2016 at 13:20
  • 2
    $\begingroup$ @HolgerFiedler : Recalling (perhaps fuzzily) from a homework assignment two decades ago: gravitational induction is > 30 orders of magnitude too small to explain galactic rotation curves. $\endgroup$ Nov 18, 2016 at 16:52
  • 2
    $\begingroup$ A detail, but Zwicky studied motion in galaxy clusters. It was Vera Rubin who studied galaxy rotation curves. You have conflated the two. $\endgroup$ Nov 18, 2016 at 22:43

When astronomers look at the orbital speeds of stars in galaxies (using the Doppler Shift), those speeds imply there is far more gravity and so far more mass in those galaxies than can be accounted for by ordinary matter.

The best measurements show that there is about 5 to 6 times as much dark matter as ordinary matter. The actual amount of dark matter in each galaxy varies. One galaxy, called Dragonfly 44, is about 99.9% dark matter.

Astronomers can also measure how much mass is in a region by using 'gravitational lensing'. Gravity bends light, and so through careful analysis, astronomers can calculate how much gravity and so how much mass is in a particular region.

One of the critical observations using gravitational lensing is of a group of galaxies called the Bullet Cluster.

In the Bullet Cluster, the collision of galaxies has caused the self interacting gas to move differently than the barely interacting stars. (Stars have such a low number density, that they almost never crash into each other, or even get very close when their galaxies collide.) There is a region of space within this cluster where the gravitational lensing shows there is mass, but there is no ordinary matter to account for it.

Now to the part of your question where you ask about the definition of matter. Dark Matter has mass, otherwise it would move at the speed of light and not stick around in a galaxy. Ordinary matter takes up space because of the electromagnetic interactions and Pauli Exclusion. Dark matter is not expected to have electromagnetic interactions with ordinary matter and since dark matter isn't made of ordinary electrons Pauli Exclusion wouldn't apply.

There are some hints that dark matter should be self interacting. That even though it only interacts gravitationally with ordinary matter, it might have an exclusive interaction with itself.

  • $\begingroup$ Dark matter may well be fermionic. Why do you assume otherwise? $\endgroup$
    – ProfRob
    Nov 20, 2016 at 13:57
  • $\begingroup$ Well it is your default assumption. The dark matter could be scalar, fermionic or of a vector boson nature. Obvious fermionic examples would be sterile neutrinos or neutralinos. The PEP applies to fermions, not just electrons as you have described. $\endgroup$
    – ProfRob
    Nov 20, 2016 at 16:20
  • $\begingroup$ @Rob Jeffries, (minor edits) I think that's probably a good default assumption. But without data from either particle accelerators (no one has reported a signal where energy goes missing, and so dark matter is produced, and then detected by its absence) or any of the direct detection WIMP experiments, it's hard to get a handle on what exactly dark matter might be. I bet the answer will surprise just about everyone. $\endgroup$
    – David Elm
    Nov 20, 2016 at 16:28

Not the answer you're looking for? Browse other questions tagged or ask your own question.