Astronomers say there is more to the Universe than meets the eye. From the way
that stars and galaxies move through space, it is clear that they are being
tugged by some invisible force. But what is responsible for that force?
THE DUTCH astronomer Jan Oort was one of the first to realise, in the early
1930s, that the only way to explain the behaviour of stars in our immediate
neighbourhood was to imagine that some dark, invisible matter filled the
greater part of space. At the time, astronomers had established that the stars
of the Milky Way Galaxy are each in orbit around a centre quite distant from
the Sun. Our Solar System lies about two-thirds of the way out from the centre
of this swirling system, in the galactic suburbs, so to speak. It is possible
to study in some detail stars in our neighbourhood, and astronomers find that
individually these objects do not move precisely in a single plane, but wobble
up and down as they orbit around the Galaxy.
Advertisement
Dark forces
Dust and gases
OF COURSE, Oort could not watch an individual star moving up and down in this
way. Such changes take thousands or millions of years. But he was able to
conclude from the overall distribution of stars above and below the plane of
the Galaxy, and from measurements of their speeds (determined by the Doppler
effect 鈥� the observed change in wavelength emitted by a star due to its
motion), that the visible stars themselves contribute only a third of the
gravitational force holding the stars in place.
Since then, other astronomers have identified, by radioastronomy and other
techniques, that as great a mass of cold clouds of gas and dust is spread
between the stars as in all the visible stars. But even that, together with
the stars themselves, only accounts for two-thirds of the gravitational mass
required to explain the local dynamics of the Galaxy. There ought to be
considerably more dark matter.
The unseen dark matter can be measured in terms of a number called the
mass-to-light ratio, M/L. This is defined as being 1 for our Sun 鈥� one solar
mass of matter, in the form of a star, produces one solar luminosity of light.
Oort鈥檚 figures tell us that in our neighbourhood M/L is about 3. But the
number is bigger for the Galaxy as a whole.
Spiral galaxies
Flat fried eggs
IN THE 1980s, spectroscopic techniques became sufficiently advanced that
astronomers could measure in detail the rotation of spiral galaxies, systems
like our own Milky Way Galaxy. The visible part of a spiral galaxy consists of
a central bulge of stars surrounded by a thin disk of stellar material 鈥� the
proportions are approximately those of the yolk to the white of a fried egg.
If a galaxy happens to be oriented 鈥渆dge on鈥� to us in the sky, it is possible
to measure the speed at which different parts of the disc rotate by placing a
narrow slit across the image of the disc at different places, and measuring
the Doppler shift in the spectrum at different distances from the central
bulge. Recently, the technique has been extended farther out from the centres
of some galaxies, using the techniques of radio astronomy to measure the
velocity of clouds of hydrogen gas, still part of the disc.
When astronomers plot the velocities of the stars and clouds orbiting the disk
of a distant galaxy at different distances from its nucleus, they obtain a
rotation curve that is usually highly symmetrical. At a distance from the
centre on one side of the galaxy the stars are moving towards us at the same
speed as stars on the other side of the galaxy, at the same distance from the
centre, move away from us. (It is necessary to subtract out from these
measurements the overall red shift caused by the expansion of the Universe.)
This was no surprise. But astronomers were surprised to find that outside the
innermost regions of a spiral galaxy, on either side of the nucleus, the speed
with which the stars move is the same all the way across the disc. In
astronomers鈥� jargon, the rotation curves are extremely flat (see Figure 1).
Bright nucleus
Invisible halo
THIS WAS a surprise as astronomers had assumed that the greatest amount of
mass in a spiral galaxy was concentrated in the bright central nucleus, where
there are many stars. If that were so, then stars farther from the nucleus
should be moving more slowly in their orbits.
This is what occurs with the outer planets of our Solar System (most mass is
concentrated in the Sun, at the centre) which move more slowly in their orbits
than the inner planets.
The simplest way to explain the flatness of the rotation curves is to consider
there is a great deal of dark matter spread around each spiral galaxy in a
huge unseen halo. If this halo is roughly spherical, then as it rotates it
will drag the visible, bright stars around with it in just the way we see. In
other words, most of the mass of a spiral galaxy like our own is not
associated with the bright stars of the nucleus (or even those of the disc),
and the mass-to-light ratio is at least 5.
Moving up from the scale of individual galaxies, the next level of structure
in the Universe is provided by clusters of galaxies, anything from a few
galaxies to many hundreds of galaxies held together in a swarm through mutual
gravitational attraction. The speed with which each galaxy in a cluster is
moving can be inferred from the Doppler effect (once again, the overall
cosmological red shift, in this case for the cluster, has to be subtracted
out), and the amount of mass in each galaxy can be estimated from its
brightness, if we assume that the mass-to-light ratio is about 1.
Swarming clusters
Familiar elements
THE FIRST person to make these sort of studies was the Swiss astronomer Fritz
Zwicky. At about the same time that Oort was discovering evidence of dark
matter close to home in the Universe, Zwicky began to find evidence of dark
matter on a much more impressive scale. If the galaxies in a cluster really
are held together in an association by gravity, while the cluster as a whole
moves through space like a swarm of bees, then the velocities of individual
galaxies in the cluster must be less than the escape velocity from the
cluster. But when Zwicky used the Doppler technique to measure the velocities
of galaxies in one group, the Coma cluster, he found that they were moving far
too rapidly, relative to one another, to be held together by the gravitational
pull of all the stars in all the galaxies of the cluster.
It looked as if the flying galaxies ought to have moved apart, dissolving the
cluster, long ago when the Universe was young. And he found the same thing
when he looked at other clusters 鈥� the galaxies in them were all moving apart
far too quickly to be held together by the gravity of the matter we can see.
For decades, even though all the evidence continued to hint that clusters of
galaxies contain large amounts of dark matter 鈥� with mass-to-light ratios as
high as 300 鈥� few astronomers worried much about this problem. In the 1930s,
the concept of the expanding Universe, and even the fact that it extended far
beyond our Milky Way Galaxy, were new ideas. The possible existence of dark
matter seemed a minor puzzle compared with developing an overall picture of
the origin and evolution of the Universe and, indeed, of the galaxies
themselves. It was only in the 1960s that the big bang model began to be
established as the standard model of the Universe. And only after the big bang
model became established did astronomers pay much attention to the detail of
finding an explanation for the dynamic behaviour of galaxies in clusters.
One of the early triumphs of the theory of the big bang was that it seemed to
explain how much matter there ought to be in the Universe, and the predictions
seemed to match the amount of matter we can see. The famous cosmic microwave
background radiation, which the American astronomers Arno Penzias and Robert
Wilson discovered in 1965, was interpreted as a leftover relic from the
fireball in which the Universe was born, and used to calibrate conditions in
the fireball. With this calibration, the standard model of the big bang
predicted that in the big bang primordial hydrogen should have been processed
into helium in just the right amount to explain why the oldest stars are made
up of about 25 per cent helium and 75 per cent hydrogen.
But the same calculation also limits the overall amount of matter there can be
in the form of hydrogen, helium and the rest of the familiar chemical elements
(the so-called 鈥渂aryonic matter鈥�). In order to match the exact conditions in
which helium was manufactured in the big bang fireball to the abundances of
hydrogen and helium seen in the Universe today, cosmologists also had to
specify the overall density of baryons in the Universe. Assuming that all the
matter in the Universe is made of baryons (the same sort of stuff that we and
everything else on Earth, and all the bright stars are made of), this density
converts into a mass-to-light ratio for the Universe as a whole 鈥� and the
number comes out as less than 100. In other words, at the very outside there
could be as much as 100 times as much matter in the form of clouds of dust and
gas, and so on, as we see in the form of bright stars.
At the beginning of the 1980s, this was starting to cause concern among
astronomers. As telescopes and observing techniques improved, the
observational evidence was mounting that in the case of clusters of galaxies,
M/L is at least 300. But a now well-established and highly successful big bang
theory said that M/L for all the baryons in the Universe must be comfortably
less than 100. A great deal of extra mass seemed to be needed 鈥� and it also
seemed clear that it could not be in the form of baryons.
Then, the big bang theory itself underwent a dramatic transformation with the
advent of an idea dubbed 鈥渋nflation鈥�, which describes the earliest era of the
Universe, which produced the fireball in which hydrogen was processed into
helium. Inflation resolved some remaining puzzles about the big bang model,
and is now part of the standard model itself. But while this removes some of
the old cosmological puzzles, it makes one firm prediction, concerning the
density of the Universe. It says that there should be enough matter in the
Universe to keep all the galaxies and clusters of galaxies in a gravitational
grip, in much the same way that clusters of galaxies are held together by
gravity. If that is so, then there must be so much matter in the Universe that
the overall mass-to-light ratio is about 1000, at least three times greater
even than the figure for clusters of galaxies. There must be at least ten
times, perhaps a hundred times, more matter in the Universe than can be
explained by all the baryonic matter produced in the big bang fireball. What
is it? And where is it? Physicists already knew of one other kind of particle
that might fit the bill. For the purposes of these calculations, electrons are
included with baryonic matter 鈥� the mass of an electron is only half of one
thousandth that of a proton and there are the same number of electrons in the
Universe as there are protons, so they make only a minor contribution to the
density. But there is another kind of particle which exists in vast
quantities. These are the neutrinos, which participate in nuclear reactions
that involve the weak force (see Inside Sciences Nos. 15 and 17). It is a firm
prediction of the standard big bang model that there should be about as many
neutrinos in the Universe as there are photons 鈥� about a billion
(109) times as many neutrinos as there are baryons.
Until the 1980s, neutrinos had traditionally been thought to have exactly zero
mass, like photons. Some particle physics theories require this, but others
allow the possibility that neutrinos might have a small mass. With so many
neutrinos in the Universe, if each one had a mass of a few tens of electron
volts (less than one ten-thousandth of the mass of an electron) they would
together provide all of the dark matter required to hold the Universe
together. (Note, particle physicists conventionally measure masses in electron
volts (eV). The electron volt is, strictly speaking, a unit of energy, but as
Einstein taught us, dividing energy by the square of the speed of light gives
a mass equivalent. The 鈥渃2鈥� term is taken as read when physicists
quote particle masses in eV.)
Experiments to measure directly the masses of neutrinos are difficult and have
so far proved inconclusive, only setting upper limits on the mass. So the
best test of whether the dark matter is in the form of neutrinos is to analyse
the distribution of galaxies in the Universe and determine whether this
matches the pattern that would be produced if neutrinos have mass.
The situation is complicated, however, because there is another possible type
of candidate for the dark matter. Particle physicists seeking to find a
unified theory of physics have suggested that there may be one or more
varieties of particle present in the Universe which have never been detected
in the laboratory. It is intriguing that this suggestion 鈥� that there might be
more stuff in the Universe than we have ever seen 鈥� has come independently of
the astronomers鈥� discovery that there is more to the Universe than meets the
eye. 快猫短视频s who operate on both the largest and smallest scales require
鈥渘ew鈥� forms of matter.
Different variations on the particle physics theme suggest different
candidates for the extra particle(s). However, some would have masses
comparable to that of the proton, but be so reluctant to interact with
everyday matter (except by gravity) that they have not yet been detected.
These hypothetical particles are sometimes called WIMPs, for 鈥渨eakly
interacting massive particles鈥�.
The generic name for such particles is 鈥渃old dark matter鈥� (CDM). The
terminology 鈥渃old鈥� refers to the fact that they have relatively large mass and
therefore emerge from the big bang travelling much more slowly than the speed
of light. By comparison, neutrinos have little mass (if any) and emerge from
the big bang travelling at very high speeds, close to the speed of light. They
are therefore known as 鈥渉ot dark matter鈥� (HDM). Over the past ten years or so,
one of the major challenges for astronomers has been to determine whether the
pattern of galaxies in the sky more closely resembles the pattern associated
with CDM or the pattern associated with HDM.
The key difference is the influence of the two kinds of dark matter in the
early Universe, just after the big bang fireball, when stars and galaxies
started to form. Hot dark matter particles would sweep everything before them,
keeping the Universe smooth and homogeneous until they slowed down and began
to allow irregularities to grow. Because the distribution of matter on smaller
scales would have been smoothed out by then, the first structures to form
would be on the scale of superclusters of galaxies, shaped like huge sheets
and filaments, which broke up to make galaxies and stars 鈥� a 鈥渢op-down鈥�
scenario as the big bits broke up first to make small bits.
Galactic superclusters
Filamentary patterns
IN A Universe dominated by cold dark matter, however, structure begins to form
on smaller scales, soon after the big bang. Clumps of dark matter attract
baryonic matter, like water flowing into a pothole, and structure builds from
the bottom up, with stars and galaxies clumping together to make superclusters
and filaments.
Both theoretical calculations and computer simulations help to indicate what
kind of clumpiness would be seen in a Universe dominated by hot dark matter,
and what kind of clumpiness we would expect in a Universe dominated by cold
dark matter. A Universe dominated by hot neutrinos is predicted to have a
rather simple structure, like the cells of a honeycomb (though not so
regular), in which bright galaxies form only in well-defined sheets and not at
all in the voids. The CDM Universe is more messy and complicated, with a
richer structure that looks more like the real Universe. Sheets and filaments
do form, but they intertwine in a complicated way, and the voids are not
really empty completely.
COBE鈥檚 finds
Next step on
BUT THE simplest versions of the CDM model cannot account for all of the
details of the distribution of galaxies across the sky. Some additional
influence is needed to explain the structure of the real Universe.
Arguments about what the additional influence might be have included the
suggestion that the nature of gravity might have to be modified, or that
ripples produced by gravitational radiation (see Inside Science No. 31) could
have played a part in determining the distribution of baryons in the early
Universe. But the simplest resolution of this puzzle comes from analysis of
the ripples in the cosmic microwave background radiation that NASA鈥檚 Cosmic
Background Explorer (COBE) satellite detected (see Inside Science No. 69).
The pattern of ripples that COBE detected was imprinted on the background
radiation some 300 000 years after the big bang, at the time when the
radiation last interacted with matter. Then the Universe had cooled to about
6000 K (roughly the temperature at the surface of the Sun today), and
electrons could combine with nuclei, for the first time, to make stable,
electrically neutral atoms, which did not interact significantly with the
background radiation.
This pattern of ripples extends over much larger scales than even the largest
supercluster filaments traced out by looking at the patterns of galaxies on
the sky. But the geometrical structure of the pattern of ripples is the same
on all scales (if you look at half the sky, you get the same kind of pattern
as for the whole sky, or for a quarter of the sky, and so on). Furthermore, it
is the same kind of pattern as the pattern made by the bright galaxies. So it
does not seem unreasonable to suppose that this scale-invariant pattern is
typical of the way matter is (and was) distributed throughout the Universe.
In order to make exactly this kind of pattern in the distribution of mass
across the sky, you need a mixture of about two-thirds cold dark matter,
one-third hot dark matter and just a smear (perhaps 1 per cent of the total
mass) of ordinary atomic (baryonic) matter. In such a 鈥渕ixed dark matter鈥�
scenario, the CDM provides the clumps on which galaxies and clusters of
galaxies grow, while the HDM fills some of the space between the clumps,
smoothing the overall density of the Universe and reducing the density
contrast between the clumps and the voids.
Atomic matter 鈥� the bright stuff of stars and galaxies 鈥� feels the
gravitational influence of both kinds of dark matter, and so the foamy
distribution of galaxies we see today represents the averaged out influence of
waves made up of hot and cold dark matter (see Box 1).
In a possibly slightly over-optimistic assessment of these data, taking all
the observations at face value, one group of researchers, at Queen Mary and
Westfield College, London, has given precise figures for the mixture of
materials. They suggest that the Universe is made of 69 per cent CDM, 30 per
cent HDM and 1 per cent baryonic matter. And they even calculate the required
mass for the HDM neutrinos 鈥� 7.5 eV. This is 0.0014 per cent of the mass of an
electron, and comfortably below the upper limit of 20 eV so far set by
experiments.
An experiment which confirmed that the mass of the neutrino is around 7-8 eV
could, therefore, be taken as evidence from the laboratory that we really do
know what the Universe is made of. It would mean that the mass of the lightest
particle (apart from those with zero mass) had been predicted from
measurements of the entire Universe.
Experiments are also now in hand to detect CDM particles. Because the theories
do not say how many of these particles there should be in the Universe, the
range of possible masses for each particle is quite large; but once again,
detection of anything in the right sort of range would be strong vindication
of the mixed dark matter model of the Universe. One way or the other, the next
step forward in understanding what most of the Universe is made of is most
likely to be taken in laboratories here on Earth.
1. Bright side outlook
THE PATTERN of the distribution of visible stuff across the Universe (see
Figure on p4) shows bright galaxies distributed in filamentary chains and
sheets, with voids containing very little bright stuff. In simple terms, this
pattern can be explained if the average distribution of matter across the
Universe (baryonic matter and dark matter) varies slightly from place to
place, like the long smooth swell of an ocean. If galaxies then form only from
exceptionally high peaks in the initial density distribution (short-wavelength
ripples on top of the swell), they will be strongly concentrated in the crests
rather than the troughs of the long-wave perturbations. The distribution of
bright stuff in galaxies will be more 鈥渃lumpy鈥� than the overall distribution
of mass (see Figure top 3).
2. When MACHOs are no match for WIMPs
SOME of the dark matter needed to explain the dynamics of spiral galaxies may
be in the form of 鈥渕assive astronomical compact halo objects鈥�, or MACHOs (the
name was deliberately chosen as a counter to the term WIMP). Three groups of
astronomers have recently claimed that they have detected such objects by
their influence on the light from stars in a nearby galaxy, the Large
Magellanic cloud. The newly discovered objects account, though for only a
tiny fraction of the dark matter.
MACHOs could be either dim stars (brown dwarfs) each with about the same mass
as Jupiter, or black holes each with a mass up to a million times that of our
Sun. But although they could account for the invisible halos needed to explain
how galaxies like the Milky Way rotate, they are themselves made from baryons
produced in the big bang, and so cannot provide the much larger amount of dark
matter needed to explain the overall structure of the Universe. Although they
are dark, in the context of cosmological discussions they are part of the 1
per cent of the Universe made of ordinary atomic stuff.
Further Reading
The Afterglow of Creation, by Marcus Chown (Arrow, 1993).
In the Beginning, by John Gribbin, (Viking, 1993).
Cosmic Coincidences by John Gribbin and Martin Rees (Black Swan, 1991).