evolutionary one, rather like the familiar lineup that leads from
crouching monkey to upright human. There was some support
for this at the time Hubble was working, and it seems reasonable
to believe that galaxies which started off as featureless balls of
stars would slowly collapse under their own gravity to form
discs. Within those discs would be spiral arms, the inevitable
result of disrupting a rotating system, and these might well then
unwind over time, completing the journey from one end of the
tuning fork to the other. It’s an attractive picture, but one that
turned out, sadly, to be complete nonsense. Galaxies don’t behave
like that. Nonetheless, Hubble’s classification still tells us about
the shape of galaxies, and it’s still used nearly a century later.
The idea that we’re talking about classifying galaxies might
seem archaic in and of itself. The discoverer of the atomic nucleus
and pioneering popularizer of science, Ernest Rutherford, dis-
missed such work as mere stamp collecting (physics alone, and
his kind of physics at that, being exempt from being stamped
with his philatelic scorn). But sorting things into categories often
marks the first attempt to understand something scientifically,
and it can be important even within the inner sanctum of phys-
ics, or, in this case, astrophysics.
Those taking a Rutherfordian view might expect classifica-
tions based on mere observation to cease to be useful as we start
52 The Crowd and The Cosmos
to really understand the physics of galaxies. Simple labelling of
what you see is fine to begin with, but Proper Science will pro-
ceed differently, and we should expect applying labels as straight-
forward as a shape to be left as a curiosity in scientific history.
Just such an upheaval might be seen to be taking place right now
in much of biology, for example, as genomic analysis rearranges
our ideas about the relationships between species. A deeper and
more useful classification of the tree of life can be found by look-
ing at where the action really is—in the DNA—rather than by
carefully observing anatomical features. But for a long time, all
biologists had to work with was the power of observation (and
plenty would still argue for traditional, rather than genetic work
as a means of making progress).
We might imagine that astronomers, too, will find some fun-
damental measurements that record a galaxy’s history and
explain its behaviour today. If such fundamental parameters do
exist, though, we’ve yet to find them and thus, lacking any celes-
tial equivalent to genetics, astronomers have little choice, more
than eighty years after Hubble, than to carry on dutifully sorting
galaxies into categories based on an antiquated tuning fork.
The difference between elliptical and spiral galaxies isn’t a
temporary one. Left in isolation, a typical spiral will have enough
fuel, primarily hydrogen, to go on forming stars for billions of years.
As mentioned earlier, recent star formation also means that most
spiral galaxies will be blue and elliptical galaxies red; the most
massive stars are blue but they are also short lived. This is some-
what counter-intuitive, as you might expect the most massive
stars to have more fuel for nuclear fusion on hand and thus to
last longer. In fact, the rate at which nuclear fusion proceeds is
exquisitely sensitive to temperature, and that in turn depends on
the pressure exerted at the core by a star’s mass. More massive
stars have much hotter cores, and so burn through their fuel
The Crowd and The Cosmos 53
much faster, and the brightest and bluest of them may live for
only a few hundred million years. Spiral galaxies, and in particu-
lar the spiral arms where most star formation takes place, are
thus speckled with brilliant blue stars, while ellipticals are, for
the most part, red.
Distinguished by their properties, the two sets of galaxies are
also separated by their environments. The two types of galaxy
often live in separate places. Take the neighbourhood we happen
to find ourselves in. The Milky Way and the Andromeda Galaxy
are two of the three large galaxies in our Local Group. The Milky
Way is a spiral, and Andromeda is also a disc rather than an ellip-
tical. Its exact morphology is somewhat ambiguous, a situation
not helped by its nearly edge-on presentation as seen from Earth,
with the most prominent feature being a ring of recent star for-
mation. But if we broaden our definitions to distinguish disc gal-
axies, whether or not they have spiral arms, from ellipticals then
the two comfortably sit within the same category.
This reflects the fact that the shapes we see are essentially a
result of the dynamics of the material within the galaxy. Disc gal-
axies like the Milky Way are composed of material which orbits
the centre in an organized fashion. The Sun, for example, travel-
ling at a little over half a million miles per hour, completes one
orbit around the centre of the Milky Way in just over 225 million
years, a length of time that is sometimes referred to as a ‘galactic
year’, and most of its neighbours will do the same. The disc of the
galaxy is relatively thin. As Monty Python’s ‘Galaxy Song’ has it,
the Milky Way measures 100,000 light years side to side, but is,
out by us, only a few thousand light years wide. Think of the gal-
axy as a large fried egg—ten centimetres across and a few milli-
metres thick—and you won’t be far wrong. By contrast, individual
stars in an elliptical galaxy are each travelling on their own orbit
around the centre of the galaxy, but they are all inclined at
54 The Crowd and The Cosmos
different angles, producing the football or rugby ball shape we
see. There is no organized set of orbits as there is in a disc, and
this reflects the real physical difference between the two sets of
galaxies.*
M33, the third and final large member of our Local Group of
galaxies, is also a spiral, and a beautiful one at that. Its multiple spiral arms are in danger of losing their unique identity, so dominated are they by the presence of young star clusters. This sort of
arrangement is sometimes known as a ‘flocculent’ spiral, its
appearance somewhat reminiscent of a scattering of tufts of
wool. That detail aside, our local neighbourhood’s large galaxy
population shows signs of segregation: three large galaxies, all of
them spirals.
Should we be surprised? The key is in understanding where we
live. If you rank the environments in which galaxies are found,
from the most densely populated regions to those which are
emptiest, you find that we live in the cosmic suburbs. The Milky
Way is not a hermit, in splendid isolation, but nor is our patch of
space especially crowded. Looking around, this sort of environ-
ment is typically dominated by spiral systems, which seem to
thrive today in less dense environments. Ellipticals dominate the
Universe
’s cities, mostly hanging out in vast clusters and super-
clusters of galaxies.
This talk of environments for galaxies would have surprised
the astronomers of just a few decades ago. One of the most
* I should probably point out that the use of these terms in the astronomical literature is often confusing, with things being made much worse by the presence of ‘S0’ galaxies that look like ellipticals but which have a disc hidden in them. People often use the terms ‘early-type’ for ellipticals and ‘late-type’ for spirals, the names deriving from the old mistaken idea that ellipticals eventually turn into discs. I’ll just use elliptical and spiral here, but really I’m trying to divide disordered systems from those galaxies, typically spiral, where a regular disc is the most prominent feature.
The Crowd and The Cosmos 55
profound and interesting discoveries of the last few years has
been that there is structure in the Universe even on the largest
scales that we can map. Take the million galaxies of the Sloan
Digital Sky Survey, for example. The survey was critically able to
make a three-dimensional map of the Universe, recording not
only the position of galaxies in the sky but also their distance
from Earth. Cepheids aren’t visible at these distances, but we can
make use of another yardstick—the expansion of the Universe
itself.
I already mentioned that Hubble and others provided evidence
that the Universe is expanding, and if you work backwards that
leads you to the idea that everything began in a hot dense state
just after something called the Big Bang which occurred 13.8 bil-
lion years ago (give or take a hundred million years or so; the
accuracy with which the age of the Universe can be determined
still astounds me). By expansion we’re not talking about the
movement of galaxies through space, but as the expansion of
space itself; think not of actors rushing away from each other,
but of a stage which itself expands, widening the space between
everyone standing on it. That expansion of space, which pro-
ceeds today at a rate such that every thirty billion billion kilome-
tres’ worth of space grows by seventy kilometres a second, has
an effect on the light travelling through it.*
The expansion stretches the light to longer wavelengths,
which correspond to redder colours. More distant galaxies cap-
tured by Sloan, therefore, look red when compared to their
local compatriots, purely because of this redshift. Their spectral
* In more sensible if less comprehensible units, astronomers would write this as 70 kilometres per second per megaparsec. A megaparsec is 3.26 million light years, or just over thirty billion billion kilometres. In whatever units, this value is known as Hubble’s constant, though it isn’t constant, but rather something that will change through time as the contents of the Universe act on the expansion.
56 The Crowd and The Cosmos
pattern—what you see when you split the light up into its com-
ponent wavelengths—shifts too. Elements emit or absorb light
at particular wavelengths, and we can create a list of those wave-
lengths by experiment in the laboratory. We still see the same
pattern of lines corresponding to elements such as oxygen in the
galaxy spectra, but shifted, and so we can compare the measured
galaxy spectrum to the standard one and measure the redshift
that way.
This measurement contains information about how the
Universe has expanded during the millions of years light from a
galaxy has been travelling towards us. (When we want to talk
about the intrinsic colour of a system, as when considering star
formation, we usually have to adjust for this effect.)
Conversely, if you understand the expansion then the redshift
becomes a measure of distance to a particular galaxy. The Sloan
survey’s efforts thus included taking a spectrum for each of
nearly a million galaxies, a task made more complicated by the
fact that only the best nights, when the air is stillest, could be
used for this delicate work. Sloan did, however, have an advan-
tage over previous surveys, in being designed to take hundreds of
spectra at once. For each patch of sky it might observe, a metal
plate was prepared, drilled with holes corresponding to the posi-
tion of each galaxy. Into each of these holes, a fibre optic cable
was plugged, carrying the light from a distant galaxy on the last
few metres of its journey to the instrument that would analyse it.
Compared to the old method of pointing the telescope at each
galaxy in turn, this provided for great efficiency, but at the cost of having to complete the laborious task of plugging fibres into the
holes in the plate. A rival survey, called 2dF, spent great time and
effort producing a fibre-handling robot to do the job. It’s a better
long-term solution, but Sloan just relied on the efforts of junior
The Crowd and The Cosmos 57
astronomers, whose labour proved an easy if unsatisfying solu-
tion to the problem night after night and plate after plate.
The results of all of that effort were spectacular. Sloan and
similar surveys revealed with clarity what previous data sets had
only led astronomers to suspect. The Universe around us is a
honeycomb, a cosmic web of clusters and of filaments which
wind their way around enormous empty voids. Each of these
superstructures is made up of thousands or hundreds of thou-
sands of galaxies; the largest of them, the Sloan Great Wall, is
more than a billion light years across.* The fact that there are dif-
ferences even on these enormous scales has consequences.
Jumping millions of light years in any direction from the Milky
Way could place you in a very different place, surrounded by dif-
ferent types of galaxies or even, if you end up in a void, with very
little company at all.
In the densest parts of these superstructures, a spiral galaxy
like our own would look rather out of place. Clusters and super-
clusters of galaxies are the realm of enormous galaxies which are
almost uniformly elliptical. Even smaller examples like the Virgo
Cluster, at 54 million light years away our nearest example, have
plenty of ellipticals among their population. The densest part of
that cluster, centred on a galaxy called M87 which itself weighs in
at 200 times the mass of the Milky Way, is almost entirely popu-
lated only by ellipticals, with spirals mostly relegated to a few
surrounding groups of galaxies which might still be in the pro-
cess of being absorbed by the main cluster.
* Studying such a vast structure is rather difficult, and arguments about the Sloan Great Wall still rage. There’s some evidence that it isn’t one, but rather three different structures superimposed onto each other. The problem is not merely of cartographic interest; understanding the odds of such a large structure forming provides neat constraints on cosmological theories.
58 The Crowd and The Cosmos
Given these facts, galaxy classification may seem to be a simple
matter. Ellipticals are red, and live in great galactic cities. Spirals are blue, and live in the sticks. Anyone wanting to select one at
the expense of the other could just sort by colour. Pick a handful
of red galaxies, and they’re likely to be ellipticals. Stick to blue
ones, and you’ll probably come up with a fistful of spirals.
Alternatively, we could take a bunch of elliptical galaxies and
expect them to be red, or collect spirals and then look and find
they were almost all blue.
This close connection between colour and shape makes it
sound like visitors to the Galaxy Zoo website were wasting their
time, but things were much more interesting than that. I’d first
had an inkling that things might not be simple when, nearly a
year to the day before the launch of the site, I’d gone up to Oxford
for a job interview. I was coming to the end of my PhD (and more
practically, to the end of the grant I’d been awarded to complete
it which was paying the bills). I was nervous, and intimidated by
the place and the department, and the most critical part of the
day consisted of giving a seminar on my work to a busy crowd.
I’d rather more of them had been distracted by whatever was on
their laptops than listening to me blather on, trying to demon-
strate my deep and abiding desire to work on extragalactic astro-
physics by reviewing work I’d just completed.
I’d been working on updating a classic, simple model of how
the details of the expansion of the Universe affects the galaxies
that form within it. Less than a decade earlier, careful measure-
ments had revealed to astronomers that the expansion of the
Universe is not slowing down under the influence of gravity, as
might have been expected, but rather that it was speeding up. We
The Crowd and The Cosmos 59
still don’t know why this is happening,* but it has profound con-
sequences for every aspect of our understanding of how the
Universe evolves.
The point of my work was to reconsider the formation of ellip-
ticals. This sort of work, more connected to theory than my nor-
mal stuff, is fun, but not my natural territory—hence being
worried about the talk. As it turned out, I’d barely started talking, and was still feeling exposed, nervous, and uncomfortable, when
I was interrupted, loudly and insistently, from the other end of
the room.
I’d just been going through the motions of explaining where
my data came from and how I’d selected a sample of reliably
The Crowd and the Cosmos: Adventures in the Zooniverse Page 8