When Edwin Hubble devised his now famous "tuning-fork" diagram in 1926, he sought to organize galaxies' many shapes into a sensible, well-ordered sequence.

The Hubble Sequence

The modern Hubble sequence consists of elliptical galaxies, whose shapes range from spherical (E0) to elongated (E7); spirals, which subdivide into those with (SBa to SBc) and without (Sa to Sc) a central bar; and lenticulars (S0), which share characteristics of ellipticals and spirals. A fourth type, irregulars, is not shown. Sprials dominate, with more than 70% of the total galaxy census.

Ville Koistinen

He realized that spirals sometimes sport a central bar and sometimes do not, and that ellipticals are gigantic star balls — all bulge and no arms. In Hubble's scheme lenticulars represent a hybrid of those types, and all the peculiar-shaped leftovers became known as irregulars.

Yet I'll admit that when I see or hear the word "galaxy," I conjure up a vision of just one of these: an immense, stately spiral of stars like the one shown at lower right. Seasoned observers can rattle off the names of dozens of pinwheels from memory: the Andromeda galaxy, M51 and M101 in Ursa Major, and M74 in Pisces, to name a few. Ellipticals and irregulars rarely come to mind.

Cosmologists spend a lot of time thinking about spiral galaxies too — not to admire their beauty but to figure out how they exist at all. Their very shape indicates a star system that's been stable and largely unperturbed for billions of years. Yet the early universe was hardly a tranquil place. Crowding caused many young galaxies to collide, merge, and tear each other asunder.

Spiral galaxy M83

The galaxy M83 has well-defined spiral arms that trace regions of active star formation.

Adam Block / NOAO / AURA / NSF

So it's a minor miracle (and a topic of considerable debate) how all the spirals we see today managed to endure all that mayhem unscathed. "The formation of spirals is a problem," admits Christopher Conselice, a galaxy specialist at the University of Nottingham. "We don't know how they formed, or how they survive all those mergers."

Now that telescopes can peer to ever-grater distances and thus farther back in time, astronomers are attempting to take a census of the shapes and numbers of galaxies that existed 6 to 8 billion years ago. Using observations from the Sloan Digital Sky Survey and from the Hubble Space Telescope, François Hammer and Rodney Delgado-Serrano (Paris Observatory) led an effort to catalog 116 local galaxies and 148 distant ones, respectively. In effect, they've created two Hubble sequences: one for the present and one for the circumstances 6 billion years ago. Their results appear in two articles just published in the European journal Astronomy & Astrophysics.

Surprisingly, the assorted beasts in the galaxy zoo were very different long ago. Peculiar-shaped irregular galaxies were far more common (52% of the total) and spirals relatively scarce (31%) — just the opposite of what's observed today (10% irregulars, 72% spirals).

Using computer models to trace the role of interstellar gas and rates of star formation during mergers, Hammer and his team conclude that many of the spiral galaxies seen today must have resulted from collisions between irregular systems. They can't prove that spirals arose phoenix-like from the ashes of titanic collisions — the Hubble views don't reveal scads of merging galactic blobs. But the model suggests that spirals were "rebuilt" following particularly gas-rich mergers.

Simulated galaxies

Simulated galaxies generated by the GALFORM computer model. The yellow objects are most distant and therefore appear as they were 13 billion years ago, while those closer are seen as they looked more recently. Click on the image for a larger view.

A. Benson / Univ. of Durham

But Tanner and Delgado-Serrano aren't the only ones trying to crack the Hubble sequence. A competing paper, just published in the Monthly Notices of the Royal Astronomical Society, concludes that spirals would not have fared well in the bump-and-grind chaos of the early universe. Instead, argue Andrew Benson (Caltech) and Nick Devereux (Embry-Riddle Aeronautical University), spirals abound today because they managed to escape violent interactions. "The dense galaxy clusters we see today are actually unusual," Benson explains. "The quiet regions of the universe, then and now, are more common than we thought."

The Benson-Devereux computer model, called GALFORM, takes a set of assumed starting conditions and runs it forward through time. In a sense, Benson notes, the more we know about the early universe, the more difficult the modeling becomes. GALFORM incorporates the effects of unseen dark matter (which helps draw galaxies together quickly) and even more perplexing dark energy (which then pushes them apart). The result is a very good (but not perfect) match to the numbers and types of galaxies observed today and at times past.

One hurdle shared by all these models is that we don't yet know the true shapes of the earliest galaxies. Benson points to the celebrated Hubble Deep Field image, for example, in which "they just look like blobs of light" that sometimes appear as weird chains and other shapes that aren't seen now.

Fortunately, a massive new effort — approved just three weeks ago — will use the revamped Hubble Space Telescope to observe the early universe with unprecedented clarity. Led by Sandra Faber (University of California, Santa Cruz) and Harry Ferguson (Space Telescope Science Institute), the Cosmology Survey Multi-Cycle Treasury Program will employ Hubble's new Wide Field Camera 3 to survey some 250,000 galaxies in five regions of the sky. With luck, the resulting views should reveal crucial details about how galaxies looked at least 12 billion years ago. Don't expect results anytime soon, though. Faber and Ferguson don't yet know how much HST time they'll need — but it's at least a 100 orbits' worth!

If you're impatient (or even if you're not), then let me suggest that you join Galaxy Zoo — a "citizen-science" effort to classify ancient galaxies and scan them for supernovas.


Image of Peter Wilson

Peter Wilson

February 17, 2010 at 9:39 am

"GALFORM incorporates the effects of unseen dark matter (which helps draw galaxies together quickly) and even more perplexing dark energy (which then pushes them apart)." Actually, the drawing-together must result in pushing-apart. It is cause-effect; action-reaction. This should not be perplexing. See http://www.dark-energy.org

You must be logged in to post a comment.

Image of Frank Skorina

Frank Skorina

February 22, 2010 at 1:22 pm

Most observable galaxies are spiral but most galaxies are ellipical. This is because most elliptical galaxies are small and dim. If we survey distance galaxies, I would guess that we would be unable to resolve small, dim galaxies. Our local group consists of ~36 galaxies. Three are spirals (Milky Way, Andromeda, Triangulum), a few are irregulars, but most are ellipticals and dwarf ellipticals.

You must be logged in to post a comment.

You must be logged in to post a comment.