Evidence from observations and computer simulations supports a picture of galaxy growth that isn't dominated by the rough-and-tumble crashes of big galaxies.

To make a giant cookie, you would probably start by grabbing two clumps of cookie dough and smooshing them together. Astronomers long thought the same went for galaxies: want to build a big galaxy? Merge two galaxies together. Voilà.

This detailed face-on view of NGC 6744, taken by the MPG/ESO's 2.2-meter telescope at the La Silla Observatory in 2000, shows a spiral galaxy similar to the Milky Way in many ways, one of which being the presence of a companion galaxy seen in this picture as a smudge at the lower right corner. Observations suggest that many galaxies do much of their growing by eating smaller companions.


But astronomers are finding that major mergers aren’t the fundamental galaxy builders they once seemed. Last year, a study of more than 400 Milky-Way-mass galaxies (i.e. those with tens of billions of solar masses) showed that, contrary to theory, disk galaxies grow their central bulges simultaneously with their disks, and not via mergers.

The less-is-more theme with mergers continued at the winter meeting of the American Astronomical Society. Sugata Kaviraj (University of Hertfordshire and University of Oxford, UK) presented his team’s study of 80 massive galaxies whose light left them more than 10 billion years ago, corresponding to a redshift of about 2. This epoch is a popular one with astronomers, because star formation in the universe peaked around this time. And because star formation builds up a galaxy’s stellar mass (which means it builds up the galaxy), astronomers can track galaxy growth by studying star formation.

The Mice

Some galaxies do suffer major mergers, including NGC 4676 (a.k.a. "The Mice"), a spectacular pair of colliding galaxies 300 million light-years away in Coma Berenices. They will eventually merge into a single giant galaxy. This Hubble composite-color image is assembled from exposures made through blue, orange, and near-infrared filters.

Kaviraj and his colleagues looked for signs of a major merger — such as extended tidal features or double cores — in these early galaxies. (They defined a major merger as one in which the two original galaxies had about the same mass.) But they found that at most 27% of star formation activity during this period arose thanks to such mergers. After comparing the star formation in non-interacting galaxies with that in galaxies suffering mergers, the team concluded that the mergers’ actual contribution to cosmic star formation in this era might be more like 15%.

In a separate study of about 330 newborn elliptical galaxies, Kaviraj’s team also found that more than 50% of the blue (a.k.a. star-forming) ball-shaped galaxies don’t seem to be driving their star formation via major mergers. That matches up with theoretical work suggesting these galaxies can form without such a clash.

Globular cluster NGC 2419

The Japanese 8.2-meter Subaru telescope took this image of the globular cluster NGC 2419. Most globular clusters occupy the halo, the large spheroid of stars and diffuse hot gas that surrounds the galaxy.

Subaru Telescope / NAOJ

Overall, it looks like galaxies power most of their star formation — and therefore their growth — with minor mergers and gas accreted from their environs. The Milky Way is a prime case, as it shows no signs of a recent major merger. Instead, it’s beefed itself up by snacking. Zhibo Ma (Case Western Reserve University) reported at the meeting that his team’s work with the SEGUE K Giant Survey, an extension of the Sloan Digital Sky Survey focusing on cool giant stars, supports this view of the Milky Way. The team distinguishes several stellar populations in the Milky Way’s halo, the large sphere of older, cooler stars that encircles the galaxy. These stellar streams suggest that our galaxy has built up its halo in an ongoing process, accreting and disrupting dwarf galaxies as it ages.

These streams have particular chemical makeups, developed when the stars were first born in whatever extragalactic clump of gas the Milky Way ripped up and ate. But the streams may be able to tell us even more about their progenitors than these compositions reveal. Heidi Newberg (Rensselaer Polytechnic Institute) and her colleagues are working on using stellar streams to figure out the dark matter content of the dwarfs our galaxy has eaten, and of our galaxy itself. They’re doing it with a citizen science project called MilkyWay@home, which uses volunteers’ computer time to create a detailed, 3D view of the Milky Way. The project is still fairly new, but it promises to provide a fascinating glimpse into how our galaxy formed.


AAS presentations (click here for meeting program):
S. Kaviraj et al. "The insignificance of major mergers in the early universe." Abstract #310.07
Z. Ma et al. "The SEGUE K Giant Survey." Abstract #336.02D (dissertation talk)
H. J. Newberg et al. "Relating Dark Matter to Tidal Streams with MilkyWay@home." Abstract #336.03

S. Kaviraj et al. "The insignificance of major mergers in driving star formation at z~2." Monthly Notices of the Royal Astronomical Society, February 11, 2013.

S. Kaviraj et al. "Newborn spheroids at high redshift: when and how did the dominant, old stars in today's massive galaxies form?" Monthly Notices of the Royal Astronomical Society, January 11, 2013.


Image of Peter Wilson

Peter Wilson

January 18, 2014 at 11:56 am

It helps to remember that stars and galaxies do not grow so much as collapse. A star begins as a diffuse cloud of gas, many light-years across, then collapses. What causes that? The core grows, but the process is one of collapse. Ditto for galaxies. They become more massive, but the phenomenon is a “falling together” into deeper and deeper gravitational wells. The universe began with matter, everywhere. There were no voids, anywhere. Matter falling into galaxies created the emptiness around them. Mass is not being added to the universe; the rate of galaxy growth is the rate of local collapse.

You must be logged in to post a comment.

Image of Dieter Kreuer

Dieter Kreuer

January 20, 2014 at 3:44 am

...but not spiral galaxies. In order for a disk to form, particles need to collide and get rid of momentum inclined to the disk plane. I've often wondered how spiral galaxies were supposed to grow from mergers of smaller galaxies, when they just throw in more or less point-like stars that would retain their original momentum. This would logically result in an elliptical galaxy. Spiral galaxies, based on their shape, instead appear to form from collapsing gas that can flatten itself to a disk, and it's encouraging to hear that the findings discussed in the article agree with just that. Now the question remains, if spirals start small, where is all the gas and when will the first spiral galaxy be caught in the act of accreting?

You must be logged in to post a comment.

Image of Gregg Weber

Gregg Weber

January 20, 2014 at 6:33 pm

Is there larger mass objects in intergalactic space that could be drawn into galaxies to increase the mass?
Could these be seen by looking for blinking in a pattern of a transit of further objects that are large enough to be a white wall with pixels blinking in a straight line?

You must be logged in to post a comment.

Image of Peter Wilson

Peter Wilson

January 21, 2014 at 10:38 am

Gregg: There are clouds of gas out there sans stars, and they do add to galaxy growth. They are pretty hard to see. Dieter: There are no small spiral galaxies. Something about their dynamics requires a large mass. Also, a non-spinning cloud of gas will not flatten or “pancake.” So spiral galaxies need a lot of mass and angular momentum, while ellipticals do not. It is true that in a collision of elliptical galaxies, the trajectories of already-existing stars will not flatten. But most of the mass in a galaxy is in the form of gas, which does pancake, and which provides the material for new star formation. The pattern seen in spiral galaxies is not the distribution of stars, per se; the arms outline where NEW stars are forming. I’m not seeing how spiral galaxies could form except by collision, like cream poured into a cup of coffee.

You must be logged in to post a comment.

Image of Roland Boudry

Roland Boudry

January 21, 2014 at 2:55 pm

Also dark matter (27 %) and dark energy (68 %) are playing an important role in growing galaxies.

You must be logged in to post a comment.

Image of Dieter Kreuer

Dieter Kreuer

January 22, 2014 at 1:57 am

@Peter Wilson

Thanks for clarifying. I suppose you are professional astronomer? I wasn't aware that small galaxies cannot be spirals (at least, M33 is much smaller than our galaxy and still a spiral). Two questions: doesn't any collapsing cloud have a random non-zero momentum? If not, it would collapse into a big black hole with at most a small disk surrounding it. Hence, collapsing clouds should usually have a chance to pancake (and so do much smaller protoplanetary disks). And second, I understand that (young) galaxies are mostly gas so they might coalesce to a pancake which forms new stars that eventually take on the spiral shape, however, they also include some finished original stars. How do you force these stars to join the disk, or if you don't, where are they? Are they just invisible members of the halo population, because old stars tend not to be luminous? Very interesting discussion, I'm eager to learn.

You must be logged in to post a comment.

Image of Richard Sauder

Richard Sauder

January 22, 2014 at 4:18 pm

@ Camille Carlisle

This was an interesting article, but it is irritating how some authors use such unscientific terms to describe scientific concepts.

Case in point: Galaxies do not "snack". They merely accumulate matter that falls into them. It adds nothing but confusion to use irrelevant terms to describe concepts that are easily described with correct terms.

You must be logged in to post a comment.

Image of Bruce Mayfield

Bruce Mayfield

January 23, 2014 at 10:08 pm

Peter, you’ve made good points about gas collapse leading to the formation of both stars and galaxies, and collapse builds both larger (galactic clusters) and smaller (planets) things too. But since conservation of angular momentum is a universal principle the formation of discs will be forced in any collapsing cloud of any size if there is significant rotation. So Dieter is correct to question your claim that there are no small spiral galaxies. They ARE rare, but dwarf spiral galaxies do exist: the Wikipedia article “Dwarf Spiral Galaxies” lists 9 examples. I would guess that the reason for this rarity compared to other types of dwarfs is that small spirals, like all dwarf galaxies, are easily torn apart by larger neighbors. Dieter also raises a good question about how Spirals grow through accretion of smaller galaxies. My guess on this point would be that today’s large spirals started out with large gas discs, and that accreted smaller galaxies mainly contribute to the halo, bulge and core star populations.

You must be logged in to post a comment.

Image of Peter


January 24, 2014 at 9:08 am

Should have said small spirals are rare. The flattening tendency is always at odds with the rounding tendency, and like so many countervailing forces in nature, one or the other can dominate, depending on circumstances, and which part we look at, or which part is visible. Sometimes it’s a draw. Am assuming ellipticals have some angular momentum, but not enough to flatten. Again, the simplest way to imagine galaxy collapse/growth is through collision, but the point of the article is that while collisions are important, they are not the dominant process. To use a cliché, there are a lot of unanswered questions regarding star and galaxy formation. And Roland? Don't get me started on dark energy!

You must be logged in to post a comment.

You must be logged in to post a comment.