Galaxy survey shows that color matters. (Red Team, Blue Team).

Using the largest survey of galaxies ever compiled, astronomers have found that the cosmos divides sharply along color lines. Old, red galaxies clump tightly, while young, blue ones are more spread out. Although the standard theory of galaxy formation predicts the same general trend, it permits a continuum, from very tight to very loose clustering. The survey, however, denies the middle ground.

There's no ready explanation for this great divide among galaxies, says Alex S. Szalay of Johns Hopkins University in Baltimore. His Johns Hopkins colleague Tamas Budavari presented the findings this week at a meeting of the American Astronomical Society in Nashville.

The astronomers have analyzed 2 million of the roughly 50 million galaxies observed so far by the Sloan Digital Sky Survey. The survey, which uses a telescope at Apache Point, N.M., is scheduled to view some 100 million galaxies over the northern sky by 2005.

According to theory, the very first galaxies condensed from regions in the early universe where the density of matter was the highest. This material consisted mainly of invisible, mystery material dubbed dark matter. As time went on and gravity continued to pull material together, more ratified regions of the universe also began to form galaxies.

The first galaxies, which condensed less than a billion years after the Big Bang, are now elderly. They appear red because they stopped forming stars long ago and the longest-lived stars radiate most of their light at red or infrared wavelengths. Galaxies that formed later look blue because they contain a significant number of young stars, which emit bluer light.

The Sloan survey differs from other large astronomical surveys because it examines galaxies at a variety of wavelengths. This color information, which enables astronomers to easily estimate the distances to millions of galaxies, has revealed the difference in clustering, notes Szalay.

"There seems to be two [distinct] populations and not a gradual transition in clustering properties between red and blue galaxies," agrees Licia Verde of Princeton University. In contrast, data from a separate survey of 250,000 galaxies that Verde has analyzed offers only a hint of color segregation, she notes.

The main inference that Verde and her colleagues culled from the 2df Galaxy Red-shift Survey is that, on average, galaxies cluster in the same way on large scales as do the vast clumps of dark matter that presumably prompted their formation (SN: 1/5/02, p. 5). The new Sloan finding "may be telling something about the distribution of dark matter," she suggests.

The Sloan study represents "an important transition," says David N. Spergel of Princeton. "Astronomers are now using the large-scale distribution of galaxies to probe the physics of galaxy formation rather than to learn about the composition of the universe."