By Robert Sanders, Public Affairs

BERKELEY– Several recent observations support a controversial theory that high velocity clouds of hydrogen gas seen
near our galaxy are left over from the formation of the Local Group of galaxies, which includes Andromeda and the Milky

If these puzzling clouds are indeed remnants of the very early history of the formation of nearby galaxies, they may
represent the earliest structures that formed in the universe some 15 billion years ago and should contain “dark matter.”
Dark matter is the missing mass astronomers know must exist in galaxies but can’t see because it doesn’t glow.

“These clouds are the building blocks of the Milky Way,” said Leo Blitz, professor of astronomy and director of the
Radio Astronomy Laboratory at the University of California, Berkeley. “They are interesting in and of themselves because
they should be remnants of the very early history of the Local Group. The clouds tie cosmology and the origins of the
universe into the history of our own local galaxies.”

Blitz details the theory and recent observations that support it in an invited talk on Jan. 14 at the national meeting of the
American Astronomical Society in Atlanta, Georgia. The talk is at 3:40 p.m. in the Centennial I and II rooms of the Hyatt
Regency Atlanta. To see movies that illustrate the theory, link to the team’s web site at

Blitz and four other astronomers from the United States and Holland proposed three years ago that these enigmatic
clouds are buzzing about the Local Group, colliding and merging with the galaxies there.

Discovered by radio astronomers in 1963, the gas clouds are moving at high velocity through space in orbits that don’t
conform to the nicely circular orbits of most other objects in the disk of the Milky Way. It became evident in 1972 that
they are not simply falling into the Milky Way, either, since some have high velocities away from us.

Several hundred of these clouds have been mapped within the margins of the Local Group, but their distances have been
hard to determine. Composed of atomic hydrogen, they have the mass of a small galaxy and are approximately 15,000
parsecs (50,000 light years) across.

According to the team’s scenario, these clouds are what remain of several thousand such clouds that formed in this region
early in the history of the universe and perhaps were the first large structures to form. Over the eons these clouds have
collided and coalesced into the galaxies we see today. Andromeda and the Milky Way are still in the process of forming
as they gobble up the remaining clouds.

Today the Local Group is a cluster of about 30 galaxies bound together by their own gravity and scattered throughout a
region about 1.5 million parsecs across, or about five million light years. Nearly 98 percent of the total mass within this
group is contained in the two huge spiral galaxies, Andromeda (M31) and the Milky Way, which contains the Sun and

No theory could explain the diverse observations of high velocity clouds until Blitz and his colleagues – David Spergel of
Princeton University; Peter Teuben of the University of Maryland, College Park; Dap Hartmann of the
Harvard-Smithsonian Center for Astrophysics in Cambridge, Mass.; and W. Butler Burton of the University of Leiden,
the Netherlands – proposed theirs in 1997. The theory was built on an analysis of radio observations of these clouds and
computer simulations of the formation of the Local Group.

“Based on this hypothesis, we made several predictions,” Blitz said. “So far, our theory has passed the tests.”

First, because these clouds would have formed long ago, the abundance of heavy elements should be much lower than in
the galaxies in the Local Group. Heavy elements like carbon, oxygen and iron are formed in stars, or as stars explode, so
their abundance was much less before star formation transformed hydrogen and helium into heavier elements.

A paper by University of Wisconsin astronomer B. Wakker in late November in Nature reported just that. The heavy
element abundance in the closest high velocity cloud is one tenth that in the Milky Way. The nearest cloud is so close to
us it stretches from horizon to horizon, and appears to be accreting onto the Milky Way, Blitz said.

A second prediction was that, if the high velocity clouds are part of the Local Group, they should show little of the
fluorescence (hydrogen alpha emission) seen in relatively nearby clouds of atomic hydrogen. Recently, a group led by Ben
Weiner of the Carnegie Institution in Pasadena was able to make measurements of the hydrogen alpha emission and
confirm the low expected levels of emission. The results are being presented separately at this week’s AAS meeting.

Thirdly, the clouds should have a low pressure and be “big and puffy,” Blitz said. A team led by Johns Hopkins University
astronomer Ken Sembach recently measured the pressure in one such cloud and found it “embarrassingly close to what
we expected,” Blitz said.

A fourth line of evidence also will be reported Jan. 14 at the AAS meeting by Blitz and graduate student Timothy
Robishaw. Looking at past radio observations of the Local Group, they discovered evidence of a large cloud of hot,
ionized, million-degree gas around the Milky Way and Andromeda that would be expected from the collision of high
velocity clouds. The gas, from perhaps a thousand cloud collisions over the history of the Local Group, could permeate
the entire group.

“This is something we would have expected, but it needs to be confirmed by direct observation,” Blitz said. The detection
of emission from highly ionized oxygen by the FUSE satellite, the subject of a press conference earlier in the week, will
provide crucial evidence.

The theory proposed by Blitz and his colleagues solves several conundrums about the formation of galaxies and stars.
Among these is the apparent lack of sufficient visible gas in many spiral galaxies like the Milky Way to continue to fuel the
currently observed amount of star formation. Rather than postulate that all the galaxies we see are in the last third of their
lives and have run out of fuel, the theory suggests instead that high velocity clouds fall in and continually replenish the gas
needed to fuel star formation.

Another implication is that these clouds are one of few places where the stuff of the early universe can be found relatively
uncontaminated by later star formation.

“In these clouds, we can still study the universe we had 15 billion years ago,” he said.

Astronomers have seen similar high velocity clouds in more distant groups of galaxies. The theory implies that they too are
left over from the formation of their group.

Among the many unanswered questions about high velocity clouds is: What is going on inside the clouds themselves? Blitz
thinks many of the clouds are associated with dwarf galaxies seen scattered about the Local Group. Since there are more
high velocity clouds than known dwarf galaxies in the group, the implication is that there are many more such galaxies than
we see today.

The recent reanalysis of radio data by Blitz and Robishaw turned up hydrogen clouds – just like the high velocity clouds –
associated with 10 out of 21 dwarf spheroidal galaxies in the Local Group. The association is one more confirmation that
the high velocity clouds are part of the Local Group, and thus, as Blitz says, “the last gasp. All other high velocity clouds
already have been gobbled up by Andromeda and the Milky Way.”

The research was supported by the National Science Foundation.