skyandtelescope.com • FEBRUARY 2020 61
Nor is G77-61 a fl uke. “Carbon dwarfs give us the poten-
tial for a lot of clues on the birth of the Milky Way,” Farihi
says. In 2018, his team analyzed the kinematics of more than
600 carbon dwarfs. Whereas only about 0.1% of stars near
the Sun belong to the halo, a whopping 30% to 60% of car-
bon dwarfs are halo members and most of the rest belong to
another old stellar population, the thick disk.
“They are waving a fl ag at you saying, ‘I am a very old
star,’” Margon says. But why should carbon dwarfs favor the
halo? First, Margon says, it takes a long time for a star to
become a red giant, then a white dwarf, and then a white
dwarf so cool and faint that a mere red dwarf outshines it.
Second, Farihi says, halo stars possess little oxygen, having
formed before supernova explosions shot much oxygen into
the galaxy. In most stars, oxygen outnumbers carbon and
soaks up the carbon by making carbon monoxide. To create
a carbon star, “you need to fl ip the carbon-to-oxygen ratio,”
Farihi says, so that carbon is free to form carbon compounds
other than carbon monoxide. Because the halo has little oxy-
gen, a halo star can convert a companion into a carbon star
by giving it just a small amount of carbon, explaining why so
many carbon dwarfs occupy the galaxy’s oldest population.
Moreover, the prototype carbon dwarf seems to be espe-
cially old. The typical halo star’s iron-to-hydrogen ratio is 2%
solar. But in 2005 Bertrand Plez (University of Montpellier,
France) and Judith Cohen (Caltech) reported they had found
that G77-61’s iron abundance is just 0.009% solar. That’s
shockingly low, even by halo standards, and suggests the star
arose near the dawn of time.
And the other carbon dwarfs? Alas, no one knows. “It’s
Carbon dwarfs are challenging amateur targets at best. Here we’ve listed a few that are well-placed during
January, February, and/or March for Northern Hemisphere observers. We’ve also listed several carbon giants
for comparison, but note that these are variables so their magnitudes are approximate.
hard work,” Green says. The stars are faint, which impedes
extracting elemental abundances from their spectra. Further-
more, deriving such numbers for carbon-rich stellar atmo-
spheres is diffi cult. Nevertheless, Margon’s team recently
obtained spectra that suggest not all carbon dwarfs are as
extreme as the prototype.
Still, G77-61’s meager iron abundance hints that some of
its carbon dwarf peers could offer insights into the primor-
dial era when the fi rst stars shone and exploded, catapulting
chemical elements into space. Although those stars have died,
the elements they forged survive on the surfaces of ancient
long-lived stars like G77-61, which means that carbon dwarfs
could help reveal the nature of the fi rst stars and in particu-
lar how massive they were.
Thus, the little red stars that most astronomers have never
even heard of may preserve fossil records of the short-lived
superstars that lit the early universe and ended the cosmic
dark ages more than 13 billion years ago.¢ KEN CROSWELL, who earned his PhD at Harvard Univer-
sity for studying the Milky Way’s halo, has written for National
Geographic, New Scientist, and Scientifi c American. He is the
author of eight books, including The Alchemy of the Heavens
and The Lives of Stars.Select Carbon Stars
Object Type Mag(v) RA Dec. Approx. distance (l-y)
G77-61 dwarf 13.9 03 h 32.6m +01° 58′ 255
2MASS J0933-0031 dwarf 14.6 09 h 33.4m −00° 32′ 487
2MASS J0818+2234 dwarf 16.2 08 h 18.1m +22° 34′ 1,320
2MASS J0742+4659 dwarf 16.7 07 h 43.0m +46° 59′ 482
UU Aur giant 5.3 06 h 36.5m +38° 27′ 1,600
BL Ori giant 6.0 06 h 25.5m +14° 4 3′ 2,280
W Ori giant 6.1 05 h 05.4m +01° 11′ 3,290
X Cnc giant 6.4 08 h 55.4m +17° 14 ′ 2,930
W CMa giant 6.6 07 h 08.1m –11° 55′ 1,800
R Lep giant 7.8 04 h 59.6m –14° 48 ′ 1,360
Right ascension and declination are for equinox 2000.0.“So right away, I got scared that my thesis
was in jeopardy.”
—PAUL GREEN