this small transit depth (Fig. 1) has been
observed repeatedly3,4. The amplitude of the
transit depth reported by Vidal-Madjar et al.
thus corresponds to an extended atmosphere
that has a radius 4.3 times that of Jupiter.
At this distance, the gravity of the planet is
no longer sufficient to retain the hydrogen
atoms, and so some fraction trickles away.
The calculated minimum escape rate for this
process would reduce the planetary mass
by only a negligible amount (0.1%) if held
constant over its roughly 5-billion-year age.
However, the data indicate a higher escape
rate, enough to trim the planetary mass
significantly.
Vidal-Madjar et al. conclude by pointing
out that this planet lies exceptionally close to
its parent star (completing a full orbit once
every 3.5 days). The authors speculate that,
at smaller orbital separations, the rate of
mass loss is correspondingly higher, and
planets that initially reside in such orbits
evaporate. Eight extrasolar planets have
similar orbital periods (between 3 and 4
days), and only recently has a planet been
found10 that has a shorter period (of 1.2
days). Seared to a temperature of 1,900 K,
this new-found planet is in the hot seat — its
existence provides an upper limit on the
evaporation rate.
Reservations about Vidal-Madjar and
colleagues’ findings centre on possible con-
taminating sources of emission that vary
with time at the wavelength concerned. Sun-
light scattered by the outermost layers of the
Earth’s atmosphere is large and variable as
viewed from the orbiting Hubble Space Tele-
scope. And the emission of the parent star
itself may vary in time or across the stellar
surface. The authors present good evidence
The rationality hypothesis brings order to much of the thinking in social sciences, and especially in economics.
According to this hypothesis, people act
solely to advance their own interests, inter-
preted in the most selfish way. But everyday
experience indicates that this is not entirely
true: people are willing to make sacrifices to
reciprocate favours or to take revenge. People
tip waiters even though they will never see
them again, and insults can lead to dangerous
fights. Experimental evidence that supports
these common-sense observations is accu-
mulating. On page 137 of this issue, Fehr and
Rockenbach1 describe experiments in which
positive and negative forms of reciprocation
— rewards and revenge — are potentially in
conflict. They find that the threat of punish-
ment for not rewarding a favour adequately
can diminish the actual reward. Such insight
into the motives that govern generosity, and
our notions of fairness, is vital in the search
for realistic principles that improve on the
rationality hypothesis.
Applied to game theory, the rationality
hypothesis gives rise to the concept of
news and views
NATURE | VOL 422 | 13 MARCH 2003 | www.nature.com/nature 125
that these effects are excluded, but their
conclusions will be all the stronger with
confirmatory data.
Modelling the outer reaches of the
planetary atmosphere in this high-energy
environment, taking these new observations
into account, should prove rewarding for
both planetary scientists and astronomers,
and there is always the promise of more
data to come. The transiting configuration of
HD209458b will ensure that it remains the
centre of attention in the immediate future,
and a keystone in our understanding of
planets outside the Solar System. n
David Charbonneau is at the California Institute of
Technology, 105-24, 1200 East California
Boulevard, Pasadena, California 91125, USA.
e-mail: dc@caltech.edu
1. Vidal-Madjar, A. et al. Nature 422, 143–146 (2003).
2. Mayor, M. & Queloz, D. Nature 378, 355–359 (1995).
3. Henry, G. W., Marcy, G. W., Butler, R. P. & Vogt, S. S.
Astrophys. J. 529, L41–L44 (2000).
4. Charbonneau, D., Brown, T. M., Latham, D. W. & Mayor, M.
Astrophys. J. 529, L45–L48 (2000).
5. Mazeh, T. et al. Astrophys. J. 532, L55–L58 (2000).
6. Seager, S. & Sasselov, D. D. Astrophys. J. 537, 916–921
(2000).
7. Brown, T. M. Astrophys. J. 553, 1006–1026 (2001).
8. Hubbard, W. B. et al. Astrophys. J. 560, 413–419 (2001).
9. Charbonneau, D., Brown, T. M., Noyes, R. W. & Gilliland, R. L.
Astrophys. J. 568, 377–384 (2002).
10.Konacki, M., Torres, G., Jha, S. & Sasselov, D. D. Nature 421,
507–509 (2003).
Behavioural science
Fair’s fair
Truman Bewley
A basic theory of behaviour holds that people act only in their own best
interests. But more complex motives are apparent in an experimental study
that shows that generosity is diminished by the unfairness of others.
The guidance molecule netrin-1 is
famous for its role in telling axons
— the long extensions sent out by
nerve cells — where they should
and shouldn’t go in the nervous
system. Reporting in Developmental
Cell (4, 371–382; 2003), Karpagam
Srinivasan et al. extend the
repertoire of netrin-1 activities to
non-neuronal tissues, showing that
it also acts to keep cells stuck
together during the development of
mammary glands.
It was already known that
netrins are secreted by many
cells outside the nervous system,
but until now no one had really
worked out what they were doing.
Srinivasan et al. looked more closely
at where netrin-1 occurs in the
developing mammary glands of
mice, and found that it surrounds
the cap cells — the single layer
of cells that caps the developing
gland, or bud — in a pattern that
is complementary to one of its
receptors, neogenin.
Extrapolating from netrin-1’s
function in the nervous system, it
might be predicted that it provides a
positional cue to guide moving cells
within the mammary gland.
Surprisingly, the authors found that,
instead, netrin-1 prevents cap-cell
movement. Loss of either netrin-1
or neogenin disrupted adhesion
between the cap-cell layer and
adjacent cells, and resulted in cap
cells moving into regions where they
would not normally go, as seen in
these images of normal (left) and
netrin-deficient (right) buds.
Furthermore, addition of netrin-1 to
isolated neogenin-producing cells
caused them to aggregate. So it
seems that netrin-1 may be required
in the developing mammary gland
simply to make sure that cells stick
together.
The authors are now
investigating the long-term
consequences of loss of netrin-1
for mammary-gland development,
and in particular the possibility
that these tissue disruptions
increase susceptibility to cancer.
At a more basic level, how does
the binding of netrin-1 to neogenin
immobilize cells? Studies of netrin-1
in the nervous system provide some
hints, but there may be more
surprises in store. Alison Schuldt
Developmental biology
Guidance molecule goes global
© 2003 Nature Publishing Group