For decades, the solar system set the template: small rocky planets close to the star, gas giants farther out, beyond a boundary called the snow line where icy grains help cores grow fast enough to grab thick atmospheres. A four-planet system orbiting a dim red dwarf star called LHS 1903 has just broken that template in the most direct way possible.
Two gas-rich planets sit in the middle of the system, sandwiched between smaller, denser rocky worlds on either side. The innermost planet is rock. The next two are swollen with gas. The outermost is rock again. That sequence, reported in May 2026 in the journal Science by a team led by astronomer Thomas Wilson at the University of Warwick, is the reverse of what every mainstream formation model would predict.
Wilson’s explanation is disarmingly simple. The gas disk that fed these planets “ran out of gas” before the outermost world could bulk up. The two middle planets formed their cores early and grabbed thick envelopes while the supply lasted. The outer core, still growing, missed its window.
Four planets, one month of orbits
All four planets hug their star tightly. Orbital periods range from about 2.2 days for the innermost world to roughly 29.3 days for the outermost, according to the team’s open preprint. Every one of them completes a full orbit in less time than Earth’s Moon takes to circle our planet. That proximity to LHS 1903, an M-dwarf star far cooler and smaller than the Sun, is what made detection possible in the first place: frequent transits across the star’s face, combined with the gravitational tugs each planet exerts on the star, gave the team both radii and masses.
Combining those two measurements yields bulk densities, and the densities tell the story. The innermost and outermost planets are compact and dense, consistent with rocky compositions similar to Earth’s. The second and third planets are larger and significantly less dense, pointing to substantial gaseous envelopes. That rocky-gas-gas-rocky architecture has no precedent among the thousands of multi-planet systems cataloged by missions like Kepler and TESS.
Why the gas ran out
Planet formation is a race against the clock. A young star’s protoplanetary disk, the swirling ring of gas and dust from which planets coalesce, does not last forever. Radiation from the star and internal processes drain the gas over a few million years. Any planetary core that has not grown massive enough to start pulling in gas before the disk disperses ends up as bare rock.
Wilson and his collaborators argue that around LHS 1903, this clock ran especially fast. The two inner cores crossed the threshold for gas accretion while the disk was still rich. The outermost core did not. The result is a system that looks, from the outside, as though someone assembled it backward.
The team also considered the most obvious alternative: that the planets formed in a more conventional arrangement and then swapped positions through gravitational scattering or collisions. Their dynamical analysis found no evidence for that scenario. The orbits are nearly circular and coplanar, lacking the eccentric, tilted paths that violent reshuffling typically leaves behind. The current layout, they concluded, reflects where the planets actually formed.
LHS 1903 is classified as a thick-disk star, a population that tends to be older and lower in metals than thin-disk stars like the Sun. Lower metallicity means less solid material available to build planet cores quickly, which could slow the onset of gas accretion and leave outer planets especially vulnerable when the disk begins to fade. That connection is physically plausible, but so few thick-disk planetary systems have been characterized in this detail that it remains an untested hypothesis rather than a confirmed trend.
What the data can and cannot tell us
The architectural fact, that LHS 1903 hosts a rocky-gas-gas-rocky sequence, rests on direct measurements. Transit photometry gives radii. Radial-velocity observations give masses. Densities follow from the combination. This methodology is the same pipeline that established compositions for benchmark systems like TRAPPIST-1, and it has been validated across thousands of exoplanets over two decades. The peer-reviewed Science paper subjects these measurements to rigorous error analysis.
The causal story is on somewhat softer ground. The gas-depletion timeline depends on theoretical models of disk evolution calibrated to other systems, not on direct observations of LHS 1903’s early environment. Quantitative estimates of the original disk mass and the rate at which gas escaped, whether through photoevaporation from stellar radiation or through accretion onto the star itself, are referenced in the paper by citation to earlier modeling work rather than independent measurements specific to this system.
There is also a question of atmospheric evolution after formation. Even if the outermost planet started with a modest gas envelope, billions of years of radiation from its nearby star could have stripped much of that atmosphere away. The current observations cannot fully separate how much of the inside-out pattern comes from the disk running dry early versus later atmospheric erosion. Disentangling those two effects will require detailed modeling of the star’s activity history alongside future measurements of what, if anything, still clings to the outer planet’s surface.
Direct atmospheric spectra, the kind of data the James Webb Space Telescope is now routinely collecting for other exoplanets, have not yet been published for any of the four LHS 1903 worlds. If JWST or a successor instrument finds that the outermost planet is truly bare rock with only a thin, secondary atmosphere, it would strongly favor the early gas-depletion scenario. If it turns out to retain a residual hydrogen envelope, the picture gets more complicated.
What changes if this holds up
For theorists who simulate how protoplanetary disks evolve around low-mass stars, LHS 1903 is an immediate challenge. Models will need to accommodate the possibility that gas dispersal timescales vary enough to produce architectures that look nothing like the solar system’s tidy arrangement. If thick-disk, metal-poor stars commonly lose their gas early, inside-out systems could represent a distinct formation pathway, not a one-off curiosity.
For observers, the system is a priority target. Its planets are close enough to their star and transit frequently enough that follow-up spectroscopy is feasible with current instruments. Confirming or ruling out residual atmospheres on the outer rocky world would convert a compelling but model-dependent interpretation into a much harder conclusion.
None of the four planets are likely habitable by Earth-like standards. They orbit too close to their star, and the rocky worlds are probably tidally locked, with one face in permanent daylight. But habitability was never the point. What LHS 1903 offers is a laboratory for testing how timing, disk chemistry, and stellar environment conspire to shape the final layout of a planetary system. The solar system gave us one answer. This small, inverted system around a faint red star suggests the question was always bigger than we assumed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
