Mosquitoes are more than an itchy nuisance. They can carry serious diseases, including Zika, West Nile, yellow fever and chikungunya. Now after testing 11 types of mosquito repellents, researchers say they’ve identified the products most effective at warding off the bloodsuckers.

Spray-on repellents with DEET or a refined tree extract called oil of lemon eucalyptus are most likely to keep you bite-free, the scientists report online February 16 in the Journal of Insect Science. The OFF! Clip-On repellent, which puffs out a vapor of the chemical metofluthrin, killed every mosquito in the cage. But Hansen says the mosquitoes couldn’t escape, so they probably got a higher dose than they would in a natural setting.
Other tested repellents such as a citronella candle simply don’t work, says study coauthor Immo Hansen, an insect physiologist at New Mexico State University in Las Cruces.

“There are a whole lot of different products out on the market that are sold as mosquito repellents, and most of them haven’t ever been tested in a scientific setting,” Hansen says.

To evaluate the repellents, the researchers used a person, safely protected from bites, as “bait.” The volunteer sat in a wind tunnel as her alluring scent — and repelling chemicals — were pulled toward a cage of Aedes aegypti mosquitoes.
The three-compartment cage allowed the mosquitoes to move toward or away from the volunteer. After 15 minutes, the researchers determined the portion of mosquitoes that had moved into the compartment closest to the volunteer.
Three deterrents did little to dissuade the insects: bracelets with geraniol oil, a sound machine that buzzes like a dragonfly and a citronella candle (which appeared to slightly attract the mosquitoes). Burning a candle releases carbon dioxide, which might have drawn the mosquitoes, which home in on a human meal by sensing exhaled CO2 (SN: 3/18/17, p. 10).

Repellents face-off
Researchers measured attraction rates of A. aegypti mosquitoes to a person one meter or three meters away who was wearing or seated next to the repellent. Attraction rates are the percentage of total mosquitoes, averaged over four tests, that flew toward the person.
These repellents were not significantly different from the no-repellent control: bracelets (Mosquito-NO!, Invisaband, Mosquitavert), Cutter Citro Guard candle and Personal Sonic Mosquito Repeller.

The moon’s origin story does not add up. Most scientists think that the moon formed in the earliest days of the solar system, around 4.5 billion years ago, when a Mars-sized protoplanet called Theia whacked into the young Earth. The collision sent debris from both worlds hurling into orbit, where the rubble eventually mingled and combined to form the moon.

If that happened, scientists expect that Theia’s contribution would give the moon a different composition from Earth’s. Yet studies of lunar rocks show that Earth and its moon are compositionally identical. That fact throws a wrench into the planet-on-planet impact narrative.
Researchers have been exploring other scenarios. Maybe the Theia impact never happened (there’s no direct evidence that the budding planet ever existed). Instead of a single colossal collision, scientists have proposed that a string of impacts created miniature moons largely from terrestrial material. Those mini moons merged over time to form one big moon.

“Multiple impacts just make more sense,” says planetary scientist Raluca Rufu of the Weizmann Institute of Science in Rehovot, Israel. “You don’t need this one special impactor to form the moon.”

But Theia shouldn’t be left on the cutting room floor just yet. Earth and Theia were built largely from the same kind of material, new research suggests, and so had similar compositions. There is no sign of “other” material on the moon, this perspective holds, because nothing about Theia was different.

“I’m absolutely on the fence between these two opposing ideas,” says UCLA cosmochemist Edward Young. Determining which story is correct is going to take more research. But the answer will offer profound insights into the evolution of the early solar system, Young says.
The moon is an oddball. Most of the solar system’s moons are way out among the gas giant planets. The only other terrestrial planet with orbiting satellites is Mars. Its moons, Phobos and Deimos, are small, and the prevailing explanation says they were probably asteroids captured by the Red Planet’s gravity. Earth’s moon is too big for that scenario. If the moon had come in from elsewhere, asteroid-like, it would probably have crashed into Earth or pulled off into space. An alternate explanation dating from the 1800s suggested that moon-forming material flew off of a fast-spinning young Earth like children tossed from an out-of-control merry-go-round. That idea fell out of favor, though, when scientists calculated that the spin speeds required were impossibly fast.
In the mid-1970s, planetary scientists proposed the giant-impact hypothesis and the mysterious planet-sized impactor (named Theia in 2000 for the Greek deity who was mother of the moon goddess Selene). The notion made sense given that the early solar system was like a game of cosmic billiards, with giant space rocks frequently colliding.

A 2001 study of lunar rocks collected during the Apollo missions cast doubt on the giant-impact hypothesis. The research showed that the Earth and moon had surprising similarities. To determine a rock’s origin, scientists measure the relative abundance of oxygen isotopes, which act something like finger-prints at a crime scene. Rocks from Earth and its moon, the scientists found, had seemingly identical mixes of oxygen isotopes. That didn’t make sense if much of the moon’s material came from Theia, not Earth. Using impact simulations, Rufu and colleagues recently estimated that the chance of a Theia collision yielding an Earthlike lunar composition is very slim.

Studies of other elements in Apollo rocks, such as titanium and zirconium, also suggest that the Earth and moon originated from the same material. Young and colleagues recently repeated the oxygen isotope measurements with the latest techniques, hunting for even the slightest difference between Earth and the moon. In January 2016, the team published the results in Science. “We measured the oxygen to the highest precision available,” Young says, “and, gosh, the Earth and moon still look identical.”
Some scientists have built simulations of a giant Theia impact that fashion a moon made mostly from terrestrial material. But the scenarios struggle to match the modern positions and movements of the Earth-moon system.

It’s time to think outside the giant-impact box, some scientists argue. Not one but many impacts contributed to the moon’s formation, Rufu and colleagues proposed January 9 in Nature Geoscience. The moon, they say, has an Earthlike composition because most of the material flung into orbit from these impacts came from Earth.

Mini-moon merger
The multi-impact hypothesis was first put forward in 1989, though scientists at the time didn’t have the computer power to run the simulations that could support it. Rufu and colleagues recently revisited the proposal with computer simulations of multiple impactors, each about a hundredth to a tenth of Earth’s mass, smacking into the early Earth.

Any impactors that were direct hits would have transferred lots of energy into the Earth, excavating terrestrial material into space. Debris from each impact combined over centuries to form a small moon, the simulations show. As more impacts rocked Earth over tens of millions of years, more moons formed. Gravity pulled the moons together, combining them. Over roughly 100 million years, according to this scenario, around 20 mini moons ultimately merged to form one mighty moon (SN Online: 1/9/17).
The multimoon explanation yields the right lunar mix in simulations roughly 20 percent of the time, better than the 1 to 2 percent for the giant-impact hypothesis, the researchers note. “The biggest takeaway is that you cannot explain everything with one shot,” Rufu says.

Planetary scientist Robin Canup finds the scenario convincing. “To me, this appears to be a real contender alongside the one big impactor hypothesis,” says Canup, of the Southwest Research Institute in Boulder, Colo.

Don’t discount Theia
But the Theia hypothesis has recently found fresh support. The odds of Theia resembling Earth’s composition enough to yield an Earthlike moon may be a lot higher than originally thought, new chemical analyses suggest. Most of the material that makes up Earth came from the same source as a type of meteorite called enstatite chondrites, planetary scientist Nicolas Dauphas of the University of Chicago reported January 26 in Nature.

Just as with oxygen, the isotopic mix of various other elements in Earth’s rocks serves as a fingerprint of the rocks’ origins. Some of these elements are iron-lovers, such as ruthenium, which quickly sink toward Earth’s iron-rich core (SN: 8/6/16, p. 22). Any ruthenium found close to Earth’s surface, in the mantle, probably arrived late in Earth’s development. Iron-indifferent elements like calcium and titanium don’t sink to the core; they stay in the mantle. Their isotopes record what went into Earth’s assembly over a much longer period of time. By looking at the iron-lovers and iron-indifferent elements together, Dauphas created a timeline of what types of space rocks added to Earth’s mass and when.
A mix of different rocks, including some resembling enstatite chondrite meteorites, supplied the first 60 percent of Earth’s mass, Dauphas says. The remaining balance came almost exclusively from the meteorites’ precursors. In total, around three-quarters of Earth’s mass came from the same material as enstatite chondrites, Dauphas estimates. If Theia formed at around the same distance from the sun as Earth, then it primarily formed from the same material, and consequently had a similar isotopic composition. So if the moon formed largely from Theia, it makes sense that lunar rocks would have a similar composition to Earth, too.
“Most of the problem is solved, in my opinion, if you admit that the great impactor’s material was no different than that of the [early] Earth,” says cosmochemist Marc Javoy at the Institute of Earth Physics of Paris. “It’s the simplest hypothesis” and would mean that the material gobbled up by budding planets in the inner solar system was fairly uniform in composition, offering insight into the arrangement of material that built the solar system.

The notion that Earth is made from the same material as enstatite chondrites “doesn’t make many people happy,” says geochemist Richard Carlson of the Carnegie Institution for Science in Washington, D.C. The isotopes in Earth’s mantle and the meteorites may match, but the relative abundance of the elements themselves do not, Carlson wrote in a commentary in the Jan. 26 Nature. An additional step in the process is needed to explain this compositional mismatch, he says, such as some of the element silicon getting stashed away in Earth’s core.

“What we have now are a lot of new ideas, and now we need to test them,” says Sarah Stewart, a planetary scientist at the University of California, Davis.

One recently proposed test for the moon’s formation is based on temperature, though it seems to be consistent with both origin stories. A new study comparing the moon’s chemistry with glass forged by a nuclear blast suggests that temperatures during or just after the moon’s inception reached a sizzling 1400° Celsius. That means any plausible moon-forming scenario must involve such high temperatures, researchers reported February 8 in Science Advances.
High heat causes rocks to leach light isotopes of zinc. The green-tinged glass forged in the heat of the 1945 Trinity nuclear test in New Mexico lack light isotopes of zinc, says study coauthor and geologist James Day of the Scripps Institution of Oceanography in La Jolla, Calif. The same goes for lunar rocks. Such high temperatures during or just after the moon’s formation fit with the giant-impact hypothesis, he says. But Rufu calculates that her multi-impact hypothesis also yields high enough temperatures.
So maybe temperature can’t resolve the debate, but probing the composition of Earth and the moon’s deep interiors could prove the mini-moon explanation right, says Rufu. Without a single giant collision, the interiors of the two worlds may not have been well mixed, she predicts. Dauphas says that measuring the compositions of other planets could lend credence to his Earthlike Theia proposal. Mercury and Venus would also have formed largely from the same kind of material as Earth and therefore also have Earthlike compositions, he says. Future studies of the solar system’s inhabitants could confirm or rule out these predictions, but that will require a new chapter of exploration.

WASHINGTON — A second cancer later in life is common for childhood cancer survivors, and scientists now have a sense of the role genes play when this happens. A project that mined the genetic data of a group of survivors finds that 11.5 percent carry mutations that increase the risk of a subsequent cancer.

“We’ve always known that among survivors, a certain population will experience adverse outcomes directly related to therapy,” says epidemiologist and team member Leslie Robison of St. Jude Children’s Research Hospital in Memphis. The project sought “to find out what contribution genetics may play.” The team presented their work at the American Association of Cancer Research meeting April 3.
“This is a nice first step,” says David Malkin, a pediatric oncologist at the University of Toronto. “The results validate the thoughts of those of us who believe there is a genetic risk that increases the risk of second malignancies.”

Five-year survival rates for kids with cancer have grown to more than 80 percent. But “there are long-term consequences for having been diagnosed and treated for cancer as a child,” notes Robison. Some survivors develop a later, second cancer due to the radiation or chemotherapy that treated the first cancer (SN: 3/10/07, p. 157).

The researchers examined 3,007 survivors of pediatric cancer who routinely undergo medical evaluation at St. Jude. About a third had leukemia as children. By age 45, 29 percent of this group had developed new tumors, often in the skin, breast or thyroid.

The team cataloged each survivor’s DNA, and looked closely at 156 genes known as cancer predisposition genes. Of the survivors, 11.5 percent carried a problematic mutation in one of the 156 genes. Some genes on the list convey a higher risk than others, so the team looked further at a subset of 60 genes in which only one mutated copy in each cell is enough to cause disease. These 60 genes also have high penetrance, meaning that a mutated copy is highly likely to lead to a cancer. Nearly 6 percent of the survivors had a problematic mutation in one of these 60 genes.

The research team also separated the survivors based on whether or not they had received radiation therapy as children. Close to 17 percent of survivors not exposed to radiation therapy had a problematic mutation in the subset of 60 genes. These survivors had an increased risk for any second cancer. Those with both a mutation in one of the 60 genes and radiation in their treatment history had a higher risk for specific kinds of second cancers: breast, thyroid or sarcomas, tumors in connective tissues.
Based on the new estimates of genetic risk, the team suggests that survivors not given radiation therapy undergo genetic counseling if a second cancer develops. Counseling is also recommended “for survivors who develop a secondary breast cancer, thyroid cancer or sarcoma in a site that received prior radiation therapy,” says St. Jude epidemiologist and project team member Carmen Wilson. Counseling can provide guidance on health practices going forward, reproductive choices and the implications for immediate family members who may have inherited the mutation, notes Robison.

The extensive amount of medical and genomic information collected for the survivors could help with cancer prevention efforts in the future, Robison says. The team would like to create prediction models that consider treatment, genetics and other clinical information, in order to place survivors into different risk groups. “It’s eventually going to have clear implications for how these patients are clinically managed, and how we either prevent or ameliorate the adverse effects,” Robison says.

Malkin notes that not only “what you got for treatment, but when you got it” is another factor influencing a survivor’s risk profile for second cancers, as treatments and doses have changed over time. He also thinks the percentage of survivors at risk reported by Robison’s team is lower than expected. “Expanding the pool of genes to look at will be very informative,” he says.

A recent upsurge in planet-warming methane may not be caused by increasing emissions, as previously thought, but by methane lingering longer in the atmosphere.

That’s the conclusion of two independent studies that indirectly tracked concentrations of hydroxyl, a highly reactive chemical that rips methane molecules apart. Hydroxyl levels in the atmosphere decreased roughly 7 or 8 percent starting in the early 2000s, the studies estimate.

The two teams propose that the hydroxyl decline slowed the breakdown of atmospheric methane, boosting levels of the greenhouse gas. Concentrations in the atmosphere have crept up since 2007, but during the same period, methane emissions from human activities and natural sources have remained stable or even fallen slightly, both studies suggest. The research groups report their findings online April 17 in Proceedings of the National Academy of Sciences.
“If hydroxyl were to decline long-term, then it would be bad news,” says Matt Rigby, an atmospheric scientist at the University of Bristol in England who coauthored one of the studies. Less methane would be removed from the atmosphere, he says, so the gas would hang around longer and cause more warming.

The stability of methane emissions might also vindicate previous studies that found no rise in emissions. The Environmental Protection Agency, for instance, has reported that U.S. emissions remained largely unchanged from 2004 to 2014 (SN Online: 4/14/16).

Methane enters the atmosphere from a range of sources, from decomposing biological material in wetlands to leaks in natural gas pipelines. Ton for ton, that methane causes 28 to 36 times as much warming as carbon dioxide over a century.

Since the start of the Industrial Revolution, atmospheric methane concentrations have more than doubled. By the early 2000s, though, levels of the greenhouse gas inexplicably flatlined. In 2007, methane levels just as mysteriously began rising again. The lull and subsequent upswing puzzled scientists, with explanations ranging from the abundance of methane-producing microbes to the collapse of the Soviet Union.

Those proposals didn’t account for what happens once methane enters the atmosphere. Most methane molecules in the air last around a decade before being broken apart during chemical reactions with hydroxyl. Monitoring methane-destroying hydroxyl is tricky, though, because the molecules are so reactive that they survive for less than a second after formation before undergoing a chemical reaction.
Neither study can show conclusively that hydroxyl levels changed, notes Stefan Schwietzke, an atmospheric scientist at the National Oceanic and Atmospheric Administration’s Earth System Research Laboratory in Boulder, Colo. The papers nevertheless add a new twist in explaining the mysterious methane rise, he says. “Basically these studies are opening a new can of worms, and there was no shortage of worms.”

Despite being conducted by two separate teams — one headed by Rigby and the other by atmospheric scientist Alex Turner of Harvard University — the new studies used the same roundabout approach to tracking hydroxyl concentrations over time.

Both teams followed methyl chloroform, an ozone-depleting substance used as a solvent before being banned by the Montreal Protocol. Like methane, methyl chloroform also breaks apart in reactions with hydroxyl. Unlike methane, though, emission rates of methyl chloroform are fairly easy to track because the chemical is entirely human-made.

Examining methyl chloroform measurements gathered since the 1980s revealed that hydroxyl concentrations have probably wobbled over time, contributing to the odd pause and rise in atmospheric methane concentrations. But to know for sure whether hydroxyl levels varied or remained steady, scientists will need to take a more detailed look at regional emissions of methane and methyl chloroform, Rigby says.

Why hydroxyl levels might have fallen also remains unclear. Turner and colleagues note that the ban on ozone-depleting substances like methyl chloroform might be the cause. The now-recovering ozone layer (SN: 12/24/16, p. 28) blocks some ultraviolet light, an important ingredient in the formation of hydroxyl. Identifying the cause of the hydroxyl changes could help climate scientists better predict how methane levels will behave in the future.

NEW ORLEANS — A relatively small brain can pack a big evolutionary punch. Consider Homo naledi, a famously puzzling fossil species in the human genus. Despite having a brain only slightly larger than a chimpanzee’s, H. naledi displays key humanlike neural features, two anthropologists reported April 20 at the annual meeting of the American Association of Physical Anthropologists.

Those brain characteristics include a region corresponding to Broca’s area, which spans parts of the right and left sides of the brain in present-day people. The left side is typically involved in speech and language.
“It looks like Homo naledi’s brain evolved a huge amount of shape change that supported social emotions and advanced communication of some type,” said Shawn Hurst of Indiana University Bloomington, who presented the new findings. “We can’t say for sure whether that included language.” Frontal brain locations near Broca’s area contribute to social emotions such as empathy, pride and shame. As interactions within groups became more complex in ancient Homo species, neural capacities for experiencing social emotions and communicating verbally blossomed, Hurst suspects.

Scientists don’t know how long ago H. naledi inhabited Africa’s southern tip. If H. naledi lived 2 million or even 900,000 years ago, as some researchers have suggested (SN: 8/6/16, p. 12), humanlike brains with a language-related area would be shocking. A capacity for language is thought to have emerged in Homo over the last few hundred thousand years at most.

Discoverers of H. naledi, led by anthropologist Lee Berger of the University of the Witwatersrand in Johannesburg, will announce an estimated age for the species and describe new fossil finds within the next few weeks, Hurst said.

Hurst and Ralph Holloway of Columbia University led a team that laser scanned the inside surfaces of several partial H. naledi skulls to create virtual casts, or endocasts, of brain surfaces. An endocast reproduces the shape and, with varying success, details of the surface of the brain that were imprinted on the walls of the braincase while an individual was alive. Such brain impressions are not always clear, which has sparked debate over how to interpret them.

Two grooves identified on an endocast from a partial H. naledi skull frame the language-related section of Broca’s area in humans today, Hurst said. H. naledi’s brain also possessed folds of tissue that largely covered a surface section where the grooves converged. Similar folds of tissue typically cover the surface of Broca’s area in modern human brains.
The general shape of that part of the frontal brain in humans differs greatly from that of living apes and fossil hominids dating to at least 700,000 to 1 million years years ago, Hurst added.

H. naledi also displays a humanlike pattern of surface features at the back of the brain, although to a lesser extent than at the brain’s front, Holloway said. Endocasts for this analysis came from two other partial H. naledi skulls.

Specific protrusions and other features at the back of H. naledi’s brain are more pronounced on the left side, Holloway said. In people today, the same left-sided bias in brain organization is associated with right-handedness.

In the past, Holloway and anthropologist Dean Falk of Florida State University in Tallahassee have sharply disagreed over how to identify neural features on fossil endocasts, including a key groove in tissue at the back of the brain. After hearing Hurst and Holloway’s presentations, Falk expressed doubt that H. naledi’s brain was as humanlike as they concluded.

Shortly after the presentations, Hurst and Falk hashed out their differences head-to-head as they jointly studied a solid cast of the partial H. naledi brain surface displaying proposed signs of Broca’s area. They agreed on much about the fossil species’ neural setup, with one major exception. “I’m skeptical that two frontal [grooves] frame an area that corresponds to Broca’s area,” Falk said. If she’s right, then H. naledi communicated much less like present-day people than proposed by Hurst. Falk plans to study the new endocasts more closely and compare them with endocasts of other fossil hominids.

The benefits of statins for people older than 75 remain unclear, a new analysis finds. Statins did not reduce heart attacks or coronary heart disease deaths, nor did they reduce deaths from any cause, compared with people not taking statins, researchers report online May 22 in JAMA Internal Medicine.

Recently published guidelines cited insufficient data to recommend statins for people older than age 75 who don’t have a history of cardiovascular disease. The new analysis considered a subset of older adults enrolled in a study of heart attack prevention and mortality conducted from 1994 to 2002. The sample included 2,867 adults ages 65 and older who had hypertension, 1,467 of whom took a statin.

There was no meaningful difference in the frequency of heart attacks or coronary heart disease deaths between those who took statins and those who did not. There was also no significant difference in deaths from any cause, both overall and among participants ages 65 to 74 or those 75 and older.

Statin use may be associated with muscle damage and fatigue, which could especially impact older adults and put them at higher risk for physical decline, the authors say.

Ancient rock artists were drawn to echo chambers. Members of early farming communities in Europe painted images in rock-shelters where sounds bounced off walls and into the surrounding countryside, researchers say.

Rock-shelters lacking such sound effects were passed up, at least in the central Mediterranean, report archaeologist Margarita Díaz-Andreu of the University of Barcelona and colleagues in the July Journal of Archaeological Science. In landscapes with many potential rock art sites, “the few shelters chosen to be painted were those that have special acoustic properties,” Díaz-Andreu says.
Some hunter-gatherer and farming groups studied over the past couple centuries believed in spirits that dwell in rock and reveal their presence via echoes. But acoustic evidence of special echoing properties at rock art sites is rare.

Díaz-Andreu’s team studied two rock art sites in 2015 and 2016. Baume Brune is a kilometer-long cliff in southeastern France. Of 43 naturally formed cavities in the cliff, only eight contain paintings, which include treelike figures and horned animals. Rock art in the Valle d’Ividoro, on Italy’s east coast, appears in an 800-meter-long section of a gorge. Only three of its 11 natural shelters contain painted images. Researchers generally date these French and Italian paintings to between roughly 6,500 and 5,000 years ago, several thousand years after the Stone Age had ended, Díaz-Andreu says.

To investigate the acoustics of the decorated and unadorned shelters, the researchers developed a new technique for determining the direction, intensity and timing of sound waves arriving at a particular point from every direction. A special microphone connected to a digital recorder measured the acoustic properties of any echoes set off by balloons popped just outside each rock-shelter. This setup was moved to various spots outside the caves to record the acoustic reach of reflected popping sounds. Echo measurements in France were taken at distances ranging from 22 to 36 meters from cliff shelters. Due to rougher terrain in Italy, measurements there were taken at distances ranging from 77 to 300 meters.
Then, the acoustic data were transformed into 3-D, slow-motion depictions of echoes, represented by moving circles, indicating where sound reflections originated. At both sites, shelters with rock paintings displayed better echoing properties than undecorated shelters, Díaz-Andreu says. And in each location, the shelter that best reflected echoes had the highest number of paintings.
“This novel technique shows a clear correlation between audible echoes and decorated shelters,” says music archaeologist Riitta Rainio of the University of Helsinki in Finland, who did not participate in the new study.

Echoes that bounce off steep rock cliffs bordering three lakes in northern Finland also attracted ancient artists, Rainio says. She and her colleagues took acoustic measurements at Finland’s painted cliffs from 2013 to 2016. Microphones placed on boats positioned at different spots on nearby lakes measured sound waves generated, in most trials, by a starter’s pistol. These Finnish paintings date to between around 7,200 and 3,000 years ago, Rainio says.

In some cases, echoes reflect directly from cliff paintings. “That, and possible drumming figures painted on the cliffs, suggest that sound played some role in rituals at these sites,” Rainio says. Her team will report its findings in an upcoming Journal of Archaeological Method and Theory.

Creators of older Stone Age cave art also appear to have focused on sites where echoes abounded, says archaeologist Paul Pettitt of Durham University in England. For instance, many roughly 14,000- to 12,000-year-old animal drawings and engravings at France’s Niaux Cave cluster in a cathedral-like chamber where sounds echo loudly.

“The new study provides convincing evidence that echoes, which were scientifically inexplicable to prehistoric people, played a determining role in how art was created,” Pettitt says.

Big, rogue planets — ones without parent stars — are rare.

A new census of free-floating Jupiter-mass planets determined that these worlds are a tenth as common as previous estimates suggested. The results appear online July 24 in Nature.

Planets can go rogue in two ways: They can get kicked out of their parent planetary systems or form when a ball of gas and dust collapses (SN: 4/4/15, p. 22).

In the new study, Przemek Mróz of the Astronomical Observatory of the University of Warsaw and colleagues estimated the number of large, rogue planets in our galaxy using a technique called microlensing. When an object with a mass of a planet passes in front of a distant, background star, the gravity of the planet acts as a gravitational magnifying glass. It distorts and focuses the light, giving up the planet’s existence.
Mróz and colleagues looked at 2,617 microlensing events recorded between 2010 and 2015 and determined which were caused by a rogue planet. For every typical star, called main sequence stars, there are 0.25 free-floating Jupiter-mass planets, the analysis suggests.

The new result sharply contrasts an estimate published in 2011, which suggested that rogue Jupiters are almost twice as common as main sequence stars. About 90 percent of stars in the universe are main sequence stars, so if that estimate were accurate, there should be a lot of free-floating Jupiters.

“That result changed our conceptual framework of the universe just a little bit,” says astronomer Michael Liu of the University of Hawaii in Honolulu. It challenged long-held ideas about how planets go rogue because the known methods wouldn’t generate enough planets to account for all the wanderers.

The 2011 result was based on a relatively small sample of microlensing events, only 474. Since then, infrared telescope images haven’t detected as many free-floating planets as expected. “Over the years, serious doubts were cast over the claims of a large population of Jupiter-mass free-floaters,” Mróz says.

David Bennett, coauthor of the 2011 study, agrees that the new census failed to find evidence for a large population of Jupiter-mass rogue planets. He notes, however, that the new data do reveal four times as many Jupiter-mass failed stars called brown dwarfs than predicted in the original census. So some of the rogues that were originally classified as planets may, in fact, be failed stars. Bennett, of NASA’s Goddard Space Flight Center in Greenbelt, Md., and colleagues are working on a new analysis of potential rogues with nearly 3,000 microlensing events and plan to compare their results with those from the new census.
Liu says the latest census is much more in line with theories of how planets form. Most rogues should be Earth-mass or a little heavier. Those lighter planets get tossed out of their planetary systems much easier than behemoths like Jupiter. Still, the smaller planets are harder to detect.

The new microlensing analysis did identify several events in which stars brightened and dimmed in less than half a day. Such short events hint at the existence of Earth-mass free-floaters because smaller planets with less gravity should brighten a distant star more briefly than more massive stars. Determining whether those small planets are really rogue and counting how many there are will take better telescopes, the team notes.