The only way to see anything is for our eyes to intercept visible radiation—light waves—bouncing off it. That makes the recipe for invisibility obvious: just make the object absorb all the light that falls on it. Last year, for instance, scientists at Duke University manufactured a cloak made out of metamaterials, which are artificial composites that interact with electromagnetic radiation in a certain way. In that case, copper rings and wires on sheets of fiberglass deflected microwaves, causing the cloak to be invisible to any being whose eyes are sensitive to microwaves rather than visible waves. Okay, not exactly what J.K. Rowling had in mind, but definitely a step in the right direction.

Now Muggle physicists have taken one more step toward reproducing Harry Potter’s invisibility cloak, this time with visible light waves. In an upcoming issue of the journal Physical Review Letters, physicists Zhichao Ruan, Min Yan, Curtis W. Neff and Min Qiu of Sweden’s Royal Institute of Technology and China’s Zhejiang University show that a cloak made of metamaterials and shaped into a column could make the Muggle, witch or or wizard inside completely invisible.

The key is that the metamaterials force light to follow a particular path. (That’s also the reason your hair looks darker when wet: the water molecules make light bounce around within the hair, with the result that fewer escape to reach your eye; less light equals darker object.) “When electromagnetic waves pass through the invisibility cloak, the cloak will deflect the waves, guide them around the object, and return them to the original propagation direction without perturbing the exterior field,” the physicists write. Done right, the column (for complicated reasons, a cylindrical shape is better for invisibility than, say, a flat rectangle) would guide light around the inner chamber, preventing it from ever reaching anything—including an eye or camera—beyond. But light scattered from objects behind the cloak might also be guided around it, meaning that when you looked at the cloak you would see whatever was behind it, just as with Harry Potter’s. As the authors say, “a cloak with the ideal material parameters is a perfect invisibility cloak.”

The next time some grumpy (and uninformed) curmudgeon tells you that the only way for the world to reduce its emissions of greenhouse gases is to drink warm beer (give up refrigeration), freeze in the dark (take electricity and heat conservation to an extreme) and drive dangerous little tin cans (have vehicles meet fuel-economy standards not through smart engineering but by downsizing cars), offer them two words: gas flaring.

When petroleum is pumped out of the ground, some natural gas comes with it. Typically, it’s burned at the wellhead, a process called gas flaring. Countries report how much flaring goes on within their borders, but the World Bank, taking a cue from Ronald Reagan, has gone the “trust but verify” route: it asked the U.S. National Oceanic and Atmospheric Administration to use the military weather satellites at its disposal to peer down at oil wells from 400 miles up in space to measure who is really burning how much. Two more words: Russia lies.

It has become a mantra of the war on cancer: early detection saves lives. Mammograms, colonoscopies, Pap tests, prostate tests—you name it, and if you undergo the screening test often enough your risk of dying of cancer decreases.

If only there were scientific support for that claim. But studies concluding just the opposite keep arriving like a drumbeat of bad news. In the latest, researchers find that a shorter time between screenings for prostate cancer do not reduce how many aggressive tumors are found—even though more prostate cancers are detected among men screened every two years than men screened every four years. The study is published online today in the Journal of the National Cancer Institute.

There is nothing more intuitive than the idea that cancer screening leads to early cancer detection, and that earlier detection leads to longer survival. Surely the earlier a cancer is caught, the greater the chance that you’ll be alive in five years, the benchmark for "survivorship."

But the effect of cancer screening on reducing cancer mortality remains unproved. And the evidence keeps going in the other direction. To be sure, some tumors discovered through screening would, if left untreated, have killed patients; some of the improvement in breast-cancer survival rates reflect earlier detection. But many other tumors are so slow-growing that detecting them “early” makes no difference to survival; catch them now or in 10 years, and modern chemotherapy and radiation can still wipe them out before they metastasize (it is metastasis that kills some 90 percent of cancer patients, not the primary tumor). The incidence of prostate cancer has soared since the prostate specific antigen, or PSA, test was introduced in the 1980s, for instance, but mortality has hardly budged: the test is picking up tumors that would have posed little mortality risk even if they had been detected only after symptoms appeared. Men may think the test saved them, and give moving testimonials to that effect, but there’s no scientific proof of that. Just the opposite. As a 2006 study concluded, "results do not suggest that screening with PSA . . . is effective in reducing mortality" from prostate cancer.

In the latest study, scientists led by Monique Roobol of Erasmus Medical Centre in Rotterdam, The Netherlands, compared men who were screened bi-annually for prostate cancer (using the standard PSA test) to those who were screened every four years. In particular, the scientists examined so-called interval prostate cancers, those discovered between screenings when the men developed symptoms of prostate cancer before it was time for their next PSA test. Over ten years, there was no statistically significant difference in the total number of interval cancers, or in the number of aggressive interval cancers, in the two groups. That means more-frequent screening did not reduce the number of interval cancers. And larger, more worrisome cancers can be found equally well whether screening occurs every two years or every four.

Government weather forecasters love El Nino. For years they have used this fluctuating current in the Pacific Ocean (in El Nino years, sea surface temperatures in the western tropical Pacific rise, while in La Nina years they fall) to forecast winter weather months in advance. In El Nino years, they basically expect the northeast to be a little warmer than the historical average, the southern tier to be colder, and the southeast to be snowier. La Nina years are supposed to bring balmier winters to almost all the continental U.S.

In fact, government forecasters love El Nino so much that even when a better basis for seasonal prediction comes along, they stick with the kid.

Since 1999, scientists led by meteorologist Judah Cohen, director of seasonal forecasting at AER Inc., of Lexington, Mass., have been gathering more and more evidence that there’s a better way. As he and physicist Christopher Fletcher of the University of Toronto report in the August issue of the Journal of Climate, the predictive power of El Nino, at least outside the tropics where its effects are directly felt, can’t hold a candle to an alternative: using the amount of snowcover in October in Eurasia and, especially, Siberia to predict upcoming winter temperatures and snowfall for the high- and mid-latitudes of the Northern Hemisphere—basically, the eastern U.S. and northern Europe and Asia. They call their predictive model sCast, where the "s" stands for snow.

If you figure that a good time to ask the boss for a raise is when she gets back from summer vacation in a few days, you’ve got it half right. Being in a good mood influences people’s judgment and decision making, meaning you have a greater chance of coming away with that extra five percent when the boss is in a good mood than when she has just been chewed out by the uber-boss. But here’s the other half: if the boss knows that you are trying to exploit her “incidental affect,” as psychologists call it, and is therefore reminded that her judgment may be clouded by incidental feelings, the effect may disappear.

So conclude Eduardo B. Andrade and Teck-Hua Ho of the Haas School of Business at the University of California, Berkeley, who tested how mood affects decision making—and how knowing that someone is trying to exploit that effect for his own ends can make the attempt backfire. They report their findings in the August issue of Psychological Science.

The world is full of examples of mood affecting judgment on issues that “should” turn on logic and rationality alone. When people are negotiating, being in a good mood makes it more likely that they’ll cooperate, while being in a bad mood makes it more likely they’ll be more competitive than cooperative. Let’s say you’re negotiating with someone—haggling over the price of a used car, say—and the seller responds, “that offer is enormously insulting!” Knowing the effect of anger on decision making, and really wanting the car, you backpedal, upping your offer. As Andrade and Ho write, “the angry feeling generated by a disliked offer is taken into account during the negotiation process.” So is a contented feeling: people feel they can ask more of someone in a good mood.

But what if the mood has nothing to do with the subject at hand—settling on a price for a car—and instead reflects something completely incidental, such as the boss's euphoria about how great his vacation was? In this case, the contentment has nothing to do with your request for a five percent raise. Your best strategy is nevertheless to ask for more when you know the boss is in a good mood, and less if he is in a bad mood.

But here’s where it gets tricky. If he knows that you know he is in a good mood, he is likely to take this into account, figure you’re trying to exploit his good humor, and overreact, telling you and your request for a raise to take a hike.

The “God of the gaps” just got squeezed out of yet another dwelling place.

This centuries-old argument for the existence of God basically goes like this. There is no way to explain (fill in the blank: lightning . . . volcanoes . . . the creation of the sun and planets . . . the origin of species . . . ) without invoking a supernatural hand; science alone falls short. Hence there must be a God. The trouble, of course, is that science has this annoying habit of eventually accounting for such mysteries through natural, not supernatural, processes. And that leaves God one less explanatory gap to inhabit.

A current favorite refuge for those who deny that mutation and natural selection (also known as Darwinian evolution) are enough to explain the diversity of life on Earth is that highly-complicated biological structures cannot have been produced natrually. It’s basically a mouse-trap argument. A complicated structure such as, say, a receptor on a cell surface, is made of numerous components, just as a mouse trap is. No single component is of much use, just as the wooden base of a mouse trap is not much use without the bar and the spring and the other components. Since evolution confers survival only on useful components, continues the argument, it strains credulity to think that the individual parts of a receptor would have just coincidentally emerged from mutations at exactly the same time to come together in a useful structure. More likely, each component emerged, found nothing to do, and was eliminated by natural selection. Want a receptor? Ask God to make it.

Not so, say scientists who have determined, for the first time, how an ancient protein evolved step by step, refuting even more strongly a key element of the “intelligent design” argument.

In 2006 biologists reconstructed a receptor that last existed on earth 450 million years ago. Now the researchers have determined the structure of this recovered ancestor and figured out that specific mutations that, step by step, produced the modern-day glucocorticoid receptor, which allows cells to respond to the stress-hormone cortisol and therefore regulates stress. Says Joe Thornton, an evolutionary biologist at the University of Oregon who led both last year’s project and this one, “We were able to see the precise mechanisms by which evolution molded a tiny molecular machine at the atomic level, and to reconstruct the order of events by which history unfolded.” The study is being published today online in Science Express, the rapid-communication version of the journal Science.

The scientists found that just seven mutations changed the ancestral receptor gene into the gene that makes today’s glucocorticoid receptor. Some mutations would have wrecked the protein unless other mutations were in place first; this allowed the scientists to infer the sequence in which the mutations occurred. One crucial mutation resculpted a big section of the protein in a way that re-positioned a cluster of atoms; a second mutation in this region then tightened up the receptor’s embrace of cortisol. Other mutations buttressed other regions of the protein so they could withstand this remodeling. “Permissive mutations stabilized specific structural elements, allowing them to tolerate later destabilizing mutations that conferred a new function,” the authors write.

The result is “an unprecedentedly detailed view into the deep past of our genes,” Thornton told me. The discovery “provides the first fully mechanistic account of how a gene [for the ancestral receptor] evolved a new function [for the modern-day glucocorticoid receptor]. By comparing the ancestral structure to the modern structures that derived from it, we were able to see the mechanisms by which present-day receptors evolved.”

Thornton is not shy about his paper’s relevance to the unending evolution wars. The discovery, he says, “shows precisely how a complex phenotype evolved through a sequence of neutral and adaptive substitutions,” including those permissive mutations that do not themselves confer a selective advantage but pave the way for later mutations that do. A new book by the intelligent-design proponent Michael Behe argues that complex structures cannot evolve if they require two or more mutations. The new discovery shows that permissive mutations make this multi-step evolution perfectly possible—and in the case of the glucocorticoid receptor, a fact.

Chimpanzees and other great apes get all the good press when it comes to the intelligence of non-human animals. Chimps use tools and teach their offspring to do so, for instance, whether it's twirling a stick into a termite nest to extract hors d'oeuvres or selecting just the right rock to crack open a tough nut, as researchers keep discovering.

Scientists are reporting today that New Caledonian crows are no slouches when it comes to sophisticated tool use. As they report online in the journal Current Biology, the crows can spontaneously combine two tools to get a snack--and in a way that suggests they solved the problem not by trial and error, but by reasoning through analogy. That is, the crows were able to see that a novel situation was essentially the same as one they'd encountered before.

"Evidence suggests that, from the earliest human stone tools, analogical reasoning has been at the core of human innovation," Russell Gray of the University of Auckland, who led the study, said in a statement. "This hallmark of human intelligence may also be at work in both the great apes and New Caledonian crows."

In the study, the researchers put food in a hole that was unreachable unaided. They also left a stick lying around, but it was also too short to reach the food. They left one more prop: a long stick in a "toolbox." This stick was long enough to reach the food--but it, too, was out of reach. No problem. The crows used the short stick to get the long tool out of the box, then used the long stick to get the food. In fact, six out of seven crows immediately tried to get the long stick with the short stick; only one dunce did the bird-brained thing of trying to get the food directly with the short stick and realizing that wouldn't work, and that he had to use the short stick to get the long stick.

Next challenge: food still at the bottom of a deep hole, small stick inside the toolbox and long stick within the crows' reach.

Those who deny the reality of human-induced climate warning should mark their calendars. Today, climatologists reported a new prediction from their computer-based model of the planet’s climate. After folding into the model precise information about the state of the ocean and the atmosphere rather than the approximate parameters other models use, they find that global warming will slow during the next few years.

Just to emphasize: even as greenhouse gases accumulate in the atmosphere, other influences will keep the world from continuing on the upward temperature path it’s been on since the 1990s. No one ever said that the greenhouse effect is the only thing driving our climate, and from time to time other factors will outweigh it. We are moving into one of those times, say the British climatologists. So when 2008, say, turns out not to set any temperature records even though 11 of the last 12 years have done so, can we please not have press conferences and headlines and op-eds and Senate floor speeches saying, “see? That whole global warming thing was a hoax.” To the contrary. The same climate models that foresee the changes chronicled in the IPCC reports also say that we’re in for a few years of reprieve from the almost unrelenting temperature rise.

What with killer spinach and poisoned Chinese toothpaste and dog food, the risks of environmental chemicals have faded into the background lately. But one group of scientists convened by the government thinks that's a mistake, at least when it comes to  a chemical called bisphenol-A. A building block of certain plastics, it is produced in massive amounts (6 billion pounds per year), and is used in the linings of some food cans, in certain plastic water bottles, baby bottles and other products. Traces of bisphenol-A are found in rivers, oceans and wildlife, in many foods, and in our blood.

Figuring it might be a good idea to know what exactly we're dealing with, given bisphenol-A's ubiquity, the National Institute of Environmental Health Sciences (part of the National Institutes of Health) convened 38 experts from around the world to assess the risks of the chemical. (This was the first time NIEHS has undertaken such a comprehensive review of the potential harm from an environmental contaminant.) The result falls somewhat short of reassuring. After examining over 700 published studies, the scientists concluded that the average level of the chemical in Americans’ blood is very high, far above levels associated with adverse effects in rats and mice, and also well above the government's "safe" level. The trouble with having so much of the chemical sloshing around us is that bisphenol-A acts like estrogen, even minuscule amounts of which can have powerful physiological effects.

Just as you'd expect from a hormone-like compound, very low doses of bisphenol-A can adversely affect the reproductive, nervous, endocrine and immune systems in rodents, especially when exposure occurs during fetal development and infancy. For example, in a study just published in Reproductive Toxicology (where the NIEHS panel's report is also published), the female offspring of rats exposed to bisphenol-A shortly before they gave birth developed uterine fibroids, ovarian cysts, and changes in the uterine lining analogous to endometriosis.

What about people?