直接上原文。

BEGINNING IN 1905, investigations into the behavior of light got positively spooky. That year, Einstein published his special theory of relativity, in which he ratcheted up M & M’s null result to an audacious level. The speed of light in empty space, he declared, is a universal constant, no matter the speed of the light-emitting source or the speed of the person doing the measuring.

　　What if Einstein is right? For one thing, if you’re in a spacecraft traveling at half the speed of light and you shine a light beam straight ahead of the spacecraft, you and I and everybody else in the universe who measures the beam’s speed will find it to be 186,282 miles per second. Not only that, even if you shine the light out the back, top, or sides of your spacecraft, we will all continue to measure the same speed.

　　Odd.

　　Common sense says that if you fire a bullet straight ahead from the front of a moving train, the bullet’s ground speed is the speed of the bullet plus the speed of the train. And if you fire the bullet straight backward from the back of the train, the bullet’s ground speed will be its own minus that of the train. All that is true for bullets, but not, according to Einstein, for light.

　　Einstein was right, of course, and the implications are staggering. If everyone, everywhere and at all times, is to measure the same speed for the beam from your imaginary spacecraft, a number of things have to happen. First of all, as the speed of your spacecraft increases, the length of everything—you, your measuring devices, your spacecraft—shortens in the direction of motion, as seen by everyone else. Furthermore, your own time slows down exactly enough so that when you haul out your newly shortened yardstick, you are guaranteed to be duped into measuring the same old constant value for the speed of light. What we have here is a cosmic conspiracy of the highest order.

　　IMPROVED METHODS OF measuring soon added decimal place upon decimal place to the speed of light. Indeed, physicists got so good at the game that they eventually dealt themselves out of it.

　　Units of speed always combine units of length and time—50 miles per hour, for instance, or 800 meters per second. When Einstein began his work on special relativity, the definition of the second was coming along nicely, but definitions of the meter were completely clunky. As of 1791, the meter was defined as one ten-millionth the distance from the North Pole to the equator along the line of longitude that passes through Paris. And after earlier efforts to make this work, in 1889 the meter was redefined as the length of a prototype bar made of platinum-iridium alloy, stored at the International Bureau of Weights and Measures in Sèvres, France, and measured at the temperature at which ice melts. In 1960, the basis for defining the meter shifted again, and the exactitude increased further: 1,650,763.73 wavelengths, in a vacuum, of light emitted by the unperturbed atomic energy-level transition 2p10 to 5d5 of the krypton-86 isotope. Obvious, when you think about it.

　　Eventually it became clear to all concerned that the speed of light could be measured far more precisely than could the length of the meter. So in 1983 the General Conference on Weights and Measures decided to define—not measure, but define—the speed of light at the latest, best value: 299,792,458 meters per second. In other words, the definition of the meter was now forced into units of the speed of light, turning the meter into exactly 1/299,792,458 of the distance light travels in one second in a vacuum. And so tomorrow, anyone who measures the speed of light even more precisely than the 1983 value will be adjusting the length of the meter, not the speed of light itself.

　　Don’t worry, though. Any refinements in the speed of light will be too small to show up in your school ruler. If you’re an average European guy, you’ll still be slightly less than 1.8 meters tall. And if you’re an American, you’ll still be getting the same bad gas mileage in your SUV. （一黑黑倆地兒）

　　THE SPEED OF LIGHT may be astrophysically sacred, but it’s not immutable. In all transparent substances—air, water, glass, and especially diamonds—light travels more slowly than it does in a vacuum.

　　But the speed of light in a vacuum is a constant, and for a quantity to be truly constant it must remain unchanged, regardless of how, when, where, or why it is measured. The light-speed police take nothing for granted, though, and in the past several years they have sought evidence of change in the 13.7 billion years since the big bang. In particular, they’ve been measuring the so-called fine-structure constant, which is a combination of the speed of light in a vacuum and several other physical constants, including Planck’s constant, pi, and the charge of an electron.

　　This derived constant is a measure of the small shifts in the energy levels of atoms, which affect the spectra of stars and galaxies. Since the universe is a giant time machine, in which one can see the distant past by looking at distant objects, any change in the value of the fine-structure constant with time would reveal itself in observations of the cosmos. For cogent reasons, physicists don’t expect Planck’s constant or the charge of an electron to vary, and pi will certainly keep its value—which leaves only the speed of light to blame if discrepancies arise.

　　One of the ways astrophysicists calculate the age of the universe assumes that the speed of light has always been the same, so a variation in the speed of light anywhere in the cosmos is not just of passing interest. But as of January 2006, physicists’ measurements show no evidence for a change in the fine-structure constant across time or across space.

呃，我不知道該說(shuō)什么，又似乎有很多想說(shuō)的。好吧，最想說(shuō)的還是：科學(xué)家還真是一種簡(jiǎn)單粗暴的生物啊……