Physicists are getting worked up about something that should have been settled long ago: Just how massive is a kilo­gram? Most units of scientific measure are now defined not by physical objects but by universal constants. A meter, for example, is the distance traveled by light in a vacuum in 1/299,792,458 of a second. The kilogram is a holdout, still defined by a cylinder of platinum-iridium kept in a vault near Paris since 1889. And the cylinder is losing weight—a grain of salt’s worth so far—demonstrating the need for a unit based on a physical constant.

Two major standardization strategies have surfaced. One proposes a numerical unit built on Avogadro’s number, derived from the number of atoms in 12 grams of carbon-12. The other approach makes use of Planck’s constant and the Watt balance, which measures the electromagnetic force needed to hold a kilogram.

For Richard Davis at the official bureau for measurements in France, there is no clear win­ner; he believes both methods are too pricey for most labs. But until researchers find a simple, reproducible means, we’re stuck with a hunk of platinum and iridium that is ever-so-slowly getting lighter.