To the left of gold in the periodic table are two elements — platinum and iridium — that are just as interesting in many ways. Like gold, platinum is malleable and ductile and widely used as a catalyst, whereas iridium is hard and brittle and extremely resistant to corrosion. And in a vault in a metrology laboratory in Sèvres, a suburb of Paris, is a cylinder made of a platinum–iridium alloy that has a mass of exactly one kilogram because, in the International System of Units (SI), the kilogram is defined as being equal to the mass of this International Prototype Kilogram.

There are seven base units in SI and the kilogram is the only one that is defined by an artefact, which is prone to change with time, rather than by a fundamental physical property, which should not change with time and can also be reproduced in other laboratories. This is why the international metrology community is looking into the possibility of redefining the kilogram in terms of the fundamental physical constants, such as the electron charge and Planck's constant.

One proposal for replacing the platinum–iridium cylinder in Sèvres requires extremely accurate methods for measuring resistance and voltage (which are not base units). The primary standard for resistance is based on the quantum Hall effect in a two-dimensional electron gas in a semiconductor heterostructure, but on page 186 of this issue Alexander Tzalenchuk and co-workers report that they can measure this effect in graphene with an accuracy that approaches that of the semiconductor method. Moreover, the graphene devices (complete with gold–titanium contacts) have the advantage that they require lower magnetic fields and can operate at higher temperatures. One of the more unusual features of graphene is that the charge carriers behave as relativistic particles that do not have mass. As is pointed out on page 171, it would be ironic if such particles were involved in a new definition of the kilogram.