The British National Standard Kilogram,

at the National Physical Laboratory in Teddington.

It's a lump of platinum-iridium alloy,

it's kept in a very very secure safe just under here,

and it weighs... well, one kilogram.

Mass measurement around the world needs to be equivalent.

So if I buy something in America that's a milligram, or a kilogram, or 100 tonnes,

I need to know that it's equivalent to a milligram,

or a kilogram, or 100 tonnes in the UK.

The real key areas of interest, where the uncertainties are very important,

are, for example, the pharmaceutical industry

where you're dealing with very, very small quantities of active ingredients

and you have to measure them very, very accurately.

So when, say, your bathroom scales were manufactured,

they were tested against equipment that was tested against equipment that,

several chains down the line later,

was tested against something like this.

And this was tested against the International Prototype Kilogram in Paris.

And this is a really weird thing to wrap your head around,

but that one in Paris? It always weighs a kilogram.

It doesn't matter if it's gained a tiny amount of mass from adsorbing air contaminants,

or lost a tiny amount when it was cleaned.

It is one kilogram.

We know that the International Prototype is probably not stable,

because all its copies are unstable relative to itself.

But there's nothing above the International Prototype that you can compare it with

to tell whether it's changing or not.

Defining international constants by a physical thing is... not ideal,

so in the next few years, scientists round the world

are deciding what the 21st century standard should be.

The formal definition will be based on physics, likely Planck's Constant,

but how do you make that into a physical thing?

There are two options: first, a fixed number of atoms.

You can manufacture 1kg sphere of almost-pure,

single-crystal silicon relatively easily.

What you have to measure here is the lattice spacing,

ie how far the atoms are apart in the sphere,

and the volume of the sphere. And both of those

are measurements you can make at a level of uncertainty

where you can generate a sphere which has an accuracy

of a part in 10-to-the-8, roughly.

The downside of that is it costs you €2.5million for a sphere.

Finding the kilogram that way seems to make sense.

We're measuring mass, after all.

But there is an alternative which could be cheaper and easier,

although it isn't quite as obvious.

The watt balance is an electrical way of measuring mass.

This is a way of demonstrating simply how a watt balance works.

We have a loudspeaker here

and if I pass a current through the coil of the loudspeaker,

you can see that the cone will move up and move down.

If I was to put a mass on here,

if I restore the position of the mass,

the current I need to do that is a measure of the weight of that mass.

The problem I've got is that I don't know the strength of the field of the magnet,

or the number of turns on the coil.

And I can do a second experiment by moving the cone of the speaker, like that.

If I was to measure the velocity with which I was moving that,

and also measure the voltage which was produced,

that would tell me that quantity and allow me to eliminate it from the two experiments.

Which, if done accurately, is enough to work out the weight of that mass.

Whichever approach gets chosen,

there will still be physical reference kilograms in the world.

They're not going away. But for the first time,

we're going to have something absolute to compare them against.

Thank you to everyone at the National Physical Laboratory!

They have a YouTube channel that you can check out,

and they occasionally hold open days. Details are on their web site.

[Translating these subtitles? Add your name here!]