World’s Most Perfect Spheres

Craft & Design Science
World’s Most Perfect Spheres

Image (1) kilogramsphere02.jpg for post 67963

These silicon spheres, manufactured by metrologists at Australia’s Commonwealth Scientific and Industrial Research Organization (CSIRO) as part of efforts to alleviate dependence on the International Prototype Kilogram, have properties that are as close to truly exact as measurable properties are likely to be for a long time to come: They weigh 1.0000000 kilograms, are smooth to the nearest 0.0000000003 meter, and are round to within 0.000000050 meters. One of them is even monoisotopic.

YouTube player

Image (2) kilogramsphere00.jpg for post 67963

Image (3) kilogramsphere01.jpg for post 67963

Image (5) kilogramsphere.jpg for post 67963

Image (6) kilogramsphere03.jpg for post 67963

58 thoughts on “World’s Most Perfect Spheres

  1. Sean Weatherford says:

    Englarge the spheres to the size of the Earth. Use the official equatorial diameter of 7926.41 miles as a reference. The resulting difference between lowest point (Marianas Trench) and highest point (Mount Everest) would then be 15.7 feet.

    1. Mike says:

      Earth ins’t perfectly round. It’s vaguely spherical, but the oceans actually rize and fall a long way due to lunar gravity. Also the earth is wider at the equator than anywhere else.

    2. Dan Schmiedeler says:

      NASA has two “more perfect” spheres that, if they were the size of the Earth, would have mountains maxing out at a mere 8 feet. See NASA’s Gravity Probe B Experiment at

  2. Wilson! says:

    I call BS on at least some of this. If these were the spheres mentioned in the story, I highly doubt they would be in anything other than a clean room, and certainly not held in the palm of someone’s hand (gloved or not). And I doubt “bearded scientist guy” would be in anything other than a full-body bunny suit. Just the act of taking them out and showing them off would affect the mass and diameter.

    Now, the orbs in the pics might be stand-ins, or something to that effect, but I seriously doubt these are the real McCoy.

    1. Sean Michael Ragan says:

      Certainly you’re free to call BS any time you like. Personally, if I’m ever going to take anybody’s word on anything, it’s likely to be these guys, as the whole point of their project is to be as exactly right about one measurable quantity as is humanly possible, and they’ve spent years and millions doing it. As for your claim that “taking it out and showing it off would affect the mass and diameter,” I can’t say if I agree or disagree. But I’m curious why you think so.

      1. Wilson! says:

        Because when something is measured to the tolerances these are, the mass and diameter will be thrown off by the abrasion of, or addition of, a few molecules. Or changing the temperature by holding it. Read further down for more elaboration.

        That’s why I figured the pictured orbs weren’t the “real” ones. Whether they were manufactured to the same tolerances or not, I’d put money that the ones in the pictures, especially the one being handled, are just “stand-ins” for the true standards.

        That’s something that always sort of amused me about these standards for mass and length, like the “official” meter kept in some vault in France. They are machined to such precise tolerances that the act of measuring them changes the value being measured.

      2. deadstatue says:

        the standards of distance are no longer physical objects as they are all defined by the distance of light traveled in a given time.

        and any measurements on size of the sphere would be taken after ample time is given for proper thermalisation.

        and the wearing away of atoms by cotton gloves would not be measurable for quite some time on a piece of polished silicon. unless your rubbing it like its your willy

    2. Dustbuster7000 says:

      Wilson, if you go to the CSIRO website ( you can read the article they wrote on this project. They made 14 spheres of which only two were the monoisotopic ‘perfect’ spheres which will used as the reference for the Avogadro Project sphere. The rest are certainly ‘the real McCoy’, but they won’t necessarily be as perfect as those two. They don’t say what the other ones were for, maybe they were initial trials or are for other less rigorous applications.

  3. Anonymous says:

    It’s “metre”, not “meter” in Australia and nearly all other parts of the world. For an article that discusses precision and metric units, it’s more than a little ironic that such flaws have been introduced. Editors of Make, please fix these typos; you’re making me cringe and ruining your own scientific cred.

    1. RocketGuy says:

      This isn’t a misspelling, it’s a linguistic gulf from the divergence of American English and UK/AU/NZ English(which probably also diverge from each other a bit, I have no data there though).

      English, in general, is quite flexible which allows for some of it’s success as a language. However, this does introduce the interesting interplay of slight differences in spelling and more importantly, definition.

      Try comparing billions sometime…

    2. Einstein says:

      My how quickly this discussion went from the spheres to ignorance. It only took one holier than thou aussie to set it off… If you ask me, I could give two toids whether its spelled meter or metre, or color or colour for that matter. I can read it, no sense in being a nit picking jerk. Next thing youll do is run around the internet looking for improper use of nouns, or correcting people on their use of your and you’re. Lifes short, 100 spins around the mass of gas and the games over. Dont spend it being annoying and petty, spend it well. You only get 1 life.

    3. Anonymous says:

      not only cringing, but “whinging” too, unless they’ve got another word for it Down Under…


  4. says:

    The processed used to make these things is surprisingly simple.

    I used to be a machinist the company that made one of the original “perfect” spheres and the process is actually kind of ghetto. Put a steel (or silicone in this case) ball between two cone shaped abrasive wheels & spin. If the bearing used in the spinning process has a true enough axis, you’ll get what you’re looking for.

    Also, metrology is as much about statistics as it is about actual measuring instruments.

    1. RocketGuy says:

      I was kinda flabbergasted how simple the process is for telescope mirror grinding was too, even for a 1/10th wave mirror. Long and boring, but simple.

  5. oskay says:

    NIST publishes a *very good* guide on the recommended use of the use of the SI: , in which “meter” is still preferred as the (US) spelling.

    Since this is a cultural difference, calling our spelling a “flaw” is not helpful, and referring to it as a “typo” is snarky at best.

    Wikipedia has an interesting discussion of (some of the) spelling differences between American English and British English:

  6. Jason says:

    With proper handling, the mass shouldn’t change but the diameter and shape can somewhat. The reason for making it a perfect sphere to begin with is so that anyone else making an identical sphere at identical temperature and pressure conditions will end up with an identical mass. The body heat of the person holding the object is enough to change it’s size (when talking about the accuracies involved in metrology). In the metrology class I took during grad school, one of our homework problems was to find out how much a 1 inch block of stainless steel would grow by having a person standing 1 meter away. It was an interesting class, but I wouldn’t want a job in the field.

    1. Zach C. says:

      At a wild guess, the steel block would expand by the approximate width of a carbon atom’s outer electron shell, provided the person standing nearby is of average height, structure, and heat output; and is wearing average cotton-polyester blend clothing; and remains in position for a sufficient amount of time for their body heat to suffuse the block.

  7. Adam says:

    If you are going to use CSIRO’s full title please use the UK/AU spelling of ‘Organisation’ as it’s a proper noun and shouldn’t be localised.

    1. Simon says:

      Since you’re commenting on Make, a US based site, shouldn’t you be griping about the locali_Z_ation?

      And I am only complaining since I live in NZ and because of that it’s my duty to take the piss out of an Aussie whenever possible (despite considering myself more an anglophile than a Kiwi since I was born in the UK).

      Bloody Aussie underarm bowlers :)

      1. Mig says:

        Not to jynx it but the Aussie bowlers have had a rotten first hour of today’s test.

        Take that America…

  8. SteelToad says:

    Oh.. “perfect silicon spheres” I see my search engine got it wrong again.

    1. Carl says:

      Not that I particularly care, but what is the remote use of making something to that level of “perfection”…?

      “efforts to alleviate dependence on the International Prototype Kilogram”


      1. Sean Michael Ragan says:

        As the linked article explains, the purpose of the enterprise is to produce a well-established physical system that defines how much one kilogram actually weighs. Of all the SI (metric) units, only the kilogram is still defined in terms of a particular physical object, which is, specifically, the “International Prototype Kilogram,” also linked to above. It is kept in a vault in Paris under carefully controlled conditions because if its weight changes, so does the weight of everything else in the universe that we say weighs a kilogram. Which changes can, on a global scale, cause massive unjustified devaluation or inflation of commodities that are measured in kilograms. The idea is to set up a definition of the kilogram something like “a kilogram is equal to the weight of a perfect sphere of monoisotopic metallic silicon exactly 93.7 mm in diameter,” or whatever, instead of “a kilogram is equal to the weight of the International Prototype Kilogram kept in Paris, France.”

  9. gyziger says:

    You know, I find it strange that there are all these gripes about spelling when regardless of any US/UK/AU differences the numbers speak for themselves. Who cares if it’s “meter” or “metre?” These are great feats of science and engineering, being fickle about the spelling of certain measurements are in a way, inefficient.

    1. Rahere says:

      Until the 1950s, this was a story – afterwards, commonplace, because the techniques became simple and therefore accessible to all. The equivalent with a surface involves two reference planes and the finest grit you can make: and much time.

  10. Jay says:

    These are not the most perfect spheres. The worlds most perfect spheres are part of gyroscopes onboard the Gravity Probe-B satellite. They are some five times more round.

    1. Sean Michael Ragan says:

      Because the spheres in question are not the same size, comparison of absolute errors is not meaningful. If you go through the math, you will find that the relative errors of the kilogram spheres and the Gravity Probe-B satellite spheres are exactly the same. If the sources of the respective wikipedia articles are to be believed, in each case, if the sphere were scaled up to the size of earth (an operation which normalizes the errors and makes their comparison meaningful), the difference between the high and low points is 8 feet. This is not too surprising, as both sets of spheres have clearly been rounded to the limits of our technological ability to do so, and it’s the same limit in each case, because it’s the same technology.

  11. Joe says:

    Wilson, don’t be an idiot. Temperature doesn’t affect mass. Which was the whole point of this exercise.

  12. duke_leeto_III says:

    shrink earth down to the size of a cue ball and earth would be smoother

Comments are closed.

Discuss this article with the rest of the community on our Discord server!

I am descended from 5,000 generations of tool-using primates. Also, I went to college and stuff. I am a long-time contributor to MAKE magazine and My work has also appeared in ReadyMade, c't – Magazin für Computertechnik, and The Wall Street Journal.

View more articles by Sean Michael Ragan
Maker Faire Bay Area 2023 - Mare Island, CA

Escape to an island of imagination + innovation as Maker Faire Bay Area returns for its 15th iteration!

Buy Tickets today! SAVE 15% and lock-in your preferred date(s).