## Perfect Sphere

For the discussion of the sciences. Physics problems, chemistry equations, biology weirdness, it all goes here.

Moderators: gmalivuk, Moderators General, Prelates

crookedeye
Posts: 5
Joined: Thu Oct 21, 2010 3:53 am UTC

### Perfect Sphere

This has mostly been something I've been thinking about for the past week or so and I wanted to get opinions on how one would tackle such a problem. So lets say you are standing in a room and given a large (we'll say greater than 10m diameter) sphere. The claim is that it is a perfect sphere, how would you go about proving that it is so? I should say this is ignoring the whole "there is no such thing as a perfect sphere in nature" and your access to tools/equipment is not constrained.

ian
Posts: 706
Joined: Fri Mar 07, 2008 3:55 pm UTC
Location: Sealand

### Re: Perfect Sphere

atomic microscopy

or hit with something hard enough to put a dent in it.

++\$_
Mo' Money
Posts: 2370
Joined: Thu Nov 01, 2007 4:06 am UTC

### Re: Perfect Sphere

Measure its purported diameter by balancing a flat sheet on top of it and measuring the height of the sheet. Then go to your lab and construct two perfect, rigid, impermeable hemispherical shells of the same purported radius as the "sphere". Fill the room with <insert favorite gas here>. Enter the room and seal the shells around the "sphere" (if it's not possible to do this, then it's not spherical, so stop right now). Leave the room. Empty the room of <favorite gas>. Enter the room. Remove the shells. If the object was not spherical, then you will be able to detect some gas in the room. Otherwise, the room will be completely free of <favorite gas>.

Qaanol
The Cheshirest Catamount
Posts: 3064
Joined: Sat May 09, 2009 11:55 pm UTC

### Re: Perfect Sphere

I’d pick a point on the object, set that point on the floor, find the highest point on the object, and measure its height. Then I’d construct a perfectly cylindrical tube of exactly that interior diameter (perhaps plus some arbitrarily small epsilon) and roll the object through the tube three times on mutually perpendicular axes. Only a sphere should be able to pass this test without hitting or moving away from the tube walls. There are other shapes of constant width, but I’m pretty sure the sphere is the only one with a constant circular cross-section of rotation. And the second and third rolls are just to double and triple check, really.
wee free kings

kvn.l
Posts: 22
Joined: Sun Sep 20, 2009 12:49 am UTC

### Re: Perfect Sphere

If it's a perfect sphere, wouldn't it damage everything it touches or itself, as whatever it touches would be tangential to it, and give that pressure is Force/Area, wouldn't that do something destructive to the sphere or object it touches? If so, we could use this to prove that it's perfect?

Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

### Re: Perfect Sphere

kvn.l wrote:If it's a perfect sphere, wouldn't it damage everything it touches or itself, as whatever it touches would be tangential to it, and give that pressure is Force/Area, wouldn't that do something destructive to the sphere or object it touches? If so, we could use this to prove that it's perfect?

The other stuff would just compress around it, like it does normally.
Some people tell me I laugh too much. To them I say, "ha ha ha!"

Mr_Rose
Posts: 380
Joined: Thu May 29, 2008 9:32 am UTC

### Re: Perfect Sphere

Determine any diameter of the alleged sphere by any suitable method. Use this to calculate the radius, then use this to calculate the volume of an ideal sphere of the same radius.

Using a suitably large volume of an appropriate liquid, submerge the sphere and measure the volume displaced.

Compare the displacement with the calculated ideal. Any discrepancy within the limits of measurement/calculation indicates an imperfect sphere.
Microevolution is a term — when used by creationists — that is the evolutionary equivalent of the belief that the mechanism you use to walk from your bedroom to the kitchen is insufficient to get you from New York to Los Angeles.

telcontar42
Posts: 430
Joined: Sat Jun 28, 2008 10:33 pm UTC
Location: Davis, CA
Contact:

### Re: Perfect Sphere

This wouldn't prove that it was a perfect sphere, just that it has the same volume as a perfect sphere. It could be slightly deformed and have the same volume and the same diameter at the point you measured.

frezik
Posts: 1336
Joined: Wed Jan 10, 2007 7:52 pm UTC
Location: Schrödinger's Box

### Re: Perfect Sphere

How about relativity? Since it's traveling at some speed through the universe, it must be elongated (however slight) along its vector. So it can only be a prefect sphere within the same reference frame.

In general, I don't think you can prove it's a perfect sphere. Any method you use to measure it would itself subject to measurement error.
I do not agree with the beer you drink, but will defend to the death your right to drink it

Moose Hole
Posts: 398
Joined: Fri Jul 09, 2010 1:34 pm UTC

### Re: Perfect Sphere

If it were so perfect it would give you a blowjob and a bunch of money.

crookedeye
Posts: 5
Joined: Thu Oct 21, 2010 3:53 am UTC

### Re: Perfect Sphere

Hmm I didn't think of considering relativistic effects. Although as far as measurements go, I would think that because we are ignoring the idea that no perfect sphere can exist in nature, it would be safe to assume that you are capable, through some unknown method, to have completely accurate measurements.

frezik
Posts: 1336
Joined: Wed Jan 10, 2007 7:52 pm UTC
Location: Schrödinger's Box

### Re: Perfect Sphere

Expanding on the above, to prove that you're in the same reference frame, you have to measure speed. To measure if it's a sphere, you have to measure the position of particles along the surface. Since you must measure closely enough for quantum effects to matter, and you can't measure both speed and position, you now have both Einstein and Heisenberg waving fingers at you.
I do not agree with the beer you drink, but will defend to the death your right to drink it

BlackSails
Posts: 5315
Joined: Thu Dec 20, 2007 5:48 am UTC

### Re: Perfect Sphere

Physical spheres are made of discrete points (ie, molecules) and therefore cannot actually form a perfect sphere.

QED.

letterX
Posts: 535
Joined: Fri Feb 22, 2008 4:00 am UTC
Location: Ithaca, NY

### Re: Perfect Sphere

BlackSails wrote:Physical spheres are made of discrete points (ie, molecules) and therefore cannot actually form a perfect sphere..

crookedeye wrote: I should say this is ignoring the whole "there is no such thing as a perfect sphere in nature" and your access to tools/equipment is not constrained.

Yeah, we know. Anyways I was assuming this thread was in general about measuring the sphere-icity of an object, and then describing how far a given circular thing deviates from being an actual sphere. In which case you might want to ask some geographers* because they've been measuring this big almost-but-not-quite spherical thing for some time, and have had a great deal of success.

* There's a more specific word for this, but it escapes me at the moment...

Dark Knight Bob
Posts: 14
Joined: Thu Oct 21, 2010 10:04 pm UTC

### Re: Perfect Sphere

Using your quantum tunneling device, teleport yourself inside and attach weight (a) at any point on the sphere.

Fill the sphere with water now attach a piece of string to the weight (a) and attach a float (b) to the other end that transmits an electircal signal whenever it tocuches the spere.

Allow gravity to roll the sphere so that the Weight(a) is at the lowest point possible due to gravity.

Now release the float so that due to viscosity and gravity the float now rises to the highest point in the sphere possible(checking with the transmitter) you now have theoretical limit for the spheres radius

now you can anchour both (a) & (b) so they cannot move and attach a spinning weight (c) to the radius

Fix a rigid rod of the equivalent diameter and now detach the float and drain the water. Now swing the float around in a rotaional pattern for the centre of the sphere via measured radius.

via induction you can now demonstrate that so long as the float is moving fast enough to never become less that the radius you can use the electircal transmitter to falsify the claim that it is not a perfect sphere.i.e. if the float ever loses contact inside.

The mathematicians will surely be up in arms but screw em. So long as you vary angle of rotation you can technically cover ANY equatorial of the sphere and thus test the hypothesis.

Do I win?

Antimony-120
Posts: 830
Joined: Wed Apr 09, 2008 4:16 am UTC
Location: Wherever you can look - wherever there's a fight, so hungry people can eat.

### Re: Perfect Sphere

letterX wrote:
BlackSails wrote:Physical spheres are made of discrete points (ie, molecules) and therefore cannot actually form a perfect sphere..

crookedeye wrote: I should say this is ignoring the whole "there is no such thing as a perfect sphere in nature" and your access to tools/equipment is not constrained.

Yeah, we know. Anyways I was assuming this thread was in general about measuring the sphere-icity of an object, and then describing how far a given circular thing deviates from being an actual sphere. In which case you might want to ask some geographers* because they've been measuring this big almost-but-not-quite spherical thing for some time, and have had a great deal of success.

* There's a more specific word for this, but it escapes me at the moment...

if you mean for the almost-but-not-quite-spherical thing, it's called a geoid. Yeah, turns out the earth is an earth-shaped objects. In other news tautologies are tautological.

Anyhow, I would shine a laser onto it at some point, use a beam spltter and recombination intereformetre setup, and then spin the "sphere" on some axis. If it's perfectly spherical there will be a static interference pattern. Repeat for a couple random other axes to be certain. That'll give you a pretty good error of measurement (a couple hundred nm). If you really want accuracy, and proove that it breaks physics by being more spherical than anything made out of atoms could be, use a suitably energetic electron interferometre in the same fashion. This of course assumes that you're capable of finding an axis to a great degree of accuracy (though there are some cute ways of doing this).
Wolydarg wrote:That was like a roller coaster of mathematical reasoning. Problems! Solutions! More problems!

****************Signature Dehosted, New Signature Under Construction************************

uncivlengr
Posts: 1202
Joined: Fri Nov 14, 2008 10:35 pm UTC
Location: N 49°19.01 W 123°04.41

### Re: Perfect Sphere

Since the sphere is phyically impossible, it's going to necessarily require to be measured by some physically impossible means, or not at all.

Once you've crossed that line, the sky's the limit - you can take your set of "perfect" calipers and hook it up to your infinitely fast robotic arm, that can measure the diameter from an infinite number of points on the surface of the sphere in a finite amount of time, stopping whenever a discrepancy in the readings is found, or when every infintessimal point on the surface has been measured.

It's actually a more interesting question when you impose real world contraints on the problem.
I don't know what to do for you

BlackSails
Posts: 5315
Joined: Thu Dec 20, 2007 5:48 am UTC

### Re: Perfect Sphere

Yeah, if the sphere can be perfect, you cannot tell, since your instruments are not perfect. And if your instruments are perfect, then its easy, you just measure the sphere.

Carnildo
Posts: 2023
Joined: Fri Jul 18, 2008 8:43 am UTC

### Re: Perfect Sphere

BlackSails wrote:And if your instruments are perfect, then its easy, you just measure the sphere.

But how do you do it with a finite number of measurements?

Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

### Re: Perfect Sphere

Carnildo wrote:
BlackSails wrote:And if your instruments are perfect, then its easy, you just measure the sphere.

But how do you do it with a finite number of measurements?

Don't, but make each measurement-equivalent take an infinitely small amount of time/make your measurements continuous rather than discrete.
Some people tell me I laugh too much. To them I say, "ha ha ha!"

Dark Knight Bob
Posts: 14
Joined: Thu Oct 21, 2010 10:04 pm UTC

### Re: Perfect Sphere

I think there has to be some kind of macroscopic limit here for where we can draw the line at what constitutes perfect given that at some point we'll be dealing with oscillating wave functions and not anything that remotely resembles an actual sphere.

Also in relation to my own ego,I think I came up with best answer using physical principles (other than people invoking the 'how to measure the height of a building using a barometer' urban)

i.e finding a way to use gravity to find the spheres radius and then using angular momentum to gauge if at any point the spheres radius ceases to be a constant.

I toyed with using a singularity at the centre with the event horizon being on the very limit of the radius thus making any imperfections pressing inwards cause the sphere to collapse inwards but the simpler the better

BlackSails
Posts: 5315
Joined: Thu Dec 20, 2007 5:48 am UTC

### Re: Perfect Sphere

Carnildo wrote:
BlackSails wrote:And if your instruments are perfect, then its easy, you just measure the sphere.

But how do you do it with a finite number of measurements?

Charge the sphere, stick a charged ball in the exact middle (as measured by your perfect ruler) and see if it moves.

Spin the sphere, measure the moment of inertia tensor.

Aelfyre
Posts: 358
Joined: Wed May 07, 2008 4:22 pm UTC
Location: 3 decades west of the teenage wasteland
Contact:

### Re: Perfect Sphere

so am I the only one who is leaning towards.. eyeball it.. proclaim it perfect, then challenge someone to prove you wrong?

I'm bad at this arent I?
Xanthir wrote:To be fair, even perfectly friendly antimatter wildebeests are pretty deadly.

Dark Knight Bob
Posts: 14
Joined: Thu Oct 21, 2010 10:04 pm UTC

### Re: Perfect Sphere

Then we'd have to extract your eyeball and find a way to prove your eyeball is a perfect optical insturment.

I thought about the moment of inertia tensor but remember it doesn't have to be a sphere in order to give a result equal to a spherical radius

Unless of course you do it for all 3 axis. I don't think you can get a shape that creates a symmetrical result there unless there's some mathematician who knows of one.

Velifer
Posts: 1132
Joined: Wed Dec 26, 2007 4:05 pm UTC
Location: 40ºN, 83ºW

### Re: Perfect Sphere

Antimony-120 wrote:if you mean for the almost-but-not-quite-spherical thing, it's called a geoid. Yeah, turns out the earth is an earth-shaped objects. In other news tautologies are tautological.

Or perhaps he was thinking of an oblate spheroid. There can be more than one way to describe things. You know that, right? Roundish, geoidal, your-mamma-esque. All the same thing, though "roundish" doesn't connote the enormous size as well.
Time flies like an arrow, fruit flies have nothing to lose but their chains -Marx

Aelfyre
Posts: 358
Joined: Wed May 07, 2008 4:22 pm UTC
Location: 3 decades west of the teenage wasteland
Contact:

### Re: Perfect Sphere

Velifer wrote:
Antimony-120 wrote:if you mean for the almost-but-not-quite-spherical thing, it's called a geoid. Yeah, turns out the earth is an earth-shaped objects. In other news tautologies are tautological.

Or perhaps he was thinking of an oblate spheroid. There can be more than one way to describe things. You know that, right? Roundish, geoidal, your-mamma-esque. All the same thing, though "roundish" doesn't connote the enormous size as well.

well of course not.. thats the a vague approximation of shape.. size is conferred from your-mamma-esque.

EDIT: oh wait my bad.. He probably meant at least 10 meters not 10 miles.. so no it isn't a very appropriate term.
Xanthir wrote:To be fair, even perfectly friendly antimatter wildebeests are pretty deadly.

martin878
Posts: 63
Joined: Mon Nov 03, 2008 12:47 pm UTC
Location: Oxford, UK

### Re: Perfect Sphere

Seems like some are trying to avoid the question by throwing smart sounding ideas at it that make it seem impossible to answer. How about we pose the question as about a measure of imperfection rather than judging whether or not it's perfect. A perfect sphere would have the property x2+y2+z2=R everywhere on the surface, for some co-ordinate system centred in the middle. So if we measure any point on the surface, we can calculate its distance from the centre, and subtract R to give r. If we do this for multiple points, stacking into a vector r we have the mean-square error, rTr. This is the measure of imperfection.

If you're concerned that you don't know the centre, first assume the sphere is perfect, pose the problem as a least-squares approximation of the unknowns x0, y0 and z0. Then take loads of measurements and find the most likely centre.

As for practicalities of measuring position of a point on the surface, how about a laser range finder?

Edit: Oops, some squares missing in there.
Last edited by martin878 on Thu Oct 28, 2010 10:20 am UTC, edited 1 time in total.

Velifer
Posts: 1132
Joined: Wed Dec 26, 2007 4:05 pm UTC
Location: 40ºN, 83ºW

### Re: Perfect Sphere

Construct a shape out of all possible tangent lines, examine the created shape to see if it is entirely space filling (well, except for a perfect sphere.)
Time flies like an arrow, fruit flies have nothing to lose but their chains -Marx

Carnildo
Posts: 2023
Joined: Fri Jul 18, 2008 8:43 am UTC

### Re: Perfect Sphere

Velifer wrote:Construct a shape out of all possible tangent lines, examine the created shape to see if it is entirely space filling (well, except for a perfect sphere.)

I think that would only tell you that you've got a closed object with no sharp edges.

Cynwulf
Posts: 208
Joined: Mon Mar 23, 2009 9:03 pm UTC
Location: California

### Re: Perfect Sphere

1. Resurrect Giotto.
2. Buy Giotto lunch and ask him to sculpt you a perfect sphere.
3. Compare Giotto-sphere to contender-sphere.
4. ???
5. Profit.

Since constructing an apparatus with perfect accuracy and precision is impossible, it would be impossible to determine whether a perfect sphere was in fact perfect. You could only measure its circularity/roundness to a ludicrously high confidence. With that said, I'm sure if we measured a macroscopic object to within +/-1 Planck length, then I would go ahead and call that perfect. How to do it? Meh, not my job.
L'homme est libre au moment qu'il veut l'être. | Man is free at the instant he wants to be.
- Voltaire

Solt
Posts: 1912
Joined: Tue Mar 27, 2007 5:08 am UTC
Location: California

### Re: Perfect Sphere

Build a continuous flat ring around the earth's surface in an evacuated chamber, using the same material the sphere is made of. Roll the sphere. It should roll forever.

BlackSails wrote:Spin the sphere, measure the moment of inertia tensor.

This. Should be the same number repeated across the diagonal and zeros elsewhere iirc. You do it by spinning it up along each of three orthogonal axes, with constant force, and measuring the resulting rotational acceleration. Maybe strap some perfect thrusters onto it.

My favorite method would be to put it out in deep space, then orbit something around it. It should take a very very long time to drift out of orbit. Actually this might be better suited as a method of measuring the cumulative gravitational force exerted by the universe on a single region of space, assuming your sphere really is perfect.
"Welding was faster, cheaper and, in theory,
produced a more reliable product. But sailors do
not float on theory, and the welded tankers had a
most annoying habit of splitting in two."
-J.W. Morris

Qaanol
The Cheshirest Catamount
Posts: 3064
Joined: Sat May 09, 2009 11:55 pm UTC

### Re: Perfect Sphere

Solt wrote:Build a continuous flat ring around the earth's surface in an evacuated chamber, using the same material the sphere is made of. Roll the sphere. It should roll forever.

Actually, a perfect sphere set rolling on a level surface with no air resistance still slows down. I wish I could remember the place I read about this effect, but it has to do with the normal force from the ground being stronger slightly in front of the center of the point of contact than slightly behind, due to the rolling motion, so there is a net torque acting to slow rotation. And in any real-life scenario it will eventually come to a stop.

Solt wrote:
BlackSails wrote:Spin the sphere, measure the moment of inertia tensor.

This. Should be the same number repeated across the diagonal and zeros elsewhere iirc. You do it by spinning it up along each of three orthogonal axes, with constant force, and measuring the resulting rotational acceleration. Maybe strap some perfect thrusters onto it.

I admit I don’t know much about tensors, but are you presupposing the object to be made of a material with uniform (or at least rotationally symmetric) density?

Solt wrote:My favorite method would be to put it out in deep space, then orbit something around it. It should take a very very long time to drift out of orbit. Actually this might be better suited as a method of measuring the cumulative gravitational force exerted by the universe on a single region of space, assuming your sphere really is perfect.

I think you already see why this doesn’t work.
wee free kings

thoughtfully
Posts: 2253
Joined: Thu Nov 01, 2007 12:25 am UTC
Location: Minneapolis, MN
Contact:

### Re: Perfect Sphere

Solt wrote:My favorite method would be to put it out in deep space, then orbit something around it. It should take a very very long time to drift out of orbit. Actually this might be better suited as a method of measuring the cumulative gravitational force exerted by the universe on a single region of space, assuming your sphere really is perfect.

You mean something like this?

Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.
-- Antoine de Saint-Exupery

Technical Ben
Posts: 2986
Joined: Tue May 27, 2008 10:42 pm UTC

### Re: Perfect Sphere

uncivlengr wrote:Since the sphere is phyically impossible, it's going to necessarily require to be measured by some physically impossible means, or not at all.

Once you've crossed that line, the sky's the limit - you can take your set of "perfect" calipers and hook it up to your infinitely fast robotic arm, that can measure the diameter from an infinite number of points on the surface of the sphere in a finite amount of time, stopping whenever a discrepancy in the readings is found, or when every infintessimal point on the surface has been measured.

It's actually a more interesting question when you impose real world contraints on the problem.

I want to see that robot!
It's all physics and stamp collecting.
It's not a particle or a wave. It's just an exchange.

Velifer
Posts: 1132
Joined: Wed Dec 26, 2007 4:05 pm UTC
Location: 40ºN, 83ºW

### Re: Perfect Sphere

Carnildo wrote:
Velifer wrote:Construct a shape out of all possible tangent lines, examine the created shape to see if it is entirely space filling (well, except for a perfect sphere.)

I think that would only tell you that you've got a closed object with no sharp edges.

Topology. How does that work?

Fine. Cut it into pieces and see if you can reassemble it into two or more identical spheres.
Time flies like an arrow, fruit flies have nothing to lose but their chains -Marx

Antimony-120
Posts: 830
Joined: Wed Apr 09, 2008 4:16 am UTC
Location: Wherever you can look - wherever there's a fight, so hungry people can eat.

### Re: Perfect Sphere

Unfortunatly that won't work either, you're most likely referring to the Banach-Tarski paradox which has nothing to do with a perfect sphere, though using it on a perfect sphere is by far the easiest to work out example of the problem, see for example Tarski's circle-squaring problem, althought that isn't quite the same principal. In any case the point is that you haven't proven it's a perfect sphere, just some that it possesses some characteristics.
Wolydarg wrote:That was like a roller coaster of mathematical reasoning. Problems! Solutions! More problems!

****************Signature Dehosted, New Signature Under Construction************************

uncivlengr
Posts: 1202
Joined: Fri Nov 14, 2008 10:35 pm UTC
Location: N 49°19.01 W 123°04.41

### Re: Perfect Sphere

Qaanol wrote:Actually, a perfect sphere set rolling on a level surface with no air resistance still slows down. I wish I could remember the place I read about this effect, but it has to do with the normal force from the ground being stronger slightly in front of the center of the point of contact than slightly behind, due to the rolling motion, so there is a net torque acting to slow rotation. And in any real-life scenario it will eventually come to a stop.
That raises another problematic issue with using the notion "perfection" - do external forces deform a "perfect" sphere? Are gravitational forces going to deform the sphere, spoiling the testing?
I don't know what to do for you

Solt
Posts: 1912
Joined: Tue Mar 27, 2007 5:08 am UTC
Location: California

### Re: Perfect Sphere

Qaanol wrote:Actually, a perfect sphere set rolling on a level surface with no air resistance still slows down. I wish I could remember the place I read about this effect, but it has to do with the normal force from the ground being stronger slightly in front of the center of the point of contact than slightly behind, due to the rolling motion, so there is a net torque acting to slow rotation. And in any real-life scenario it will eventually come to a stop.

Yes, because in any real life scenario it would NOT be a perfect sphere, and it would deform. You are actually describing one of the two mechanisms by which a perfectly round real wheel experiences friction, the second being through continuous physical deformation at the point of contact (deformation requires energy).

A perfect sphere must be infinitely rigid which creates many other problems, clearly, but one of those problems would NOT be deformation and thus the mechanism you describe would not happen, because the sphere would only touch the surface at a single point. Literally, a single geometric point. Or maybe the area of contact would be a planck's length squared or something. I don't know that part should be answered by the physicists.
"Welding was faster, cheaper and, in theory,

produced a more reliable product. But sailors do

not float on theory, and the welded tankers had a

most annoying habit of splitting in two."

-J.W. Morris

thoughtfully
Posts: 2253
Joined: Thu Nov 01, 2007 12:25 am UTC
Location: Minneapolis, MN
Contact:

### Re: Perfect Sphere

Solt wrote:I don't know that part should be answered by the physicists.

The physicists are staying the hell away from this problem until people stop treating it like a math problem and pick a sane, real-life definition for "perfection".

Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.
-- Antoine de Saint-Exupery

Solt
Posts: 1912
Joined: Tue Mar 27, 2007 5:08 am UTC
Location: California

### Re: Perfect Sphere

Yea it can never really be mathematically round.

And I can't think of a good way to define it as perfectly round using physical terms. Calculus shows us that a finite collection of small segments can never be a continuous curve, no matter how small you take the segments to be, unless you take an infinite number of infinitely small segments. So what is the most round an object in nature can actually be? How round is a star? How round is a collection of phospholipids immersed in water? In both cases, not very. How round is a black hole?

If I had to choose I guess I'd say a perfect sphere is actually a spherical shell exactly 1 atom thick, with a radius chosen such that the gaps between every pair of neighboring atoms is equal. Of course such a sphere could only exist in a vacuum and with no external forces on it, and I don't know if inter-atomic bonding would even be strong enough.

For some reason this makes me want to build a sphere in Minecraft.
"Welding was faster, cheaper and, in theory,

produced a more reliable product. But sailors do

not float on theory, and the welded tankers had a

most annoying habit of splitting in two."

-J.W. Morris