20
$\begingroup$

Is there a geometric explanation for why a sphere has surface area $4 \pi r^2$ ?
Ie equal to 4 times its cross-section (a circle of radius r).

$\endgroup$
  • 4
    $\begingroup$ This link does not give a full answer, but it may help a little: en.wikipedia.org/wiki/On_the_Sphere_and_Cylinder $\endgroup$ – Eivind Mar 27 '11 at 13:16
  • $\begingroup$ I would add to the comment of Eivind: the map from the cylinder to the sphere given by orthogonal projection from the axis is area-preserving. It's a nice exercise to show that it shrinks horizontal infinitesimal distances by the same factor as it expands vertical infinitesimal distances. $\endgroup$ – user8268 Mar 27 '11 at 13:36
  • $\begingroup$ What does cross-section mean here? $\endgroup$ – Rasmus Mar 27 '11 at 13:38
  • 1
    $\begingroup$ Here's a cute interpretation of the problem: On a spherical wedge of angle 90°, the curved outer surface has the same surface area as the two planar semicircular ends put together. One can think of these as two non-minimal surfaces on the same boundary curve. Why do they have the same area? (Of course, the answer may just be that it is a coincidence.) $\endgroup$ – Rahul Mar 27 '11 at 16:38
  • $\begingroup$ See here mathcentral.uregina.ca/QQ/database/QQ.09.99/wilkie1.html $\endgroup$ – leonbloy Mar 27 '11 at 17:39
13
$\begingroup$

Let $Z$ be a cylinder of height $2r$ touching the sphere $S_r$ along the equator $\theta=0$. Consider now a thin plate orthogonal to the $z$-axis having a thickness $\Delta z\ll r$. It intersects $S_r$ at a certain geographical latitude $\theta$ in a nonplanar annulus of radius $\rho= r\cos\theta$ and width $\Delta s=\Delta z/\cos\theta$, and it intersects $Z$ in a cylinder of height $\Delta z$. Both these "annuli" have the same area $2\pi r \Delta z$. As this is true for any such plate it follows that the total area of the sphere $S_r$ is the same as the total area of $Z$, namely $4\pi r^2$.

$\endgroup$
  • $\begingroup$ A "plate" here just means a very short cylinder, right? $\endgroup$ – LarsH Mar 14 '15 at 20:34
  • $\begingroup$ @LarsH: That's right. $\endgroup$ – Christian Blatter Mar 15 '15 at 7:17
6
$\begingroup$

One geometric explanation is that $4\pi r^2$ is the derivative of $\frac{4}{3}\pi r^3$, the volume of the ball with radius $r$, with respect to $r$. This is because if you enlarge $r$ a little bit, the volume of the ball will change by its surface times the small enlargement of $r$.

So why is the volume of the full ball $\frac{4}{3}\pi r^3$? By slicing the ball into disks, using Pythagoras, you get that its volume is $$ \int_{-r}^r \pi (r^2-x^2)\mathrm{d}x $$ which is indeed $\frac{4}{3}\pi r^3$.

$\endgroup$
  • 1
    $\begingroup$ Is this true for all manifolds, dV/dr=S?, where V is the n-volume and S is the n-1-surface and r is the distance from a point in the interior to the surcface ? $\endgroup$ – TROLLHUNTER Mar 27 '11 at 13:58
  • $\begingroup$ @solomoan: I think you will have trouble defining "the distance from a point in the interior to the surface" for a general manifold with boundary. I don't see how to generalise the result in a meaningful way to general manifolds. $\endgroup$ – Rasmus Mar 27 '11 at 14:02
  • $\begingroup$ @solomoan: If $$f: B\to{\mathbb R}^3, \quad (u,v)\mapsto f(u,v)$$ produces a surface $S$ with unit normal $n(u,v)$ then $$x: \ B\times[0,\epsilon]\ \to\ {\mathbb R}^3, \quad (u,v,t)\mapsto f(u,v)+ t n(u,v)$$ produces a plate of thickness $\epsilon$. You can compute the volume $V(\epsilon)$ of this plate by means of the Jacobian of $x$, and calculating the limit $$\lim_{\epsilon\to0}{V(\epsilon)\over\epsilon}$$ you get the formula for the surface area $\omega(S)$. $\endgroup$ – Christian Blatter Mar 27 '11 at 14:18
1
$\begingroup$

In the first step consider a hemi-sphere radius $R$ and its enveloping cylinder at the equator of sphere of length $R$.

$ \cos \phi = r/R = dz/dl $

$ d Area = 2 \pi r dl $

This results in same area

$$ 2 \pi R^2 $$

so a finite slice width has the same area either for a cone or cylinder of same width $\Delta z $.

Now consider the second step. it is comfortable to calculate flat areas, use the well known method of area calculation:

Cut the free boundary into an infinity of divisions of triangles, cut and discard every alternate infinitesimal narrow triangle,this will again cut into half its area, now to:

$$ 2 \pi R^2. $$

EDIT1

The area of a sphere segment is seen from integration as $ 2 \pi R $ times axial length of the spherical segment, which is the diameter for the full sphere.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.