This Google Books search finds a page that says this:

74. Legendre's theorem

${}\qquad\qquad{}$ If each of the angles of a spherical triangle whose
sides are small when compared with the radius of the sphere
be diminished by one third of the spherical excess, the triangle
may be solved as a plane triangle whose sides are equal to
the sides of the spherical triangle, and whose angles are these
reduced angles.

"Spherical excess" I take to mean the amount by which the sum of the three angles exceeds a half-circle.

Similarly imprecise statements are found by additional googling. The sides are "small when compared to", etc., and the solution is approximate. So I'm thinking there should be something maybe with a Landau little-o, saying as the size goes to $0$, the error goes to $0$ like some power of the size. Maybe.

Does anyone know of a mathematically precise statement of this proposition?

  • $\begingroup$ Your help is appreciated. But I've altered it further since then. $\endgroup$ – Michael Hardy Sep 6 '13 at 17:41
  • $\begingroup$ That last change looks much better! Software was outwitted in the end... $\endgroup$ – rschwieb Sep 6 '13 at 17:41

There is a paper about Legendre's theorem by Zbyněk Nádeník, Legendre theorem on spherical triangles. The paper states and proves several versions of the theorem (with different error terms).

As I understand, here is what Legendre originally proved (see also Wikipedia for historical references):

Theorem [Legendre, stated in 1787, proved in 1798]. Consider a spherical triangle with sides $a$, $b$, and $c$ and angles $\alpha$, $\beta$, and $\gamma$. Then the plane triangle with sides $a$, $b$, and $c$ has angles $\alpha'$, $\beta'$ and $\gamma'$:

\begin{align*} \alpha' &= \alpha - \varepsilon/3 + O(\varepsilon \cdot M^2),\\ \beta' &= \beta - \varepsilon/3 + O(\varepsilon \cdot M^2),\\ \gamma' &= \gamma - \varepsilon/3 + O(\varepsilon \cdot M^2), \end{align*} where $\varepsilon = \alpha + \beta + \gamma - \pi$, $M = \max(a,b,c)/R$, and $R$ is the radius of the sphere. Note that $\varepsilon = O(M^2)$.


Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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