Pompeïu's Theorem states that for an equilateral $\Delta ABC$ and a point $P$ that is not on the circumcircle of the triangle, the lengths $PA, PB, PC$ form the sides of a triangle. Any such triangle is degenerate iff $P$ lies on the circumcircle.

The proof for this is simple enough, and can be obtained by a simple transformation. But, what about the converse? I was not able to find a proof for that anywhere.

EDIT: The converse states that if in a triangle $ABC$, for every point $P$ in its interior, $PB, PC, PA$ are the sides of a triangle, then $ABC$ is equilateral.

  • $\begingroup$ Pompeiu's Theorem seems to be about points in the plane of a triangle, not just the interior. Be that as is may ... We can prove the contrapositive of the converse: "A non-equilateral triangle admits a point $P$ such that possibly-degenerate segments $\overline{PA}$, $\overline{PB}$, $\overline{PC}$ do not form a triangle." In the "plane" version, take $P$ to be a vertex where two non-congruent sides of the triangle meet. In the "interior" version, convince yourself that taking $P$ sufficiently close to that vertex creates segments that cannot form a triangle. $\endgroup$ – Blue Dec 15 '13 at 14:06
  • $\begingroup$ @Blue: Yes, you are right. I should have mentioned that this is a slightly weaker version of the converse, so to speak. But it holds true, nonetheless. $\endgroup$ – Gerard Dec 15 '13 at 14:18
  • $\begingroup$ RE Blue's comment. You can also take P close to the vertex but outside the triangle and not on the circumcircle ,so we get the full converse $\endgroup$ – DanielWainfleet Nov 1 '15 at 15:57

enter image description here

We can prove for any none equilateral $ΔABC$ we will find a point $K$ that KA,KB,KC can't form a triangle.

WOLG, let $BC$ is max side,$AB$ is min side, find $CD=AB, E$ is midpoint of $DB$, take $F$ on $EB$,make a red circle with $r=FB$, and get $GB=FB$ on $AB$. take H on $GB$. line $CH$, corss red circle at $I$, take $K$ on $IH$,

then $AK<AB=CD,KB<FB=ED \implies AK+KB<CD+EB=CE$,

but $CK>CI>CF>CE >AK+KB $

you may need to add more words and details to make the statement more solid.

| cite | improve this answer | |

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.