ELMO 2014 SL G8
by Wolstenholme, Aug 1, 2014, 9:43 PM
In triangle
with incenter
and circumcenter
, let
be the points of tangency of its circumcircle with its
-mixtilinear circles, respectively. Let
be the circle through
that is tangent to
at
, and define
similarly. Prove that
have a common point
other than
.
Proof:
Lemma 1: If
,
are the tangency points of the
-mixtilinear incircle with
respectively then
is the midpoint of
.
Proof: This is a well-known result, and moreover can be proved immediately with a limiting case of Sawayama's Theorem, but I shall provide a more elementary proof. Let
be the midpoints of arcs
respectively. Then by Archimedes' Lemma we have that
and
are collinear. Then by Pascal's Theorem on hexagon
we have that
are collinear. But since
and
this immediately implies the desired result.
Lemma 2: If
denotes the midpoint of arc
then
are collinear
Proof: Consider the homothety centered at
that takes the
-mixtilinear incircle to the circumcircle of triangle
. This takes
to
and
to
so by Lemma 1 it takes
to
, the midpoint of
. So it suffices to show that
are collinear. Some quick angle-chasing yields the fact that
is a parallelogram which implies the desired result.
Now, taking the inversion centered at
with radius
, we have that
is mapped to the line parallel to
passing through the reflection of
over
. We obtain two similar lines for
and
. Now, reflect all three lines about
.
So, letting
be the line through
parallel to
and defining
similarly it suffices to show that these three lines concur at the circumcenter of the excentral triangle of triangle
. Let
be the
-excenters of triangle
respectively. It is well-known that
is the midpoint of
and that
so
is the perpendicular bisector of
. This immediately implies the desired result, and so we are done.













Proof:
Lemma 1: If






Proof: This is a well-known result, and moreover can be proved immediately with a limiting case of Sawayama's Theorem, but I shall provide a more elementary proof. Let








Lemma 2: If



Proof: Consider the homothety centered at












Now, taking the inversion centered at









So, letting













This post has been edited 1 time. Last edited by Wolstenholme, Aug 1, 2014, 9:45 PM