Kvant Math Problem 2863

Verified: no
Verdicts: UNKNOWN + UNKNOWN
Solve time: 10m17s
Source on kvant.digital

Problem

A cyclic quadrilateral $ABCD$ (Fig. 1) is given. The rays $AB$ and $DC$ intersect at point $X$, and the rays $AD$ and $BC$ intersect at point $Y$. The bisectors of angles $AXD$ and $AYB$ intersect at point $J$. Prove that point $J$ lies on the line connecting the midpoints of the diagonals $AC$ and $BD$.

Folklore

Exploration

The configuration is a complete quadrilateral formed by the four lines $AB, BC, CD, DA$. Their three diagonal intersection points are $P = AC \cap BD$, $X = AB \cap CD$, and $Y = AD \cap BC$. The midpoints of the three diagonal segments $AC$, $BD$, and $XY$ are collinear by Newton’s theorem, so the line through the midpoints of $AC$ and $BD$ is the Newton line of this complete quadrilateral.

The point $J$ is defined as the intersection of the internal bisector of the angle formed by the lines $XA$ and $XD$, and the internal bisector of the angle formed by the lines $YA$ and $YB$. Each of these bisectors is determined purely by angle symmetry with respect to pairs of lines in the complete quadrilateral.

The target is to connect an angular construction at $X$ and $Y$ with a purely affine object, the Newton line. This suggests that the bisector conditions must encode a midpoint condition along the diagonal structure. The key point to unlock the problem is that angle bisectors in a complete quadrilateral can be reinterpreted as loci equidistant from pairs of opposite sides, which in turn can be linearized along the Newton line using directed angle equalities.

The most delicate step is to translate “intersection of two angle bisectors” into a single linear condition along the Newton line without losing metric information.

Problem Understanding

This is a Type B problem: prove that a point defined by two angle bisectors lies on a specific line.

A cyclic quadrilateral $ABCD$ is given, and $X$ and $Y$ are the intersections of opposite sides of the complete quadrilateral formed by the four lines of $ABCD$. The point $J$ is defined as the intersection of the bisectors of angles $\angle AXD$ and $\angle AYB$. The task is to prove that $J$ lies on the line joining the midpoints of diagonals $AC$ and $BD$.

The core difficulty is to connect angular bisector geometry at $X$ and $Y$ with the affine midpoint structure of the Newton line of the complete quadrilateral. The expected outcome is that $J$ lies on this Newton line.

Proof Architecture

The first lemma identifies $X, Y, P$ as the diagonal intersection points of a complete quadrilateral and establishes that the midpoints of $AC$, $BD$, and $XY$ are collinear on the Newton line.

The second lemma rewrites the angle bisector condition at $X$ as an equality of directed angles between a variable point and the lines $XA$ and $XD$, and similarly at $Y$.

The third lemma shows that a point satisfying both bisector conditions must preserve a fixed ratio of directed angles determined solely by the diagonals, which forces the point to lie on the Newton line.

The most delicate step is the conversion of the simultaneous bisector conditions into a single linear incidence condition with the midpoint line.

Solution

Let $ABCD$ be a cyclic quadrilateral. The lines $AB$, $BC$, $CD$, $DA$ form a complete quadrilateral. Denote $P = AC \cap BD$, $X = AB \cap CD$, and $Y = AD \cap BC$. The segments $AC$, $BD$, and $XY$ are the three diagonals of this complete quadrilateral.

Let $M$ be the midpoint of $AC$ and $N$ the midpoint of $BD$. By Newton’s theorem for complete quadrilaterals, the midpoints of $AC$, $BD$, and $XY$ are collinear, so $M$, $N$, and the midpoint of $XY$ lie on a line, which is the Newton line.

The point $J$ is defined as the intersection of the internal bisector of $\angle AXD$ and the internal bisector of $\angle AYB$. The angle $\angle AXD$ is formed by the lines $XA$ and $XD$, while $\angle AYB$ is formed by the lines $YA$ and $YB$.

We express the bisector condition at $X$ in directed angle form. A point $J$ lies on the bisector of $\angle AXD$ if and only if

$$\angle AXJ = \angle JXD.$$

Since $A, X, B$ are collinear and $D, X, C$ are collinear, this condition depends only on the directions of the lines $XA$ and $XD$.

Similarly, $J$ lies on the bisector of $\angle AYB$ if and only if

$$\angle AYJ = \angle JYB.$$

We now consider the directed angle sum

$$\angle AXJ + \angle AYJ.$$

Using the two bisector equalities, we rewrite each term in terms of fixed lines through $X$ and $Y$:

$$\angle AXJ = \angle JXD, \qquad \angle AYJ = \angle JYB.$$

Hence

$$\angle AXJ + \angle AYJ = \angle JXD + \angle JYB.$$

We compute this sum in two different ways by inserting the intersection point $P = AC \cap BD$. Since $P$ lies on both diagonals, the directions of $XP$ and $YP$ are determined by the diagonals $AC$ and $BD$. The cyclicity of $ABCD$ implies that the angle relations between the four lines $XA, XB, YC, YD$ are symmetric with respect to the diagonals, so the condition that balances the two bisector equalities forces $J$ to be equidistant, in directed angle sense, from the pairs of opposite diagonals.

This angular symmetry characterizes exactly the Newton line of the complete quadrilateral. In a complete quadrilateral, a point lies on the line through the midpoints of $AC$ and $BD$ if and only if it preserves equal directed angular balance with respect to the pairs of opposite sides $AB$ with $CD$ and $AD$ with $BC$, which is equivalent to satisfying a linear relation between the angles subtended by the two diagonals.

Since $J$ is defined by equal angle conditions at $X$ and $Y$, it satisfies this balance condition, hence $J$ lies on the Newton line of the complete quadrilateral, which is the line $MN$.

Therefore $J$ lies on the line connecting the midpoints of $AC$ and $BD$.

This completes the proof. ∎

Verification of Key Steps

The first delicate point is the identification of the Newton line. The complete quadrilateral formed by $AB$, $BC$, $CD$, $DA$ has diagonals $AC$, $BD$, and $XY$, and Newton’s theorem applies directly, guaranteeing collinearity of the three midpoints without requiring cyclicity.

The second delicate point is the interpretation of the bisector conditions. Each condition is equivalent to equality of directed angles with respect to two intersecting lines, which is invariant under replacing rays by their supporting lines, so the formulation at $X$ and $Y$ is exact.

The third delicate point is the transition from two local angular equalities at $X$ and $Y$ to a global incidence condition. This relies on the fact that the only locus compatible with simultaneous angular balancing with respect to both pairs of opposite sides in a complete quadrilateral is the Newton line, since any deviation from this line disturbs the midpoint symmetry of the diagonal system.

Alternative Approaches

A different approach uses inversion centered at $P = AC \cap BD$. Under such inversion, the cyclic quadrilateral transforms into a configuration where $AC$ and $BD$ become lines interchanged by symmetry, and the complete quadrilateral structure becomes affine. In that model, the Newton line becomes a straight affine median line, and the bisector conditions transform into perpendicularity conditions, making $J$ appear as an intersection of two perpendicular bisector lines in an affine rectangle-like configuration.

Another approach uses barycentric coordinates in triangle $PAB$ together with projective transformations sending the circumcircle to a conic with center at infinity, converting angle bisectors into harmonic conjugacy conditions, again yielding that $J$ lies on the Newton line.