Circle Angle Theorems

Fix a chord, move points around the circumference, and watch one hidden choice — which arc the angle intercepts — generate four classical circle theorems.

Loading interactive simulation...

The centre sees twice what the circumference sees 🖖

Join O to A, B, and P. Each radius makes an isosceles triangle, so its two base angles are equal. Chasing those repeated base angles around P leaves the angle at O exactly twice ∠APB. The green arc matters: the central angle must stand on the same arc as the inscribed angle, which is why it can be reflex rather than the smaller angle AOB.

The invariant survives motion, not an endpoint crossing 🖖

Keep A and B fixed and slide P along one arc. The triangle changes shape continuously, but ∠APB does not move because it keeps intercepting the same opposite arc. Cross A or B and the intercepted arc switches to the rest of the circle. The new angle is supplementary to the old one — a jump in meaning caused by crossing a boundary, not a numerical glitch.

Thales is the one-line special case 🖖

Make AB a diameter. The relevant central angle is then a straight 180°, whichever semicircle contains P, and the angle at P is half of it: 90°. This also runs backwards: if a triangle has a right angle, its three vertices lie on the circle whose diameter is the hypotenuse. The midpoint of the hypotenuse is therefore equally distant from all three vertices.

A cyclic quadrilateral is two arcs that make a full circle 🖖

Place P and Q on opposite sides of AB. Angle APB sees one A-to-B arc and angle AQB sees the other. Those arcs total 360°, so their half-angles total 180°. That is the entire cyclic-quadrilateral theorem. Its converse is useful too: if a quadrilateral's opposite angles sum to 180°, its four vertices lie on one circle.

Problems solved in full

  1. Proving the centre sees twice, with one line the tool never draws 6 steps

    A and B sit at 330° and 110° on the circle, and P is at 220°. The panel says the angle at P is 70° and the angle at the centre is 140°. Prove the factor of two, rather than reading it — and then find what that proof quietly assumes about where P is.

    O A B P D α β
    1. Measure the arc first. Going anticlockwise from A at 330° round to B at 110° sweeps 140°, and that arc is the one P does not sit on. The angle at the centre standing on it is 140° by definition, because a central angle simply is its arc.

    2. Now the construction, and it is the whole proof: draw the line from P through the centre O and let it leave the circle at D. Nothing else is added.

    3. OA and OP are both radii, so triangle OAP is isosceles and the angles at A and at P are equal. Call that angle α, so ∠OPA = ∠OAP = α.

    4. The angle ∠AOD is exterior to triangle OAP at O, and an exterior angle equals the sum of the two interior angles it does not touch: ∠AOD = α + α = 2α.

    5. The same argument on the other side gives ∠DOB = 2β, where β = ∠OPB. Add the two, which works while OP produced meets the circle on the arc AB: ∠AOB = 2α + 2β = 2(α + β), and α + β is exactly ∠APB.

    6. So the central angle is twice the inscribed one, and here 2 × 70° = 140°. Read back what the proof used: two radii, one isosceles triangle each side, one exterior angle. No length, no coordinate, and no measurement of the arc.

    Answer

    140° = 2 × 70°, and the panel prints 70.0000° wherever you put P. Drag P from 150° to 290° and the two isosceles triangles change shape in opposite directions, their base angles absorbing the difference exactly rather than approximately. That is what an invariant looks like from the inside. Past 290° the drawing changes and the proof needs its second case: OP produced now crosses the circle beyond B instead of on the arc AB, the base angles come out 75° and 5°, and the two central angles have to be subtracted rather than added. The angle at P never notices. Thales then costs nothing: make AB a diameter and the central angle is a straight line, 180°, so every inscribed angle is 90° and a triangle in a semicircle is right-angled wherever you put its apex. It is not a separate theorem, it is this one with the arc set to half the circle. The construction line in the sketch is the part the panel cannot show you, because the panel draws the angles and this draws the reason.

  2. Where to stand for the widest angle 7 steps

    Both marked points see the chord AB at 60°. Work out what a point off the circle would see instead, then decide where a kicker should stand to get the widest view of the posts.

    1. A and B sit at 320° and 80°, so the arc between them that P does not stand on measures 120°. Half of it is 60°, and Q reads the same because it stands on the same arc as P.

    2. Now the chord itself. The central angle on that arc is twice the inscribed one, and a chord is 2R sin of half the central angle, so AB = 2 sin 60° = √3 on this circle of radius 1.

    3. Read that line backwards. Hold the chord and let the circle vary: sin of the inscribed angle is AB ÷ 2R, so a smaller circle through A and B gives a larger angle, and the smallest one has AB as a diameter and gives 90°. Which circle you stand on decides everything.

    4. Off the circle the angle is no longer 60°. From inside, the two lines cut a second arc as well and the angle is half the sum of the two: a little way in from the arc it reads about 71.6°. From outside it is half the difference, about 51.4°. The circle is the boundary between seeing more and seeing less.

    5. That settles the kick. Rugby posts stand 5.6 m apart; a try is scored 10 m out from the near post, so from the foot of the run-back line the posts are 10 m and 15.6 m away. Every point that sees them at the same angle lies on one circle through both.

    6. Each point on the run-back line lies on exactly one such circle, and the angle is largest on the smallest circle that reaches the line at all — the one tangent to it. The touch point is a tangent from the foot, and a tangent length is the geometric mean of the two secant pieces: x² = 10 × 15.6 = 156.

    7. So x = √156 = 12.49 m, where the angle is 12.64°. At the 10 m mark it is 12.34°, at 20 m it has fallen to 11.39°, and at 30 m to 9.04°.

    Answer

    Stand 12.49 m back, the geometric mean of the two distances rather than their average of 12.8. Backing off further gives away angle as surely as standing too close does, and the maximum is a genuine interior one rather than an endpoint. The 12.8 m you did not use is not wasted either: it is the radius of the circle you end up standing on. All of it is the same-segment theorem read the other way round, asking which circle through A and B passes through the point you are on instead of what angle a point on this circle sees.

Learning path

The triangle rules, and where they come from

Leads to The unit circle the central angle as twice the inscribed angle, equal angles in the same segment, and Thales' right angle.

References (2)
  • the central-angle, same-segment, cyclic-quadrilateral and semicircle theorems Euclid. Elements, Book III, Propositions 20, 21, 22 and 31.
  • a modern synthetic-geometry treatment of circle theorems Coxeter, H. S. M. and Greitzer, S. L. (1967). Geometry Revisited. Mathematical Association of America.

Example problems

  • the angle that won't move - A 140° central arc produces a 70° angle at the circumference — exactly half, always.
  • two points · same angle - Slide P anywhere on the same arc: the angle stays at 60°, because the arc it looks at never changes.
  • a diameter forces 90° - A diameter subtends 90° from either side. Thales, and the one case you can check by eye.
  • opposites make 180° - Cross the chord and equality becomes supplementarity: 100° on one side, 80° on the other.
  • the hidden reflex arc - The angle is an obtuse 110° because it intercepts the 220° reflex arc, not the minor one.