In Euclidean geometry, Ptolemy's theorem is a relation between the four sides and two diagonals of a cyclic quadrilateral (a quadrilateral whose vertices lie on a common circle). The theorem is named after the Greek astronomer and mathematician Ptolemy (Claudius Ptolemaeus).[1] Ptolemy used the theorem as an aid to creating his table of chords, a trigonometric table that he applied to astronomy.
If the vertices of the cyclic quadrilateral are A, B, C, and D in order, then the theorem states that:
AC ⋅ BD=AB ⋅ CD+BC ⋅ AD
This relation may be verbally expressed as follows:
If a quadrilateral is cyclic then the product of the lengths of its diagonals is equal to the sum of the products of the lengths of the pairs of opposite sides.
Moreover, the converse of Ptolemy's theorem is also true:
In a quadrilateral, if the sum of the products of the lengths of its two pairs of opposite sides is equal to the product of the lengths of its diagonals, then the quadrilateral can be inscribed in a circle i.e. it is a cyclic quadrilateral.
Ptolemy's Theorem yields as a corollary a pretty theorem[2] regarding an equilateral triangle inscribed in a circle.
Given An equilateral triangle inscribed on a circle and a point on the circle.
The distance from the point to the most distant vertex of the triangle is the sum of the distances from the point to the two nearer vertices.
Proof: Follows immediately from Ptolemy's theorem:
qs=ps+rs ⇒ q=p+r.
Any square can be inscribed in a circle whose center is the center of the square. If the common length of its four sides is equal to
a
a\sqrt{2}
More generally, if the quadrilateral is a rectangle with sides a and b and diagonal d then Ptolemy's theorem reduces to the Pythagorean theorem. In this case the center of the circle coincides with the point of intersection of the diagonals. The product of the diagonals is then d2, the right hand side of Ptolemy's relation is the sum a2 + b2.
Copernicus – who used Ptolemy's theorem extensively in his trigonometrical work – refers to this result as a 'Porism' or self-evident corollary:
Furthermore it is clear (manifestum est) that when the chord subtending an arc has been given, that chord too can be found which subtends the rest of the semicircle.[3]
A more interesting example is the relation between the length a of the side and the (common) length b of the 5 chords in a regular pentagon. By completing the square, the relation yields the golden ratio:[4]
\begin{array}{rl} b ⋅ b =&a ⋅ a+a ⋅ b \ b2 -ab =&
| ||||
a |
-
ab | |
a2 |
=&
a2 | \ \left( | |
a2 |
b | |
a |
\right)2-
b | |
a |
+\left(
1 | |
2 |
\right)2=&1+\left(
1 | |
2 |
| ||||
\right) |
-
1 | |
2 |
\right)2=&
5 | |
4 |
\
b | |
a |
-
1 | |
2 |
=&\pm
\sqrt{5 | |
If now diameter AF is drawn bisecting DC so that DF and CF are sides c of an inscribed decagon, Ptolemy's Theorem can again be applied – this time to cyclic quadrilateral ADFC with diameter d as one of its diagonals:
ad=2bc
⇒ ad=2\varphiac
\varphi
⇒ c=
d | |
2\varphi |
.
whence the side of the inscribed decagon is obtained in terms of the circle diameter. Pythagoras's theorem applied to right triangle AFD then yields "b" in terms of the diameter and "a" the side of the pentagon [6] is thereafter calculated as
a=
b | |
\varphi |
=b\left(\varphi-1\right).
As Copernicus (following Ptolemy) wrote,
"The diameter of a circle being given, the sides of the triangle, tetragon, pentagon, hexagon and decagon, which the same circle circumscribes, are also given."[7]
The animation here shows a visual demonstration of Ptolemy's theorem, based on Derrick & Herstein (2012).[8]
Let ABCD be a cyclic quadrilateral.On the chord BC, the inscribed angles ∠BAC = ∠BDC, and on AB, ∠ADB = ∠ACB.Construct K on AC such that ∠ABK = ∠CBD; since ∠ABK + ∠CBK = ∠ABC = ∠CBD + ∠ABD, ∠CBK = ∠ABD.
Now, by common angles △ABK is similar to △DBC, and likewise △ABD is similar to △KBC.Thus AK/AB = CD/BD, and CK/BC = DA/BD;equivalently, AK⋅BD = AB⋅CD, and CK⋅BD = BC⋅DA.By adding two equalities we have AK⋅BD + CK⋅BD = AB⋅CD + BC⋅DA, and factorizing this gives (AK+CK)·BD = AB⋅CD + BC⋅DA.But AK+CK = AC, so AC⋅BD = AB⋅CD + BC⋅DA, Q.E.D.[9]
The proof as written is only valid for simple cyclic quadrilaterals. If the quadrilateral is self-crossing then K will be located outside the line segment AC. But in this case, AK-CK = ±AC, giving the expected result.
Let the inscribed angles subtended by
AB
BC
CD
\alpha
\beta
\gamma
R
AB=2R\sin\alpha
BC=2R\sin\beta
CD=2R\sin\gamma
AD=2R\sin(180\circ-(\alpha+\beta+\gamma))
AC=2R\sin(\alpha+\beta)
BD=2R\sin(\beta+\gamma)
\sin(\alpha+\beta)\sin(\beta+\gamma)=\sin\alpha\sin\gamma+\sin\beta\sin(\alpha+\beta+\gamma)
from which the factor
4R2
Now by using the sum formulae,
\sin(x+y)=\sin{x}\cosy+\cosx\siny
\cos(x+y)=\cosx\cosy-\sinx\siny
\begin{align} &\sin\alpha\sin\beta\cos\beta\cos\gamma+\sin\alpha\cos2\beta\sin\gamma\ +{}&\cos\alpha\sin2\beta\cos\gamma+\cos\alpha\sin\beta\cos\beta\sin\gamma. \end{align}
Here is another, perhaps more transparent, proof using rudimentary trigonometry.Define a new quadrilateral
ABCD'
A,B,C
ABCD
D'
|\overline{AD'}|=|\overline{CD}|
|\overline{CD'}|=|\overline{AD}|
ACD
C
A
A
C
D
ABCD'
ABCD
\alpha
\beta
\gamma
AB,BC
AD'
ABCD
ABCD'
\begin{align} Area(ABCD)&=
1 | |
2 |
AC ⋅ BD ⋅ \sin(\alpha+\gamma);\\ Area(ABCD')&=
1 | |
2 |
AB ⋅ AD' ⋅ \sin(180\circ-\alpha-\gamma)+
1 | |
2 |
BC ⋅ CD' ⋅ \sin(\alpha+\gamma)\ &=
1 | |
2 |
(AB ⋅ CD+BC ⋅ AD) ⋅ \sin(\alpha+\gamma). \end{align}
Choose an auxiliary circle
\Gamma
r
A'B'+B'C'=A'C'.
A'B',B'C'
A'C'
Note that if the quadrilateral is not cyclic then A', B' and C' form a triangle and hence A'B'+B'C' > A'C', giving us a very simple proof of Ptolemy's Inequality which is presented below.
C
A\mapstozA,\ldots,D\mapstozD
zA,\ldots,zD\inC
\zeta:= | (zA-zB)(zC-zD) |
(zA-zD)(zB-zC) |
\inC ≠ 0
\begin{align} \overline{AB} ⋅ \overline{CD}+\overline{AD} ⋅ \overline{BC} &=\left|zA-zB\right|\left|zC-zD\right|+\left|zA-zD\right|\left|zB-zC\right|\\ &=\left|(zA-zB)(zC-zD)\right|+\left|(zA-zD)(zB-zC)\right|\\ &=\left(\left|
(zA-zB)(zC-zD) | |
(zA-zD)(zB-zC) |
\right|+1\right)\left|(zA-zD)(zB-zC)\right|\\ &=\left(\left|\zeta\right|+1\right)\left|(zA-zD)(zB-zC)\right|\\ &\geq\left|(\zeta+1)(zA-zD)(zB-zC)\right|\\ &=\left|(zA-zB)(zC-zD)+(zA-zD)(zB-zC)\right|\\ &=\left|(zA-zC)(zB-zD)\right|\\ &=\left|zA-zC\right|\left|zB-zD\right|\\ &=\overline{AC} ⋅ \overline{BD} \end{align}
\zeta
zA,\ldots,zD
C
\zeta\inR>0
From the polar form of a complex number
z=\vertz\vertei\arg(z)
\begin{align} \arg(\zeta)&=\arg
(zA-zB)(zC-zD) | |
(zA-zD)(zB-zC) |
\\ &=\arg(zA-zB)+\arg(zC-zD)-\arg(zA-zD)-\arg(zB-zC)\pmod{2\pi}\\ &=\arg(zA-zB)+\arg(zC-zD)-\arg(zA-zD)-\arg(zC-zB)-\arg(-1)\pmod{2\pi}\\ &=-\left[\arg(zC-zB)-\arg(zA-zB)\right]-\left[\arg(zA-zD)-\arg(zC-zD)\right]-\arg(-1)\pmod{2\pi}\\ &=-\angleABC-\angleCDA-\pi\pmod{2\pi}\\ &=0 \end{align}
\pi
\angleABC
\angleCDA
\arg\left[(zA-zB)(zC-zD)\right]=\arg\left[(zA-zD)(zB-zC)\right]=\arg\left[(zA-zC)(zB-zD)\right]\pmod{2\pi}
C
\arg
\overline{AB} ⋅ \overline{CD}+\overline{AD} ⋅ \overline{BC}=\overline{AC} ⋅ \overline{BD}
(zA-zB)(zC-zD)+(zA-zD)(zB-zC)=(zA-zC)(zB-zD).
In the case of a circle of unit diameter the sides
S1,S2,S3,S4
\theta1,\theta2,\theta3
\theta4
\sin\theta1\sin\theta3+\sin\theta2\sin\theta4=\sin(\theta1+\theta2)\sin(\theta1+\theta4)
Applying certain conditions to the subtended angles
\theta1,\theta2,\theta3
\theta4
\theta1+\theta2+\theta3+\theta
\circ | |
4=180 |
Let
\theta1=\theta3
\theta2=\theta4
\theta1+\theta2=\theta3+\theta
\circ | |
4=9 0 |
\sin\theta1\sin\theta3+\sin\theta2\sin\theta4=\sin(\theta1+\theta2)\sin(\theta1+\theta4)
2(\theta | |
\sin | |
1+\theta |
2)
2\theta | |
\sin | |
1=1 |
Let
\theta2=\theta4
2x
x=S2\cos(\theta2+\theta3)
It will be easier in this case to revert to the standard statement of Ptolemy's theorem:
\begin{array}{lcl} S1S3+S2S4={\overline{AC}} ⋅ {\overline{BD}}\\ ⇒ S1S3+{S
2={\overline{AC}} | |
2} |
2\\ ⇒ S1[S1-2S2\cos(\theta2+\theta3)]+{S
2={\overline{AC}} | |
2} |
2\\ ⇒
2-2S | |
{S | |
1 |
S2\cos(\theta2+\theta
2\\ \end{array} | |
3)={\overline{AC}} |
The cosine rule for triangle ABC.
Let
\theta1+\theta2=\theta3+\theta
\circ. | |
4=90 |
Then
\sin\theta1\sin\theta3+\sin\theta2\sin\theta4=\sin(\theta3+\theta2)\sin(\theta3+\theta4)
Therefore,
\cos\theta2\sin\theta3+\sin\theta2\cos\theta3=\sin(\theta3+\theta2) x 1
Formula for compound angle sine (+).[11]
Let
\circ | |
\theta | |
1=90 |
\theta2+(\theta3+\theta
\circ | |
4)=90 |
\sin\theta1\sin\theta3+\sin\theta2\sin\theta4=\sin(\theta3+\theta2)\sin(\theta3+\theta4)
\sin\theta3+\sin\theta2\cos(\theta2+\theta3)=\sin(\theta3+\theta2)\cos\theta2
\sin\theta3=\sin(\theta3+\theta2)\cos\theta2-\cos(\theta2+\theta3)\sin\theta2
Formula for compound angle sine (−).
This derivation corresponds to the Third Theoremas chronicled by Copernicus following Ptolemy in Almagest. In particular if the sides of a pentagon (subtending 36° at the circumference) and of a hexagon (subtending 30° at the circumference) are given, a chord subtending 6° may be calculated. This was a critical step in the ancient method of calculating tables of chords.[12]
This corollary is the core of the Fifth Theorem as chronicled by Copernicus following Ptolemy in Almagest.
Let
\circ | |
\theta | |
3=90 |
\theta1+(\theta2+\theta
\circ | |
4)=90 |
\sin\theta1\sin\theta3+\sin\theta2\sin\theta4=\sin(\theta3+\theta2)\sin(\theta3+\theta4)
\cos(\theta2+\theta4)+\sin\theta2\sin\theta4=\cos\theta2\cos\theta4
\cos(\theta2+\theta4)=\cos\theta2\cos\theta4-\sin\theta2\sin\theta4
Formula for compound angle cosine (+)
Despite lacking the dexterity of our modern trigonometric notation, it should be clear from the above corollaries that in Ptolemy's theorem (or more simply the Second Theorem) the ancient world had at its disposal an extremely flexible and powerful trigonometric tool which enabled the cognoscenti of those times to draw up accurate tables of chords (corresponding to tables of sines) and to use these in their attempts to understand and map the cosmos as they saw it. Since tables of chords were drawn up by Hipparchus three centuries before Ptolemy, we must assume he knew of the 'Second Theorem' and its derivatives. Following the trail of ancient astronomers, history records the star catalogue of Timocharis of Alexandria. If, as seems likely, the compilation of such catalogues required an understanding of the 'Second Theorem' then the true origins of the latter disappear thereafter into the mists of antiquity but it cannot be unreasonable to presume that the astronomers, architects and construction engineers of ancient Egypt may have had some knowledge of it.
See main article: Ptolemy's inequality.
The equation in Ptolemy's theorem is never true with non-cyclic quadrilaterals. Ptolemy's inequality is an extension of this fact, and it is a more general form of Ptolemy's theorem. It states that, given a quadrilateral ABCD, then
\overline{AB} ⋅ \overline{CD}+\overline{BC} ⋅ \overline{DA}\ge\overline{AC} ⋅ \overline{BD}
where equality holds if and only if the quadrilateral is cyclic. This special case is equivalent to Ptolemy's theorem.
Ptolemy's theorem gives the product of the diagonals (of a cyclic quadrilateral) knowing the sides, the following theorem yields the same for the ratio of the diagonals.[13]
AC | = | |
BD |
AB ⋅ DA+BC ⋅ CD | |
AB ⋅ BC+DA ⋅ CD |
Proof: It is known that the area of a triangle
ABC
R
l{A}=
AB ⋅ BC ⋅ CA | |
4R |
Writing the area of the quadrilateral as sum of two triangles sharing the same circumscribing circle, we obtain two relations for each decomposition.
l{A}tot=
AB ⋅ BC ⋅ CA | |
4R |
+
CD ⋅ DA ⋅ AC | |
4R |
=
AC ⋅ (AB ⋅ BC+CD ⋅ DA) | |
4R |
l{A}tot=
AB ⋅ BD ⋅ DA | |
4R |
+
BC ⋅ CD ⋅ DB | |
4R |
=
BD ⋅ (AB ⋅ DA+BC ⋅ CD) | |
4R |
Equating, we obtain the announced formula.
Consequence: Knowing both the product and the ratio of the diagonals, we deduce their immediate expressions:
\begin{align} AC2&=AC ⋅ BD ⋅
AC | |
BD |
=(AB ⋅ CD+BC ⋅ DA)
AB ⋅ DA+BC ⋅ CD | |
AB ⋅ BC+DA ⋅ CD |
\\[8pt] BD2&=
AC ⋅ BD | |||
|
=(AB ⋅ CD+BC ⋅ DA)
AB ⋅ BC+DA ⋅ CD | |
AB ⋅ DA+BC ⋅ CD |
\end{align}