# Ptolemy's inequality

Four points and their six distances. The points are not co-circular, so Ptolemy's inequality is strict for these points.

In Euclidean geometry, Ptolemy's inequality relates the six distances determined by four points in the plane or in a higher-dimensional space. It states that, for any four points A, B, C, and D, the following inequality holds:

It is named after the Greek astronomer and mathematician Ptolemy.

The four points can be ordered in any of three distinct ways (counting reversals as not distinct) to form three different quadrilaterals, for each of which the sum of the products of opposite sides is at least as large as the product of the diagonals. Thus the three product terms in the inequality can be additively permuted to put any one of them on the right side of the inequality, so the three products of opposite sides or of diagonals of any one of the quadrilaterals must obey the triangle inequality.[1]

As a special case, Ptolemy's theorem states that the inequality becomes an equality exactly when the four points lie in cyclic order on a circle. The inequality does not generalize from Euclidean spaces to arbitrary metric spaces. The spaces where it remains valid are called the Ptolemaic spaces; they include the inner product spaces, Hadamard spaces, and shortest path distances on Ptolemaic graphs.

## Assumptions and derivation

Ptolemy's inequality is often stated for a special case, in which the four points are the vertices of a convex quadrilateral, given in cyclic order.[2][3] However, the theorem applies more generally to any four points; it is not required that the quadrilateral they form be convex, simple, or even planar.

For points in the plane, Ptolemy's inequality can be derived from the triangle inequality by an inversion centered at one of the four points.[4][5] Alternatively, it can be derived by interpreting the four points as complex numbers, using the complex number identity

to construct a triangle whose side lengths are the products of sides of the given quadrilateral, and applying the triangle equality to this triangle.[6] One can also view the points as belonging to the complex projective line, express the inequality in the form that the absolute values of two cross-ratios of the points sum to at least one, and deduce this from the fact that the cross-ratios themselves add to exactly one.[7]

A proof of the inequality for points in three-dimensional space can be reduced to the planar case, by observing that for any non-planar quadrilateral, it is possible to rotate one of the points around the diagonal until the quadrilateral becomes planar, increasing the other diagonal's length and keeping the other five distances constant.[6] In spaces of higher dimension than three, any four points lie in a three-dimensional subspace, and the same three-dimensional proof can be used.

## Four concyclic points

Main article: Ptolemy's theorem

For four points in order around a circle, Ptolemy's inequality becomes an equality, known as Ptolemy's theorem:

In the inversion-based proof of Ptolemy's inequality, transforming four co-circular points by an inversion centered at one of them causes the other three to become collinear, so the triangle equality for these three points (from which Ptolemy's inequality may be derived) also becomes an equality.[5] For any other four points, Ptolemy's inequality is strict.

## In general metric spaces

A cycle graph in which the distances disobey Ptolemy's inequality

Ptolemy's inequality holds more generally in any inner product space,[1][8] and whenever it is true for a real normed vector space, that space must be an inner product space.[8][9]

For other types of metric space, the inequality may or may not be valid. A space in which it holds is called Ptolemaic. For instance, consider the four-vertex cycle graph, shown in the figure, with all edge lengths equal to 1. The sum of the products of opposite sides is 2. However, diagonally opposite vertices are at distance 2 from each other, so the product of the diagonals is 4, bigger than the sum of products of sides. Therefore, the shortest path distances in this graph are not Ptolemaic. The graphs in which the distances obey Ptolemy's inequality are called the Ptolemaic graphs and have a restricted structure compared to arbitrary graphs; in particular, they disallow induced cycles of length greater than three, such as the one shown.[10]

The Ptolemaic spaces include all CAT(0) spaces and in particular all Hadamard spaces. If a complete Riemannian manifold is Ptolemaic, it is necessarily a Hadamard space.[11]

## References

1. Schoenberg, I. J. (1940), "On metric arcs of vanishing Menger curvature", Annals of Mathematics, Second Series, 41: 715–726, doi:10.2307/1968849, MR 0002903.
2. Steele, J. Michael (2004), "Exercise 4.6 (Ptolemy's Inequality)", The Cauchy-Schwarz Master Class: An Introduction to the Art of Mathematical Inequalities, MAA problem books, Cambridge University Press, p. 69, ISBN 9780521546775.
3. Alsina, Claudi; Nelsen, Roger B. (2009), "6.1 Ptolemy's inequality", When Less is More: Visualizing Basic Inequalities, Dolciani Mathematical Expositions, 36, Mathematical Association of America, pp. 82–83, ISBN 9780883853429.
4. Apostol (1967) attributes the inversion-based proof to textbooks by R. A. Johnson (1929) and Howard Eves (1963).
5. Stankova, Zvezdelina; Rike, Tom, eds. (2008), "Problem 7 (Ptolemy's Inequality)", A Decade of the Berkeley Math Circle: The American Experience, MSRI Mathematical Circles Library, 1, American Mathematical Society, p. 18, ISBN 9780821846834.
6. Apostol, Tom M. (1967), "Ptolemy's inequality and the chordal metric", Mathematics Magazine, 40: 233–235, MR 0225213.
7. Silvester, John R. (2001), "Proposition 9.10 (Ptolemy's theorem)", Geometry: Ancient and Modern, Oxford University Press, p. 229, ISBN 9780198508250.
8. Giles, J. R. (2000), "Exercise 12", Introduction to the Analysis of Normed Linear Spaces, Australian Mathematical Society lecture series, 13, Cambridge University Press, p. 47, ISBN 9780521653756.
9. Schoenberg, I. J. (1952), "A remark on M. M. Day's characterization of inner-product spaces and a conjecture of L. M. Blumenthal", Proceedings of the American Mathematical Society, 3: 961–964, doi:10.2307/2031742, MR 0052035.
10. Howorka, Edward (1981), "A characterization of Ptolemaic graphs", Journal of Graph Theory, 5 (3): 323–331, doi:10.1002/jgt.3190050314, MR 625074.
11. Buckley, S. M.; Falk, K.; Wraith, D. J. (2009), "Ptolemaic spaces and CAT(0)", Glasgow Mathematical Journal, 51 (2): 301–314, doi:10.1017/S0017089509004984, MR 2500753.