Chapter V: Desargues’s Theorem¶
Kapitel V: Der Desarguessche Satz — David Hilbert (1899 / 1902) — §§ 22–30
Chapters V and VI pivot from metric Euclidean geometry to projective-affine foundations, investigating what happens when Congruence (`IV`) is dropped altogether:
§ 22 — Desargues’s Theorem (Theorem 32): Formulates the affine Desargues's Theorem (Theorem 32) for perspective triangles with parallel sides, observing that it follows effortlessly either from 3D spatial incidence (I 1–7, II, III) or from 2D planar congruence (I 1–2, II–IV via Chapter III coordinates).
§ 23 — Hilbert’s Non-Desarguesian Plane (Theorem 33): Constructs a celebrated planar countermodel inside the Euclidean $XY$-plane: straight lines cutting the ellipse $x^2 + 4y^2 = 1$ have their interior chord replaced by the circular arc passing through the two boundary intersections $P, Q$ and the exterior point $F(\tfrac{3}{2}, 0)$ (since $\tfrac{3}{2} > \tfrac{\sqrt{7}}{2}$, no circle meeting the ellipse four times can pass through $F$). This deformed plane satisfies all planar axioms `I 1–2, II, III, IV 1–5, V` while violating Desargues’s Theorem—proving planar Desargues unprovable without either 3D spatial incidence (I 3–7) or triangle congruence (IV 6)!
§§ 24–26 — Congruence-Free Segment Calculus: Fixes two intersecting axes $OE, OE'$ and a unit line $EE'$ in any plane satisfying I 1–2, II, III + Desargues (Theorem 32), defining addition $c = a + b$ and multiplication $c = ab$ purely by parallel projections. Using intricate multi-triangle Desargues configurations (Figures 41–45, drafted with Hans von Schaper), Hilbert proves commutative and associative addition (§ 25), associative multiplication $a(bc) = (ab)c$, and both left and right distributive laws (§ 26)—all without Congruence!
§§ 27–30 — Skew-Field Coordinatization and 3D Embeddability (Theorems 34–35): Proves every line has a left-linear equation $ax + by + c = 0$ (Theorem 34, whereas $xa + yb + c = 0$ need not be a line when multiplication is non-commutative!), shows the segments form an ordered associative division ring (Desarguessches Zahlensystem, rules 1–11, 13–16 of § 13), reconstructs 3D space over any Desarguesian division ring (§ 29), and crowns the chapter with Theorem 35: Desargues’s Theorem is the necessary and sufficient condition for a planar geometry (I 1–2, II, III) to be embeddable as a plane inside a 3D space (I, II, III).
Sections¶
§ 22: Desargues’s Theorem and Its Demonstration for Plane Geometry by Aid of the Axioms of Congruence
Of the axioms given in §§ 1–8, pp. 1–26, those of groups II–V are in part linear and in part plane axioms. Axioms 3–7 of group I are the only space axioms. In order to show clearly the significance of these axioms of space, let us assume a plane geometry and investigate, in general, the conditions for which this plane geometry may be regarded as a part of a geometry of space in which at least the axioms of groups I–III are all fulfilled.
§ 23: The Impossibility of Demonstrating Desargues’s Theorem for the Plane Without the Help of the Axioms of Congruence
We shall now investigate the question whether or no in plane geometry Desargues’s theorem may be deduced without the assistance of the axioms of congruence. This leads us to the following result:
§ 24: Introduction of an Algebra of Segments Based Upon Desargues’s Theorem and Independent of the Axioms of Congruence
In order to see fully the significance of Desargues’s theorem (theorem 32), let us take as the basis of our consideration a plane geometry where all of the axioms I 1–2, II–III are valid, that is to say, where all of the plane axioms of the first three groups hold, and then introduce into this geometry, in the following manner, a new algebra of segments independent of the axioms of congruence.
§ 25: The Commutative and the Associative Law of Addition for Our New Algebra of Segments
In this section, we shall investigate the laws of operation, as enumerated in § 13, in order to see which of these hold for our new algebra of segments, when we base our considerations upon a plane geometry in which axioms I 1–2, II–III are all fulfilled, and, moreover, in which Desargues’s theorem also holds.
§ 26: The Associative Law of Multiplication and the Two Distributive Laws for the New Algebra of Segments
The associative law of multiplication
§ 27: Equation of the Straight Line, Based Upon the New Algebra of Segments
In §§ 24–26, we have introduced into the plane geometry an algebra of segments in which the commutative law of addition and that of multiplication, as well as the two distributive laws, hold. This was done upon the assumption that the axioms cited in § 24, as also the theorem of Desargues, were valid. In this section, we shall show how an analytical representation of the point and straight line in the plane is possible upon the basis of this algebra of segments.
§ 28: The Totality of Segments Regarded as a Complex Number System
We see immediately that, for the new algebra of segments established in Section 24, theorems 1–6 of Section 13 are fulfilled. Moreover, by aid of Desargues’s theorem, we have already shown in Sections 25 and 26 that the laws 7–11 of operation, as given in Section 13, are all valid in this algebra of segments. With the single exception of the commutative law of multiplication, therefore, all of the theorems of connection hold.
§ 29: Construction of a Geometry of Space by Aid of a Desarguesian Number System
Suppose we have given a desarguesian number system $D$. Such a system makes possible the construction of a geometry of space in which axioms I, II, III are all fulfilled.
§ 30: Significance of Desargues’s Theorem
If, in a plane geometry, axioms I, 1–2, II, III are all fulfilled and, moreover, if the theorem of Desargues holds, then, according to §§ 24–28, it is always possible to introduce into this geometry an algebra of segments to which the laws 1–11, 13–16 of § ! 13 are applicable. We will now consider the totality of these segments as a complex number system and construct, upon the basis of this system, a geometry of space, in accordance with § ! 29, in which all of the axioms I, II, III hold.