The Timpson Lyceum seal

The Timpson Lyceum · Geometry (2024)

Comparing and Measuring Segments – Lecture 6 (Geometry)

Comparing and Measuring Segments – Lecture 6

Overview

This lecture turns the qualitative idea of congruence into a quantitative one: it shows how to assign a length to a segment and why every segment has a well-defined length. Three big ideas organize the chapter: a length is a number together with a unit; once you fix a unit, every segment has a length in that unit; and conversely, for any positive real number there is a segment of that length. Getting there requires a short tour of the number systems (naturals, integers, rationals, reals) and the fact that every real number has a decimal expansion. Two new postulates — the Archimedean postulate and the Completeness postulate — make the measuring process rigorous, and a handful of theorems tie length back to the congruence relation studied earlier.

Key Concepts

Recap of congruence. Congruence is an undefined term whose meaning comes from postulates. It is a two-term relation (a predicate that returns true/false for a pair of segments). The key facts carried in from the previous chapter:

The three big ideas of this chapter.

Number systems. Each system fills a gap left by the previous one:

Repeating decimals and rationals coincide. Every real number has a decimal expansion. A number is rational if and only if its decimal expansion eventually repeats (terminating decimals count — they end in repeating 00s). Irrationals are exactly the reals whose decimals never settle into a repeating pattern. Doing long division always forces a remainder to recur, so every rational repeats; conversely the "multiply, shift, subtract" trick converts any repeating decimal back into a fraction. The reals are precisely all numbers with a decimal expansion.

Inequalities on the reals (distinct from the segment order, despite the shared symbols):

The measuring process. To measure AB‾\overline{AB}: pick a unit — any reference segment (a meter is the standard rod in France; a foot was once the king's foot). Lay copies of the unit end to end until you land on BB or overshoot; that whole-number count is the integer part. Then refine with a smaller sub-unit (a tenth of the unit is natural but any subdivision works), repeating to get successive decimal digits. The result is the segment's decimal expansion in that unit. Some lengths never terminate: the diagonal of a unit square is 2\sqrt{2} (by the Pythagorean theorem), an infinite non-repeating expansion — Aristotle called such pairs incommensurable.

Postulate — Archimedean. If AB‾\overline{AB} lies on line LL and CD‾\overline{CD} is any unit segment, then there exist a natural number nn and points A=A0,A1,A2,…,AnA=A_0, A_1, A_2, \dots, A_n on LL with CD‾≅A0A1‾≅A1A2‾≅⋯≅An−1An‾,\overline{CD} \cong \overline{A_0A_1} \cong \overline{A_1A_2} \cong \cdots \cong \overline{A_{n-1}A_n}, and either An=BA_n = B (the unit fits exactly) or BB lies between An−1A_{n-1} and AnA_n (you stop one step past BB). This guarantees the second big idea: any segment can be measured by repeatedly laying down a unit.

Postulate — Completeness. For any real number xx and any point AA on a line LL, there exists a point BB on LL such that AB‾=x\overline{AB} = x (in the chosen units). This is the third big idea: every real length is realized by some segment.

Notation. AB‾\overline{AB} with no further mark denotes the segment (a geometric figure); written as a length AB‾\overline{AB} it denotes the number that is its measure.

Length theorems.

Worked Example

Goal: show that the repeating decimal 0.9‾0.\overline{9} equals 11, i.e. that this repeating decimal has a rational (in fact integer) value. The method — set the number equal to xx, shift the decimal by multiplying by a power of 1010, subtract to cancel the repeating tail, then solve — is the general recipe for converting any repeating decimal to a fraction.

x=0.9‾10x=9.9‾10x−x=9.9‾−0.9‾9x=9x=1\begin{aligned} x &= 0.\overline{9} \\ 10x &= 9.\overline{9} \\ 10x - x &= 9.\overline{9} - 0.\overline{9} \\ 9x &= 9 \\ x &= 1 \end{aligned}

The same procedure handles an arbitrary repeating block. For x=137.2164‾x = 137.\overline{2164}, the block has length 44, so multiply by 10410^4 to align the tails:

10000x−x=1372164.2164‾−137.2164‾9999 x=1372027x=13720279999\begin{aligned} 10000x - x &= 1372164.\overline{2164} - 137.\overline{2164} \\ 9999\,x &= 1372027 \\ x &= \frac{1372027}{9999} \end{aligned}

Every repeating decimal collapses to a fraction this way, confirming the one-to-one correspondence between rationals and repeating decimals.

Summary