Theme

Things that cannot be done — page 10

Results that close a door rather than open one — and the peculiar difficulty of drawing a picture of something that does not exist.
The most nearly equal dissections of a square into 3, 5 and 7 triangles found. 3 triangles, areas 0.25000, 0.50000, 0.25000, spread 2.5000e-1; 5 triangles, areas 0.19098, 0.21353, 0.19098, 0.19098, 0.21353, spread 2.2543e-2; 7 triangles, areas 0.14151, 0.14458, 0.14151, 0.14458, 0.14151, 0.14174, 0.14458, spread 3.0711e-3. Computation

How nearly equal an odd number of triangles can be

A square cannot be cut into an odd number of triangles of equal area. It can be cut into five triangles whose areas differ by about two hundredths, seven that differ by three thousandths, nine by a ten-thousandth and a half — the spread falling by a factor of seven or more with every two triangles added, closing on equality and never reaching it. A search for the closest finds that the best five it can make have the golden ratio in their areas.

The factorials, and three smooth curves through all of them. Γ(x+1), Γ(x+1)(1 + 0.5 sin² πx) and the degree-six interpolating polynomial through 0!…6!, on a log scale over 0 ≤ x ≤ 5.4; at x = 1/2: 0.88623, 1.32934, -3.58301. Analysis

Only one curve through the factorials bends the right way

Infinitely many smooth curves pass through 1, 1, 2, 6, 24, 120, and some of them even obey the factorial's own rule, f(x + 1) = (x + 1)·f(x), at every x. Ask that the logarithm of the curve bend upwards everywhere and exactly one survives — Euler's integral. A wiggle of any size breaks the condition somewhere, the smaller the wiggle the further out, and the proof that nothing else survives is a squeeze that computes √π along the way.

Langford pairings of order 3, 4 and 7. Rows of boxes with arcs joining equal numbers: order 3: 3 1 2 1 3 2; order 4: 4 1 3 1 2 4 3 2; order 7: 7 3 6 2 5 3 2 4 7 6 5 1 4 1. Discrete

A sum that forbids half the pairings

Put two 1s, two 2s, …, two ns in a row so that between the two ks there are exactly k other numbers. For n = 3 there is one way, 312132; for n = 4 one way; for n = 5 and 6 none at all, and the reason is a single sum — adding up the positions of every entry in two ways forces n(3n − 1)/2 to be even. A search confirms the count forbids nothing that exists and finds 26 arrangements at n = 7 and 108,144 at n = 12, but the sum is all anyone knows about why.

Malfatti's three circles against the three largest, one at a time. A triangle with angles 40, 70, 70 degrees: Malfatti's circles cover 69.1% of it, the greedy choice of incircle then largest remaining circles covers 73.7%. Geometry

Three circles that touch and are not the largest

In 1803 Gian Francesco Malfatti asked how to cut three round columns from a triangular block of marble with as little waste as possible, and answered with three circles each touching the other two and two sides. The circles exist in every triangle and are found by solving three equations. They are never the answer to his question: the plain greedy rule — the inscribed circle first, then the largest circle that still fits, twice — always does better, by 1.4 per cent on the equilateral triangle and by almost double on a thin one.

Four dominoes and a match: the same word along the top and the bottom. Dominoes b/ca, a/ab, ca/a, abc/c; the sequence 2, 1, 3, 2, 4 spells abcaaabc on both rows. Logic

Dominoes that no algorithm can match

Emil Post's puzzle has no machines and no programs in it: a handful of dominoes, each with a string on top and a string underneath, and the question whether some row of them, repeats allowed, spells the same word along both edges. No algorithm can answer it for every set of dominoes, because a set can be built whose only matches are a machine's run written out twice, one step out of phase — and three dominoes with strings of three characters already need thirty-six in a row before the edges agree.

No pair of weights modulo ten catches everything. Two 10 × 10 grids over pairs of alternating weights modulo ten, shaded by the share of single errors and of swaps each catches. No pair is full in both; weights 3 and 1, ringed, catch every single error and 88.9% of swaps. Computation

Ten digits need a symmetry that does not commute

A check digit should catch one wrong digit and two neighbours swapped. Over eleven symbols a weighted sum does both; over the ten decimal digits no weighted sum can, no scheme of any shape built on adding modulo ten can, and the reason is the same as the reason Euler's thirty-six officers cannot be paraded. What works is the ten symmetries of a pentagon, which do not commute.

Three truth values, two of which count as holding. Truth tables of negation, conjunction and disjunction in the logic of paradox, with the values that count as holding — both and true — shaded. Logic

A contradiction that stays where it is

In classical logic one contradiction proves everything: from p and not-p, any q at all. Add a third truth value — both true and false — and count it as holding, and the explosion stops. The classical laws all survive as laws; what goes is exactly the reasoning that carries a contradiction somewhere it was not. Run the other way, the same three tables make a logic of gaps instead of gluts.

Pascal's triangle mod 2, and the room for the 2-dimensional projective space. Pascal's triangle with odd entries filled, rows 0 to 15, with row 3 marked as the tangent ledger 1 + a + a² and row 1 as the normal ledger 1 + a. Topology

The room a projective space needs, read off Pascal's triangle

The projective plane cannot sit in three-dimensional space without crossing itself, and the reason can be written as arithmetic: a polynomial that records how a shape twists, which a room must cancel. For the n-dimensional projective space that polynomial is a row of Pascal's triangle read mod 2, its inverse is another row, and the inverse's last term says how many extra dimensions the room must have — exactly enough, at every power of two.

The twenty maps x ↦ ax + b modulo 5, as permutations of five roots. A four-by-five grid of pentagons; in each, arrows show where the map ax + b modulo 5 sends each of the five corners. Rows are the slopes 1, 4, 2, 3; columns the shifts 0 to 4. Algebra

The quintics that have a formula

No formula solves every equation of degree five, and yet x⁵ − 2 is solved by a fifth root and x⁵ − 5x + 12 by a longer expression of the same kind. A quintic can be solved by radicals exactly when the symmetries of its roots fit inside one group of twenty — the maps x ↦ ax + b on the numbers modulo 5 — and two exact tests on its coefficients say whether they do.

Cardano's formula for x³ − 3x + 1, drawn: three real roots from conjugate cube roots. The complex plane with u³ = −0.50 + 0.866i and its conjugate, their three cube roots each, and the vertical pairings whose sums are the real roots −1.879, 0.347, 1.532. Algebra

Three real roots and no real radicals

The cubic x³ − 3x + 1 has three real roots, and Cardano's formula reaches every one of them by way of the cube roots of a complex number. That detour cannot be removed: Hölder proved in 1891 that no expression built from real radicals gives a root of an irreducible cubic whose roots are all real. The proof is three lines of the Galois correspondence, and its general form says that real radicals reach all-real roots only when square roots alone would.

A closed knight's tour of the 8 × 8 board. A 8 by 8 chequered board with a closed polygon joining the centres of all 64 squares in the order a knight visits them. Discrete

The boards a knight can tour

A knight can visit every square of a chessboard once and land a move from where it began. On a 5 × 5 board it cannot, on a 4 × 100 board it cannot, and on a 3 × 8 board it cannot, though on 3 × 10 it can in sixteen ways. Allen Schwenk found in 1991 the complete list of rectangles without a closed tour, and the reasons on it are of two kinds — a colouring that counts squares, and a second colouring that a tour would have to make agree with the first.

The five lunes that can be squared. Five crescents with arc ratios 2:1, 3:1, 3:2, 5:1, 5:3 and their equal-area kites; θ = 45.000°, 34.265°, 26.812°, 23.439°, 16.794°. Geometry

The five crescents that can be squared

The circle cannot be squared, but some regions bounded by circular arcs can: Hippocrates of Chios found a crescent with exactly the area of a triangle around 440 BC. The trick is to make two multiples of π cancel, and it works for a crescent whose two arcs turn through angles in the ratio m to n exactly when an angle defined by a polynomial can be drawn with ruler and compass. That happens for five ratios and no others.

The Lorenz orbits LR and LLR, linked. Projection of the periodic orbits LR, LLR of the Lorenz system, linking number 1. Topology

Every loop of the butterfly links every other

The Lorenz equations have infinitely many periodic orbits, each a closed loop winding round the two lobes in its own order. Name each loop by that order — LR, LLR, LLRLR — and the linking number of any two can be read off the names, without solving anything: it is half the number of times a left-going strand passes a right-going one. Every such crossing has the same sign, so every pair of loops links, positively. Computing forty-five pairs from the equations themselves, the Gauss integral lands on the predicted whole number every time.

A trefoil, and how far its tangent turns. The (2, 3) torus knot with tube radius 0.45; total curvature 16.8556 = 5.3653π. Topology

A knot must turn twice, and a little more

Follow a closed curve and add up how far its direction turns. Any closed curve turns through at least one full circle; a knotted one turns through more than two. The proof is a count of hilltops — the total turning equals 2π times the number of highest points the curve has, averaged over every direction to look from — and a curve with a single hilltop in any direction cannot be knotted.

Every swap-proof scrambling, graded on twins and jump swaps. 34040 permutations; twin errors caught range 50–86 of 90, jump transpositions 600–848 of 900; Verhoeff's at (848, 86). Computation

The best of thirty-four thousand scramblings

Verhoeff's check digit multiplies digits in the symmetry group of a pentagon after scrambling each one a different number of times, and 34,040 scramblings make it catch every swap of neighbours. Graded on the rarer errors, none of them catches everything, the best catch 86 of 90 doubled-digit slips and 848 of 900 swaps across a digit — and the scrambling Verhoeff published in 1969 is one of the forty that are best at both.

Every light on, on a board where most patterns are out of reach. A 5×5 Lights Out press pattern of 15 presses that turns all 25 lights on; the board's quiet space has dimension 2. Computation

Every light on, from every light off

In Lights Out, pressing a light toggles it and its neighbours. On a five-by-five board only one pattern in four can be reached from all off — and yet all on is always one of them. The same is true on every board and on every network of lights whatever, provided the influence runs both ways. The proof is a single observation about symmetric matrices over a field with two elements, and it fails, as a three-arrow example shows, the moment influence runs one way.

Every finite kaleidoscope on up to eight mirrors. 1: A₁; 2: A₂ B₂ H₂ G₂; 3: A₃ B₃ H₃; 4: D₄ A₄ B₄ H₄ F₄; 5: D₅ A₅ B₅; 6: E₆ D₆ A₆ B₆; 7: E₇ D₇ A₇ B₇; 8: E₈ D₈ A₈ B₈. Geometry

One determinant for every kaleidoscope

Three mirrors close up into a kaleidoscope when the angles of their triangle add to more than half a turn. In more dimensions there is no triangle to add up, but there is a matrix — the inner products of the mirrors' directions — and the mirrors close up exactly when it is positive definite. Trying every arrangement of up to nine mirrors against that one test finds Coxeter's list: three families in every dimension, a fork that runs out at eight mirrors because its determinant reaches zero at nine, and five exceptions. The same search finds the hyperbolic kaleidoscopes, and finds that they stop at five mirrors.

When a press skips its own light: one board lit, one proved impossible. 4 by 4: solved with 10 presses; 5 by 5: unsolvable, witness set of 5 lights each seeing an even number of the set. Computation

When a press skips its own light

In Lights Out every board can be turned from all off to all on, because each press toggles its own light. Change the rule so a press toggles only the neighbours and the guarantee is gone: 310 of the 3,600 boards up to sixty by sixty cannot be lit. A single comparison of powers of two predicts every one of them, and on general graphs the answer depends on whether the number of vertices is odd or even, for a reason that is one line of algebra.

The two-handled surface's octagon, eight to a corner. Regular hyperbolic octagon with angles π/4 and 64 neighbouring copies in the Poincaré disc. Topology

An octagon only the hyperbolic plane can hold

Glue an octagon's sides in the pattern that makes a surface with two handles and its eight corners become one point, so their angles must add to a single turn. A flat octagon's add to three. The octagon that fits has corners of 45° and lives in the hyperbolic plane, where its area is forced to be exactly 4π and its four gluing moves are motions of the plane whose commutators multiply to the identity.

Fictitious play that never settles. Shapley's 3 × 3 game, 34 runs in 2,000,000 rounds; final mixtures row 0.371, 0.456, 0.173, column 0.253, 0.544, 0.203. Applied

A best reply to the past

Let two players each answer the whole history of the other's play as though it were a fixed plan. In a zero-sum game the history closes in on the equilibrium, as Julia Robinson proved in 1951. In Lloyd Shapley's three-by-three game it never does: play runs round six profiles for ever, each run the one before plus the one three before plus two, so the runs grow by the real root of x³ = x² + 1 and every run is a fixed share of all the history.

The tree of solutions of x² + y² + z² = 3xyz. Markov triples from (1, 2, 5) to depth 4; 231 Markov numbers below 10^15; first 1, 2, 5, 13, 29, 34, 89, 169, 194, 233, 433, 610, 985, 1325. Geometry

The numbers that come after the golden ratio

The golden ratio is the number worst approximated by fractions: no constant larger than √5 works in Hurwitz's theorem. Set it aside and √2 is worst, with √8. Set that aside and the next is a number whose continued fraction repeats 2, 2, 1, 1, with √221/5. Markov proved in 1879 that the list continues for ever below 3, one entry for each solution of x² + y² + z² = 3xyz.

A walk on Gaussian primes stopped by a moat. Step ≤ 2 from 1 + i: 720 primes reached, farthest 17 + 42i at 45.310. Number

A walk on Gaussian primes stopped by a moat

Start at 1 + i and step from Gaussian prime to Gaussian prime, never more than 2 at a time: the walk reaches 720 primes and stops at 17 + 42i, surrounded by a moat. Allow steps of √10 and it reaches a quarter of a million primes before stopping at distance 1,024; steps of 4 carry it to 4,313. Whether some fixed step lets a walker reach infinity is a question Basil Gordon asked in 1962, and it is open.

A fixed p-value favours the null more as the sample grows. Even odds at p=0.05 near n=76, at p=0.01 near n=1202, at p=0.001 near n=83176; p=0.05 at n=10⁶ gives 116.9 to one for the fair coin; PEAR data z=3.614, p=0.00030, BF₀₁=11.90. Probability

A result at five per cent that favours the null

Toss a coin a million times and find a count of heads exactly at the edge of significance, p = 0.05. A test at the usual level rejects fairness. A Bayes factor comparing a fair coin with a coin of unknown bias says the opposite, by 117 to one. Both computations are right, and the sample size that separates them is about seventy-six tosses.

Four ways to say which pixels touch. curve 8, rest 4: curve 1 pieces, rest 2 pieces; curve 8, rest 8: curve 1 pieces, rest 1 pieces; curve 4, rest 4: curve 12 pieces, rest 2 pieces; curve 4, rest 8: curve 12 pieces, rest 1 pieces. Topology

A curve of pixels needs two kinds of touching

On a screen a closed curve is a ring of pixels, and whether it has an inside depends on which pixels count as touching. Let diagonal pixels touch, and the inside leaks out through the corners; forbid it, and a curve can fence in a region while being twelve separate dots. Of the 1,048,575 sets of pixels in a small window, every simple closed curve obeys the Jordan curve theorem exactly when the curve and the rest of the plane are given different rules.

All themes