The Straightedge Said Fourteen and It Was Wrong

Nomography (Nomogram Design and Construction)

The straightedge said fourteen. Three times six is not fourteen, and yet there it was, my own pencil line crossing the middle scale exactly where I’d written a small, confident 14. An evening of careful draughting had produced a device that lied to me on the first honest question I asked it.

A nomogram is a calculator you draw once and then never touch again except with a ruler. You lay out two or three scales — think of them as number lines, though they rarely space their numbers evenly — and you arrange them so that a straight line laid across two known values crosses the third scale at the answer. That crossing line has a name, isopleth, and the whole appeal is that it turns arithmetic into a physical gesture. No keys, no batteries, no order of operations. You put a ruler down and read.

I came at this sideways, the way I come at most things. Back in March I opened a drawer and found my grandfather’s Pickett slide rule, worn smooth at the trig scales, and spent a small-hours evening relearning that a slide rule adds two logarithms by adding two lengths. A nomogram is that same idea let off its rails. Instead of sliding one log scale against another, you print both as fixed lines and let the ruler do the sliding. The E6-B in my flight bag — the circular whiz-wheel every student pilot spins for groundspeed and fuel burn — is exactly this, a log nomogram bent into a disc. I’ve been reading nomograms at three thousand feet for years without once building one.

The arithmetic behind the tick marks

The simplest useful design solves anything of the form f₁(u) + f₂(v) = f₃(w) with three parallel vertical lines. Here is the part nobody tells you until you’ve already got it wrong: the spacing is not free. If the two outer scales sit distances a and b apart from the answer line, the answer scale has to land at the fraction a / (a + b) between them, and its unit length gets multiplied by ab / (a + b). That factor is called the modulus — millimetres per unit — and if you pick the three moduli independently because they “look about right,” your straightedge will cross the middle scale a stubborn few millimetres off every single time.

My fourteen was cruder than a modulus error. I’d drawn the centre scale running the wrong way — numbers increasing downward instead of up — so the isopleth was reading my multiplication backwards. It is the nomographer’s version of botching the decimal point on a slide rule: the mechanism is perfect and the human has installed it upside down. Flip the scale, re-ink, and 3 across to 6 crosses cleanly at 18. Because the scales are logarithmic, that addition-of-lengths is a multiplication — the same trick, frozen to the page.

There is a deeper thing under all this that I only half-believed until I saw it work. Three points lie on one straight line if and only if a certain 3×3 determinant equals zero:

| x₁  y₁  1 |
| x₂  y₂  1 |  =  0
| x₃  y₃  1 |

So designing a nomogram is really algebra with a geometric payoff: you shove your equation around until it is that vanishing determinant, and then each row hands you the (x, y) recipe for plotting one scale. Expressing a relationship as a form a machine can execute — that is just programming with a fountain pen, and it scratched precisely the same itch.

What it can and can’t promise

A hand-drawn nomogram is honest to about three significant figures, which is to say roughly a percent. The line has width; that width is your error bar, and pretending otherwise only manufactures false confidence. This is not a flaw the field apologizes for. It’s why nomograms outlived the slide rule in exactly the places where a fast, powerless, uncrashable answer matters more than a ninth digit — the Rumack-Matthew chart an emergency doctor still lays a ruler across to decide whether an overdose needs treating. Maurice d’Ocagne, who coined the word in 1899, was barely ahead of powered flight; the tool and the aeroplane grew up together.

Tonight I have one working multiplication chart and a smudge of ink on the side of my hand. The next one I want is a density-altitude nomogram — pressure altitude and temperature in, the number that decides whether a short strip is a good idea out — because I already trust the answer the E6-B gives me, and there is something bracing about the prospect of drawing that trust from scratch and checking whether my own geometry deserves it.