A game from Inge Lehmann’s paper “P′” (1936)

Find Your Inner Core with Lehmann

The Earth is a ball. When the ground shakes somewhere, the shaking runs right through the ball and comes out on the far side. You are the earthquake: a wave sent through the ball. Find out how it bends, through your listening posts, which record what comes out on the other side.

Choose the direction of your wave: does it go straight through the core, just touch it, or pass it by?

In 1936 everyone thought the Earth had two layers: a mantle, and one core inside it. But after real earthquakes most listening posts shook hard, and one band of them only quivered. If an Earth with one core can do the same, it is right. If it can’t, something is missing, and there might be a core within the core.

Does the Earth have one core, or a core within the core?
lights: a strong shakequivers: a faint onethe band that only quivers
Dive45°
0° straight down90° along the ground

The verdict
Imagine one corethe band stays silent
Imagine a core within the corethe band quivers, as recorded
What you did

What Lehmann found

What happened next

Read further: where the game takes liberties
1936
Find Your Inner Core with Lehmann

How to play

Does the Earth have one core, or a core within the core?

The picture is the Earth cut open. Ochre is the mantle, rock where a wave runs fast. Rust is the core, where a wave runs slower, so going in it bends toward the middle and coming out it bends back. A wave runs straight and bends only where the colour changes. The burst at the top is the earthquake. The dots round the right-hand edge are eighteen listening posts.

  1. The facts. Three real earthquakes, each shown when you press the button. You see the waves run through the Earth along the paths everyone agreed on, and what the posts recorded each time, kept as three rows of dots. Most light up. One band only quivers. That band is then marked. There is nothing for you to do here but watch.
  2. Imagine one core. Now you are the earthquake, in an imagined Earth with one core. Aim by dragging on the picture or sliding the dial: 0° dives straight down through the middle, 90° runs along the ground. The − and + buttons move a tenth of a degree. Fire, and the post nearest to where you come up lights. Try to make the band quiver. Four goes, then one last try.
  3. Every direction at once. The game then fires every direction you could have chosen. This is not a go of yours; just watch.
  4. Imagine a core within the core. Three goes in an imagined Earth with a second, smaller core inside the core. Only a dive steeper than 16° gets that deep. Then every direction at once again, and the verdict.
Find Your Inner Core with Lehmann

About

Built from Inge Lehmann, “P′”, Publications du Bureau Central Séismologique International, Série A, Travaux Scientifiques 14:87–115 (1936), read in full from a scan of the original. For what came after, and for a guided reading, see Martina Kölbl-Ebert, “Inge Lehmann’s paper: ‘P′’ (1936)”, Episodes 24(4):262–267, December 2001, at episodes.org. Game built from the paper; the paper is the authority where the two disagree.

What the paper tells

Earthquake waves arrive late on the far side of the Earth, and in a band of the far side the ordinary ones hardly arrive at all. By 1936 this was explained by a core of slower material: waves entering it are bent toward the middle. Lehmann opens by drawing the paths exactly, on an Earth simplified on purpose to two shells of constant speed, 10 km/s in the mantle and 8 km/s in a core five ninths of the Earth’s radius: straight rays, bent only at the boundary. The ray that just grazes the core leaves at 33.75° and comes up 112.5° round; the core rays come back no nearer than 153.9°. Nothing lands in between.

But stations inside the real shadow did record a wave, which she calls P′3. It had been thought small. She shows that it is not: the sideways movement is small, but the up-and-down movement can be considerable, and it grows with distance. Her Figure 3 prints the records of the Buller earthquake in New Zealand, 16 June 1929, from Irkutsk (110.8°), Sverdlovsk (135.0°) and Baku (137.4°). The accepted story was diffraction: waves creeping round the edge of the core. She writes that this can hardly be considered probable any more, and suggests a third shell: an inner core, radius 8/10·r0·sin 16°, about 1400 km, in which waves run faster, 8.6 km/s. Rays leaving at 16° just touch it; between 16° and 14.86° they are totally reflected, a ray she calls of quite small intensity; steeper rays pass through and come up from 118.8° round out to 180°, steeply, and stronger with distance. Those are the characteristics observed. She offers it as a hypothesis that cannot be proved from the data at hand. The second half of the paper reads four stations near 150° for the same earthquake and ends by asking other investigators to look for suitable material.

The paper never says that the inner core is solid; it gives it one property, a higher speed. Gutenberg and Richter computed its size two years later; Jeffreys showed in 1939 that diffraction could not account for the records; Birch proposed in 1940 that it is solid, and Dziewonski and Gilbert showed it in 1971 from the free oscillations of the Earth.

What the game does

It starts from the facts and then imagines two Earths. First, with nothing to do but watch, three real earthquakes, shown slowly one after another and kept as three rows of dots: each time most listening posts light up and one band of them only quivers. Meanwhile the picture of the Earth’s inside alternates between the two possibilities, one core and a core within the core. The band is then marked. Next you are asked to imagine an Earth with one core, and to be the earthquake in it: you choose a direction, fire, and see where you come up. After the first go the game asks you to make the band quiver; after four it asks what is wrong and gives one last try; then it fires every direction you could have chosen, the band stays silent, and the game owns up that this was a trap: with one core it cannot be done. A card states Lehmann’s idea. You then imagine an Earth with a core within the core and have three goes; a dive between about 12° and 16° comes up in the band and a post there quivers. Every direction at once then makes the band quiver, as the real earthquakes did. The verdict depends on whether you made it quiver yourself, says what Lehmann said, and is followed by Read further, which lists where the game simplifies.

The mechanism, and what it claims

Kind: ported. The game runs Lehmann’s own model: straight rays, one constant speed per shell, the law of refraction at each boundary, total reflection where the sine would exceed one, and the epicentral distance summed from the chord geometry. Every number she gives comes out of it: the grazing ray at 33.75° coming up at 112.5°, the same ray refracted to 186.2°, the core rays running back to 153.9°, the ray at 16° touching the inner core and coming up at 160.7°, total reflection from 14.86° coming up at 118.8°, and her Table II of rays through the inner core, to the tenth of a degree.

The game demonstrates what the two model Earths predict: with one core, a band of the far side that no wave reaches; with a core within the core, a few steep waves turned into that band.

The liberties, all ours, and all listed for the player under Read further. (1) The three earthquakes are real and are the ones the paper discusses (the Pacific earthquake of 26 June 1924, Buller on 16 June 1929, Hawke’s Bay on 2 February 1931), but what the posts do on screen is a simple picture of what the paper reports: that waves are observed inside the real shadow with smaller amplitudes, with three records of the Buller earthquake printed as her Figure 3. The game turns that into a band of four posts that quiver and puts the band where the model’s shadow is. (2) You choose one direction at a time; a real earthquake sends waves every way at once, which the game shows only as “every direction at once”. (3) In play one post always answers: a wave that comes up outside the band lights the nearest post outside it, and a wave that comes up inside the band makes the nearest post there quiver. With the three-degree ear that decides what “every direction at once” reaches, many goes would light nothing, and three of the four band posts would need a dive right to a tenth of a degree. (4) The game’s reason for the faintness, that few waves land there, stands for her intensity calculation. (5) Diffraction, the rival explanation the paper argues against at length from the strong up-and-down movement of the records, is left out of play, which makes the choice look easier than it was.

The game assumes the model’s simplifications, which Lehmann chose on purpose: straight rays where the real Earth’s curve, one speed per shell where speed really changes with depth, and no loss of energy at any boundary. The posts every ten degrees, the three-degree ear within which a post lights, and the numbers of goes are ours.

The paper establishes that, in the two-shell model, no ray lands between 112.5° and 153.9°; that an inner core of about 1400 km with a higher speed sends rays into that band from 118.8° on, arriving steeply; and that this fits what the stations in the shadow recorded better than diffraction does. It does not establish that the inner core exists: Lehmann says so herself.

Where the discovery comes from. The shadow and its filling follow from the model’s geometry, not from our rules: the calibration check shows the same four posts unreached with the ear at one, two, three or four degrees, and the same posts reached with her inner core; with the posts shifted by five degrees the story is the same except that a post at 115°, in the sliver between 112.5° and 118.8° that her own model leaves dark, stays unreached.

The numbers

Every number sits in the config block at the top of the mechanism, tagged with its page: the radii and speeds of the three shells, the paper’s own ray results and Table II used by the check, the three stations of Figure 3, the later dates, and everything ours (posts, ear, waves, the words).

What is left out

Lehmann’s travel-time curves and intensity calculations; the S-waves; the curved rays of the real Earth beyond a line here; her layer of gradually decreasing speed that explains the weak waves just beyond 100°; the second half’s reading of the records near 150°; most of the biography.

Credits

Built with Claude under the paper-to-game skill. Type: Fraunces (Undercase Type) and DM Mono (Colophon Foundry), both SIL Open Font License, embedded. Source, design memo and calibration check: github.com/steffenhuck/paper-to-game.