Inge Lehmann sat in a small office in Copenhagen in 1936, sorting through stacks of paper seismograms from a 1929 earthquake near New Zealand, and noticed something that should not have been there. Waves from the quake were arriving at stations on the far side of the planet in a zone where, according to the physics of the day, no waves should reach at all. She published her explanation in a three-page paper titled simply P’, arguing that Earth’s core was not a single ball of molten metal but held a distinct inner sphere nested inside a liquid outer shell. Nearly a century later, every textbook diagram of the planet’s interior still rests on the shadow she read.
No human has ever seen the core. No drill has come within a thousand miles of it. The deepest hole ever bored, the Kola Superdeep Borehole in Russia, reached about twelve kilometres before the rock became too hot and plastic to cut. The distance to the centre of the Earth is 6,371 kilometres. Everything known about what lies between those two numbers has been inferred from vibrations.
How a wave becomes a map
When an earthquake ruptures a fault, it releases energy in two main kinds of body waves. P-waves are compressional, pushing and pulling rock in the direction they travel, and they move through solids and liquids alike. S-waves are shear waves, wobbling rock sideways, and they cannot travel through liquid at all. That single distinction is what turned seismology into a tool for imaging the inside of a planet.
Waves slow down in denser or hotter material. They bend when they cross a boundary between layers of different composition, following the same rules of refraction that steer light through a lens. They reflect off sharp interfaces. And they vanish, at least from certain stations, when a barrier absorbs or deflects them away.
By the early twentieth century, seismologists had noticed a strange gap. Between roughly 105 and 140 degrees of arc from any large earthquake, P-waves went missing and S-waves disappeared entirely. That ring around the planet became known as the shadow zone. In 1914 the German seismologist Beno Gutenberg calculated that the shadow implied a liquid core beginning at a depth of about 2,900 kilometres, because only a molten layer could stop S-waves and bend P-waves so sharply.
The 1929 New Zealand quake
Lehmann worked as the head of the seismology department at the Royal Danish Geodetic Institute, in an era when computation meant paper, pencil, and cardboard boxes she used to organise her data by hand. She had been tracking arrivals from a major earthquake that struck New Zealand’s South Island in June 1929. Some P-waves were arriving inside the shadow zone. Faintly, later than expected, but unmistakably there.
Gutenberg’s liquid-core model could not account for them. If the core were entirely molten, those waves should have been swept aside. Lehmann proposed a simpler fix: put a second boundary inside the core. A second inner sphere, roughly 1,220 kilometres in radius, would refract some of the incoming P-waves so sharply that they would emerge into the shadow zone as faint late arrivals. She traced the ray paths by hand and matched them to the seismograms. The paper was published in 1936 in the Bureau Central Séismologique International. Its title, P’, referred to the anomalous wave phase itself.
It took more than a decade for the community to accept the idea, and longer still to work out what the inner sphere was made of. Lehmann’s data could show that a second boundary existed, but not directly that the material inside it was solid; her seismograms carried only P-waves, which travel through liquids as readily as solids. Francis Birch argued in 1940, and Keith Bullen in 1946, that the inner core had to be solid, and in 1971 Adam Dziewonski and Freeman Gilbert confirmed it, using the way the whole planet rings after a great earthquake, its normal modes, to show that the inner core resists shear the way a solid does. By then Lehmann was in her eighties. She lived to 104, long enough to see her three-page paper become the foundation of modern deep-Earth science.
What the shadows say the core is made of
The layers Lehmann and her contemporaries mapped are now known in fine detail. The outer core, from about 2,890 to 5,150 kilometres deep, is liquid iron alloyed with nickel and a lighter element, probably sulphur, silicon, or oxygen. Convection currents in that molten metal generate Earth’s magnetic field. The inner core, from 5,150 kilometres to the centre, is solid iron under pressure exceeding three million atmospheres, at a temperature close to that of the Sun’s surface. It is solid not because it is cool but because the pressure squeezes the atoms into a crystalline lattice they cannot escape.
The inner core is smaller than the Moon but denser, and it is growing. Heat leaks outward from the centre, and liquid iron at the boundary crystallises onto the solid ball at a rate of roughly a millimetre per year. That slow freezing is what drives convection in the outer core, and therefore what powers the magnetic field that deflects the solar wind and keeps the atmosphere from being stripped away.
Reading the boundaries
Every layer of the interior announces itself as a discontinuity, a place where seismic wave speed changes abruptly. The Mohorovičić discontinuity, discovered by Andrija Mohorovičić in 1909 after studying a Croatian earthquake, marks the base of the crust. The 410 and 660 kilometre discontinuities in the mantle correspond to mineral phase transitions where olivine rearranges its atoms under pressure. The core-mantle boundary at 2,890 kilometres, called the Gutenberg discontinuity, is the sharpest interface in the planet. P-wave speed drops from about 13.7 to 8.1 kilometres per second across it, and S-waves stop entirely.
Modern seismology has pushed far beyond Lehmann’s ray-tracing. Techniques called seismic tomography treat the Earth like a giant CT scanner, combining arrivals from thousands of earthquakes recorded at thousands of stations to build three-dimensional images of wave speed anomalies. Those images have revealed continent-sized blobs at the base of the mantle called large low-shear-velocity provinces, one under Africa and one under the Pacific, whose origin is still argued. They have found ultralow velocity zones just above the core-mantle boundary, patches perhaps twenty kilometres thick where waves slow dramatically, possibly pockets of partial melt or dense material sinking through the mantle.
The signals nobody has explained yet
Some of what the waves are still saying remains unread. A class of signals known as PKP precursors, which arrive a few seconds ahead of the expected core-piercing P-wave, has puzzled seismologists for decades. A 2024 analysis published in a study of scattering in the lower mantle suggested the precursors are scattered by kilometre-scale heterogeneities near the core-mantle boundary, but the specific nature of those scatterers, whether chunks of subducted ocean floor, unmixed primordial material, or something else, is still debated. Newsweek quoted researchers plainly saying they do not know what they are.
The inner core itself has turned out to be less static than Lehmann’s picture implied. In February 2025, a team at the University of Southern California reported evidence that the inner core’s surface is deforming, changing shape over decades in ways detectable by comparing waveforms from repeating earthquakes. Other studies have argued the inner core rotates slightly faster or slower than the mantle above it, and may have recently reversed direction. The debates are sharp because the data are indirect, filtered through every layer of rock the waves crossed on the way out.
Earthquakes as instruments
Large quakes are also, in a real sense, planetary experiments no laboratory could reproduce. The 2011 magnitude 9.0 Tohoku earthquake off Japan released so much energy that its seismic waves struck the core and bounced back, giving researchers a fresh set of reflections to interpret. Every great earthquake is a chance to re-image the planet with slightly better resolution. Recent experimental work has reconstructed how seismic waves speed up through the D” layer, just above the core, by squeezing minerals in diamond anvil cells to the pressures found there and watching how sound moves through them.
That is the strange arrangement seismology has settled into. The interior is unreachable, but every rupture along every fault sends signals through it, and each signal carries a coded record of everything it passed through. The instruments read the code. The models translate it into layers, boundaries, and textures. What Lehmann did with a pencil and a stack of station records, tomography now does with server farms, but the underlying trick is the same one she used in 1936: watch where the waves should be, notice where they are instead, and let the difference draw the map.
Lehmann is buried in Copenhagen. The inner core she found, freezing outward at a millimetre a year, is now roughly ninety millimetres larger than it was when her paper appeared. Every earthquake since has confirmed it is still there.
