We are currently drilling through a big pile of mud and sand on the seafloor (the biggest such pile of mud and sand in the world), and I’m spending most of my day sitting next to the “silver bullet” in this photo:
If you can’t see the sign in the photo, this is the superconducting rock magnetometer (SRM) on the JOIDES Resolution. We use it to measure the record of Earth’s ancient magnetic field in rocks and sediments. Right now, we’re running sections of marine sediment cores through the machine. The SRM tells us what direction your compass would have pointed if you were standing here hundreds of thousands – or even millions – of years ago.
Muck in the oceans builds up, layer upon layer, so that older mud eventually gets covered with younger stuff. If you look closely at the muck, you’d see it was composed of lots of tiny particles. These are pieces of clay, silt, and sand formed from the detritus of eroded mountain ranges, the decaying bodies and shells of tiny fossil creatures, dust from the air, tiny crystals that form in the oceans, and even microscopic meteorites. Some of those particles are magnetic. For the most part, those contain the magnetic iron oxide magnetite , which can be part of the dregs of continental erosion, or it can be made by bacteria in the ocean, or by a number of other things. As the tiny magnetic particles fall through the water, they turn so that they are magnetized in line with Earth’s magnetic field – just like little compasses . After they fall into the sediment accumulating on the seafloor, the magnetic particles get buried, “locked” in position by the other particles surrounding them. If Earth’s magnetic field switches polarity, the “tiny compasses” in new sediment being deposited will align with Earth’s new magnetic field, but the ones already locked in the sediment will stay as they were.
At least, that’s how the typical story goes about how sediment records the direction of Earth’s magnetic field. In reality, it’s not so simple. For one thing, all kinds of creatures live in the sediment – like whoever lived in this burrow:
This sediment core is actually full of fossil burrows. But sediments full of burrows can record Earth’s magnetic field just fine. We think it might be because the creatures burrowing in the sediment stir up the muck just enough that it settles back in line with Earth’s magnetic field again. It’s just that the sediment “locks in” the record of the magnetic field after the burrows themselves get buried. That seems reasonable until you realize that this burrow and others like it did not record a magnetic field in the same direction as the sediment around it . This burrow is filled with pyrite, which, though iron-bearing, is not itself magnetic in the same way as magnetite . Some geologists think that something happened to make new magnetic materials form or old ones dissolve around burrows like this one.
To make things even more complex, the area we are looking at on the Bengal Fan was not formed by sediments settling out in quiet water. Instead, much of the sand and mud deposited here was dumped very quickly from places close to land . Do the magnetic particles in these tremendous currents full of churning sand and mud even have time to be pulled by Earth’s magnetic field, or are the forces in the currents too great? It looks like, at least in the muddy parts of deposits like the ones we’re studying, the sediment does keep a mostly faithful record of Earth’s magnetic field.
In the end, the story we tell about how sediments become magnetized is probably fundamentally OK, but there are parts of it we still don’t fully understand. Those parts of the story we’re still curious about are what keep us doing science!
 Magnetite is Fe3O4. To a certain extent, hematite (Fe2O3) and goethite (FeOOH) can also be incorporated into marine sediments, along with other magnetic minerals that can grow there.  Unlike in igneous rocks, where the magnetic minerals “lock in” a record of Earth’s magnetic mineral as they cool. The minerals in igneous rocks DO NOT move.  See Abrajevitch, A., Van Der Voo, R., and Rea, D., 2009. Variations in abundances of goethite and hematite in Bengal Fan sediments: Climatic vs. diagenetic signals, Marine Geology 267:191-206.  Pyrite is paramagnetic, meaning that it can be magnetized only in the presence of a magnetic field, not after the field is gone; magnetite is ferromagnetic, meaning that it can be permanently magnetized. This is called a turbidity current, and the sand and mud deposits it leaves behind are called turbidites.