The narwhal has been called the unicorn of the sea for so long that we mostly stopped asking the obvious question, which is: how does a whale grow a spear? Not a tusk that curves, not an antler that branches. A single dead-straight tooth that can stretch to three metres. It's the only straight tusk in nature. Every other elongated tooth or horn on Earth, elephant ivory, warthog tusks, boar spears, curls. This one doesn't. And until now nobody really knew why.

Researchers have now worked out the mechanism, and as usual in biology the answer is stranger than the myth. The tusk is a modified left canine tooth, usually in males, that grows in a slow spiral. The twist, if you'll pardon it, is that the spiral itself is what keeps it straight: the tooth rotates as it grows, and that rotation cancels out the natural asymmetric drift that curves every other tusk on the planet.

Sit with that for a second. Every other long tooth curves because the tissue on one side of the growing tooth lays down slightly faster than the other. It's a manufacturing flaw so universal we treat it as a design feature. The narwhal solved the flaw by spinning the flaw around the axis. If one side grows faster but the whole tooth is slowly turning, the faster side spends equal time pointing in every direction, and the drift averages out to zero. The curve eats itself.

That mesmerizing helical groove you see on every narwhal tusk in every museum cabinet? We've been displaying the mechanism for generations without recognizing it. The spiral isn't decoration and it isn't a strengthening ridge. It's the ghost of the rotation, a record of the correction written in dentin.

Which raises the follow-up question I couldn't stop chewing on at about two in the morning: why? Straightness has to confer something or the mechanism doesn't evolve. A curved tusk would be useless for whatever the narwhal is doing with this thing, and the honest answer about what narwhals do with their tusks is that we still don't fully know. The tusk is connected to sensitive nerve tissue, which has fed a long-running argument that it works as a kind of sensory organ, streaming information about the Arctic water around it. Males have been seen rubbing tusks together, which might be dominance behavior or might be something more like reading each other.

So the leading picture is an organ that may have to be straight to function as a probe and a lance, and the animal evolved a growth trick to force straightness on a body plan that doesn't naturally produce it. A whale that wanted a straight line and refused to accept a curved one.

I've written before about how the questions are better than the answers, and this one mostly proves the point in reverse, because the answer is so good it opens three more doors. Does the rotation happen in the socket or in the tissue? Is the growth rate tuned, could a narwhal produce a deliberately curved tusk if the rotation slowed, and are the rare crooked tusks sitting in museum collections evidence of exactly that failure? Somebody is going to spend a career on the crooked ones. I hope they do.

Humans have been fascinated by this object for a very long time, and the fascination was hiding a physics problem all along. Quietly spinning its way out of a whale's jaw. Three metres of self-correcting tooth, the only straight line in the animal kingdom, drawn by refusing to stop turning.