An "Anomaly" in a Cambridge Lab
In 2013, Professor Malcolm Burrows of the University of Cambridge's Department of Zoology was doing something that looked simple—filming insects jumping with a high-speed camera.
His subject was *Issus coleoptratus*, a planthopper common in European gardens. Less than a centimeter long, it is a superb jumper, leaping up to 100 times its body length.
But one thing kept puzzling Burrows: when this insect jumps, the timing difference between the launch of its two hind legs is less than 30 microseconds.
For perspective: a microsecond is one-millionth of a second. A human blink takes roughly 300 milliseconds—300,000 microseconds. The planthopper's leg synchronization is about 10,000 times more precise than a human blink.
This didn't add up.
Nerve conduction in insects runs at only a few meters per second. From the "jump decision" to the signal reaching the leg muscles takes at least a few milliseconds. Even if both legs received commands simultaneously, tiny differences in signal path length should produce millisecond-level timing differences.
Yet the planthopper achieved 30 microseconds—two orders of magnitude faster than nerve signal conduction itself.
Burrows and his colleague Gregory Sutton decided to look at the insect's legs—not with the naked eye, but with a scanning electron microscope.
They saw something that shouldn't exist.
Gears
At the base of each hind leg of the planthopper nymph (the juvenile stage before it becomes an adult), at the joint where the leg meets the body, there is a toothed structure.
Each strip is about 400 micrometers long—roughly the diameter of a human hair—with 10 to 12 regularly arranged teeth.
When the two hind legs are cocked, these toothed strips mesh with each other.
This is a pair of gears.
Not mimicry, not something that merely "looks like gears"—a genuine, functional mechanical gear train: two rigid bodies transmitting rotary motion through tooth engagement, forcibly synchronizing the phase of both legs.
This was the first time gears had been found in a living organism. To date, it remains the only one.
A Detail That Made Engineers Sit Up
When Burrows and Sutton magnified the electron micrographs, they noticed a chilling detail.
The root of each tooth is not a right angle but a rounded arc transition.
Anyone with mechanical engineering training will immediately recognize this: it's called a fillet, a standard practice in gear design to reduce stress concentration at the tooth root and prevent breakage under load.
Human engineers invented this technique during the 18th-century Industrial Revolution. The planthopper had it hundreds of millions of years ago.
This isn't merely "convergent evolution" in the loose sense—it is evolution independently arriving at a hallmark of human engineering, down to the details (fillets to reduce stress).
Why can evolution "invent" gears? Because a gear's essence is "two periodic structures synchronizing phase through contact." Given sufficiently hard, sufficiently regular structures (chitin at the leg joints), natural selection can gradually approach gear morphology through small shape changes.
Even so, this is nature's only known example. Every other jumping insect—fleas, grasshoppers, jumping spiders—uses other synchronization mechanisms. The planthopper's gear is a lone branch on the tree of life.
A 21-Microsecond Miracle
The gears' effect is stunning.
High-speed footage shows that when a nymph launches, the timing difference between the two hind legs is only 21 microseconds (a more precise later measurement)—two orders of magnitude faster than nerve conduction.
The gears accomplish what the nervous system cannot.
The mechanism: before jumping, the nymph cock its hind legs against its body, and the gears engage. The muscles then contract, storing energy like drawing a bowstring. Upon release, both legs must push off simultaneously—if one leg fires a few microseconds early, the insect tumbles in mid-air and lands short.
The gears guarantee "simultaneous." As long as they're meshed, the two legs' rotation angles are physically locked—off by no more than one tooth pitch.
This is true "hard synchronization"—not signal coordination, but physical structure.
But Adults Throw Away the Gears
Here the story could end as a tidy "nature outsmarts engineers" tale. But Burrows found something that complicates it considerably.
When the planthopper becomes an adult, the gears disappear.
After the final molt into a winged adult, the toothed strips at the leg joints are gone. Adults synchronize their legs another way—friction pads between the trochanters. When the legs are cocked, the pads contact and synchronization relies on friction.
This is strange. The gears are precise to 21 microseconds; the friction pads are far less precise. Why would adults "downgrade"?
Burrows and Sutton offered an explanation even more fascinating than the gears themselves.
The Gear's Fatal Weakness
Every mechanical engineer knows a gear's Achilles' heel: if one tooth breaks, the whole gear is ruined.
Teeth mesh rigidly. A chipped tooth causes the engagement to jam, and synchronization precision collapses—often worse than having no gear at all, because a gear is designed for complete engagement; a damaged one becomes not a synchronizer but a fault source.
The planthopper's teeth are only tens of micrometers wide, made of chitin, and under the shock of high-speed jumps they can chip.
Nymphs don't care. They molt. Each molt discards the old exoskeleton—including the old gears—and grows a new one, new gears included. The gears are consumables: broken, then replaced.
Adults never molt. The final exoskeleton must last a lifetime.
If adults kept the gears, a single chipped tooth—perhaps on the 100th jump, perhaps the 1,000th—would permanently disable them. No repair mechanism. No spare parts. The gear would turn from a precise synchronizer into a permanent fault source.
So adults use friction pads. Less precise, but with a huge advantage: they don't fail catastrophically from local damage. A worn pad has slightly less friction but still works. It degrades gracefully rather than all-at-once.
Repairability Determines Design Choices
What struck me most about this story is not the gears but the decision to abandon them.
Natural selection chose gears for the nymph—because they're repairable. Natural selection abandoned gears in the adult—because they're not.
One species, two life stages, two design philosophies.
Behind this lies a principle deeper than gears: a system's design is determined not by what is most precise, but by whether it can be repaired when broken.
If the system is repairable (the molting nymph), use the "precise but fragile" solution—gears. If it isn't (the non-molting adult), use the "less precise but more robust" one—friction pads.
Engineering has a formal name for this: maintainability constraints design choice.
Cross-Domain Analogies: Software, AI, and Planthoppers
This principle is everywhere in human engineering.
Software: microservices vs. monoliths. Microservices are "gears"—precise, efficient, low-latency, but one failing service can trigger cascading collapse. Monoliths are "friction pads"—less elegant, but local failures don't propagate. Which to choose depends on your repair capacity: with strong monitoring, circuit breakers, and auto-recovery (= molting), use microservices; without them, use a monolith.
AI systems: hardcoded rules vs. probabilistic inference. Hardcoded rules are "gears"—precise, interpretable, but one wrong rule crashes the system. Probabilistic inference is "friction pads"—less precise, but tolerant of bad data. The choice depends on data quality control: clean, iteratively updatable data (= molting) favors hard rules; dirty, drifting data favors probability.
Even relationships: contracts vs. trust. Contracts are "gears"—precise, enforceable, but one breach collapses everything. Trust is "friction pads"—fuzzy, unquantifiable, but tolerant of small frictions. The choice depends on repairability: long-term interactions with apologies and compensation (= molting) favor trust; one-shot, unrepairable games favor contracts.
A planthopper's nymph and adult demonstrate a universal design principle with two pairs of legs.
Evolution's "Engineering Wisdom"
The planthopper's gears also reveal something else: evolution is more pragmatic than human engineers.
Engineers tend to fall in love with an "elegant" design and apply it everywhere. Gears are great—so use gears everywhere. The result: systems overly dependent on rigid synchronization that collapse entirely when a part fails.
Evolution doesn't care about elegance. It only cares about surviving to reproduce.
So evolution can produce a seemingly contradictory design: gears for the nymph, friction pads for the adult—one species, two synchronization mechanisms, switched by life stage. For evolution there is no "optimal solution"—only "the optimal solution under current constraints."
When the constraint changed (from moltable to non-moltable), the design changed.
That may be the planthopper's final lesson: don't ask "which design is better"; ask "which design is repairable under my constraints".
Because all systems break. The difference lies in what happens after: do you replace a tooth, or throw away the whole gear?
---
*Postscript*: When Malcolm Burrows published this work in 2013, he was already in his seventies—a professor of insect neurobiology at Cambridge who had spent his career studying how insects jump. The paper, published in *Science*, is titled "Interacting Gears Synchronize Propulsive Leg Movements in a Jumping Insect." A seventy-something scientist, looking at the legs of a sub-centimeter insect under an electron microscope, found nature's only gears.
Sometimes the best discoveries come not from chasing hot topics, but from looking at one thing for a lifetime until you finally see what no one else saw.