If you try to imagine a speedy predator, you might envision the muscular body of a cheetah at full sprint, or the aerodynamic pose of a peregrine falcon dive-bombing a pigeon. Both are marvels of evolution, optimised over countless generations to push against the boundaries of speed and acceleration. They move their relatively large bodies at mind-boggling speeds.

But there is another class of speedy predator: one that channels all of its power into moving just one or a handful of specific body parts, while the rest of the animal remains virtually motionless. By doing so, these predators have moved beyond the acceleration rates of the swiftest cat or bird. Far, far beyond.

Take the smashing mantis shrimp. The attacks it performs with its club-like mouthparts are so fast, and so brief, that their exact speed went unappreciated by scientists until about 15 years ago – when our high-speed camera technology finally caught up.

When those acceleration rates were formally clocked, they were breathtaking: thousands of times faster than anything previously seen in nature.

And what does the mantis shrimp do with its astonishingly quick weapons? It uses them to attack an animal virtually synonymous with sluggishness: a snail.

"It's super counterintuitive," says Sheila Patek at Duke University in Durham, North Carolina. "It's changing what biologists think speed and acceleration are actually for."

Mammals and birds are out-accelerated by animals that generally lack a reputation for speed

This is the world of super-fast acceleration: a world in which things do not always make immediate sense.

We can probably all accept that humans are not the swiftest members of the animal kingdom. But many of our closest relatives, the mammals, are capable of astonishing speeds and accelerations. So, for that matter, are many birds.

It probably helps that mammals and birds can control their body temperature. Even on a chilly morning, these animals can keep their muscles relatively warm, which means those muscles can respond quickly to provide a sharp burst of speed. So if any four-limbed animal is capable of super-fast acceleration, it seems like it would make sense for it to be one of these furry or feathered species.

But this is not so.

In fact, mammals and birds are out-accelerated by animals that generally lack a reputation for speed: reptiles and amphibians.

The extent to which the acceleration was higher was unexpected

A tiny reptile called the rosette-nosed chameleon holds the current record among the amniotes: that is, animals that give birth on land, meaning reptiles, birds and mammals. It can flick out its tongue so rapidly that it briefly accelerates at 2,590m/s/s - about 170 times faster than the 15m/s/s maximum acceleration of the cheetah or the peregrine falcon.

Christopher Anderson, now at the University of South Dakota, published this extraordinary finding in January 2016.

Anderson knew that many chameleons can accelerate their tongues at rates exceeding 500m/s/s to catch flies and other insects. But he also knew that biologists had tended to study larger chameleons with a body length – not including the tail – of at least 10cm. These species are easier to handle.

However, many chameleons are much smaller than that, including the 5cm-long rosette-nosed chameleon. These chameleons feed in the same way as their larger cousins, by flicking out their tongue.

Muscle simply cannot accelerate at these astonishing rates

In a 2012 study, Anderson had shown that the muscles controlling the tongue are larger, relative to body size, in small chameleons than in big ones. This suggested that small chameleons should be capable of greater tongue accelerations than larger species.

His hunch proved correct, although the 2,590m/s/s figure he calculated still came as a surprise. "The extent to which the acceleration was higher was unexpected," he says.

Although it is the rosette-nosed chameleon's relatively large muscles that help it outcompete other species, the real secret to its accelerating tongue – and, in fact, the secret to super-fast acceleration in general – lies elsewhere.

Muscle simply cannot accelerate at these astonishing rates. No matter how finely-tuned or powerful a muscle is, it can never contract quickly enough.

Apart from human technology like catapults, elastic structures are rarely used in our corner of the animal kingdom

However, a contracting muscle can instead stretch an elastic structure and hold it in a stretched state – or, better still, keep it stretched with some sort of latch. When the latch is released, the elastic structure snaps back into its normal length. This releases energy so explosively that a small and lightweight structure, like the chameleon's tongue, can be accelerated at a tremendous rate.

This means that understanding super-fast acceleration means identifying those elastic structures, and the latches that keep them stretched tight and ready to release. In the case of the chameleon, it appears to be elastic fibres in the tongue itself that help it to unfurl so rapidly.

In theory, mammals could also benefit from these elasticated weapons. But apart from human technology like catapults, elastic structures are rarely used in our corner of the animal kingdom. It is only reptiles and amphibians that have evolved them. For instance, some salamanders can accelerate their tongues at almost 4,500m/s/s.

Why are these animals, but not mammals and birds, optimised for this sort of super-fast attack?

Anderson, working with his colleague Stephen Deban of the University of South Florida in Tampa, has devised an explanation. Paradoxically, it might be precisely because reptiles and amphibians are sometimes slow and sluggish that they have become kings of acceleration.

Chameleons and salamanders might have evolved to be super-fast because of their tendency to be super-slow

Because mammals and birds keep their muscles warm enough to work efficiently no matter the ambient temperature, they rarely experience any shortfall in performance. A predatory mammal can chase down its prey in the cool of dawn or on a warm afternoon.

Reptiles and amphibians do not have this luxury. If the ambient temperature is cold, their muscles are generally cold, stiff and slow to respond.

Elastic structures offer these animals a way to catch a meal even when their muscles are cold. These elastic structures show less of a drop in performance with temperature than muscles do. Anderson and Deban think that might be why so many "ectotherms" – animals that do not control their body temperature – have evolved them.

The idea suggests that chameleons and salamanders might have evolved to be super-fast because of their tendency to be super-slow.

"I think their hypothesis is a good one," says Patek. "Chris and Steve's papers have shown beautifully that this is a great way to circumvent temperature dependence of motion."

The temperature in that tiny volume of water briefly rises to levels normally associated with the surface of the Sun

Patek's interest in super-fast acceleration began over a decade ago while she was a postdoctoral fellow at the University of California in Berkeley. There she began studying the way some species of mantis shrimp use their club-like mouth parts to repeatedly thump marine snails into submission, smashing open the hard shells to feed on the soft tissue within.

However, the research was difficult, because each smashing mantis shrimp attack unfolded far too quickly to be visible to the naked eye.

The turning point came when a BBC film crew with a high-speed camera offered to rent the equipment to Patek and her colleagues. "That made it possible to see these movements for the first time," says Patek. "It opened up the next 12 years or so of research."

Those first measurements, taken at 5,000 frames per second, showed that peacock mantis shrimps could accelerate their clubs at an eye-popping 104,000m/s/s; comfortably 20 times as fast as any amphibian or reptile.

The peacock mantis shrimps accelerate their clubs so rapidly, they generate an area of unusually low pressure in the water. This encourages the formation of vapour bubbles. Those bubbles quickly collapse again, but when they do they release a massive amount of energy in the form of heat. The temperature in that tiny volume of water briefly rises to levels normally associated with the surface of the Sun.

Maybe plenty of animals achieve these super-fast accelerations

The poor snail finds itself facing a double-whammy: the impact of the shrimp's club itself, and a super-hot shockwave.

Back in 2004, when Patek and her colleagues published their findings on peacock mantis shrimp, the acceleration they had documented was a world record. The elastic spring in the shrimp's exoskeleton could trigger extreme accelerations that no other creature seemed capable of matching.

A key feature behind the shrimp's accelerative prowess was a tiny saddle-shaped spring in its exoskeleton: a spring similar in appearance to those used in some engineering applications. Such springs distribute stress in three dimensions to reduce the chance of failure.

But why would one shrimp stand head and shoulders above all other animals when it comes to accelerating?

There were a few potential answers to the question.

Two vicious-looking pincers are held open in a pose that is ominously reminiscent of an old-fashioned mantrap

Most obviously, perhaps the shrimp is not terribly unusual. Maybe plenty of animals achieve these super-fast accelerations.  But because they are so fast, and so likely to involve small structures, they have simply gone unnoticed by biologists.

To some extent this seems to be true. Equipped with their state-of-the-art recording equipment, Patek and her colleagues began hunting for other examples of super-fast acceleration, and within a couple of years they came across insects that can comfortably beat the peacock mantis shrimp's record.

But the way these insects harness their abilities also hints at a fundamental problem with super-fast acceleration: one that might reduce its likelihood of evolving.

Trap-jaw ants live in the tropics. As their name implies, they have an unusual set of jaws.

They also turn their power on their own bodies

Two vicious-looking pincers are held open in a pose that is ominously reminiscent of an old-fashioned mantrap. The comparison is apt, given that the ant can release an elastic structure in its head to snap those jaws shut on its prey.

Like the mantis shrimp, the trap-jaw ant hauls open its jaws using muscle power, stretching elastic structures in its head in the process. When that elastic energy is released the jaws shut.

Perhaps partly because those jaws are operating in air rather than water, and so meet less resistance, they accelerate faster than mantis shrimp clubs. They reach a peak acceleration in the order of 1,000,000m/s2: ten times greater than the shrimp.

The trap-jaw ants have found an intriguing use for their powers of acceleration. As well as snapping their jaws shut on small insects to stun or kill them, they also turn their power on their own bodies.

By carefully orientating its body just before triggering its jaws, a trap-jaw ant can ensure that its jaws strike the ground just as they snap shut. This sends the ant's head whiplashing upwards with such force that it is launched off the ground and cartwheels through the air.

Paradoxically, super-fast weapons are very slow

The ants use these inelegant leaps to quickly dodge out of the way of a predator. As it happens they are a favourite food of reptiles armed with elastically-powered super-fast tongues – but the ants can accelerate far faster than the reptiles' tongues.

Patek and her colleagues speculate that this behaviour might have come about by accident. What is now a useful escape mechanism might have begun as an inconvenient side-effect of running around armed with a potent weapon.

In other words, super-fast weapons might be difficult to control, and that could reduce their likelihood of evolving.

There is another factor that might explain why super-fast acceleration appears to be rare in the animal kingdom. Paradoxically, super-fast weapons are very slow.

After their study of the peacock mantis shrimp, Patek and her colleagues decided to turn their attention to a related group that had the potential to achieve even great accelerations.

I truly expected the fish-catching mantis shrimp to be faster

Spearing mantis shrimp such as the zebra mantis shrimp have mouthparts shaped like tiny, hydrodynamic javelins, rather than fat clubs. They use these sharp javelins to pierce the flesh of passing fish, which they then eat.

Fish can clearly outpace snails, so logic would suggest that spearing mantis shrimp accelerate their streamlined weapons at rates that far outpace anything seen in smashing mantis shrimp. But they do not. In a 2012 study, Patek and her colleagues discovered that spearing mantis shrimps achieve peak acceleration rates about 100 times lower than their smashing shrimp cousins.

"I truly expected the fish-catching mantis shrimp to be faster," says Patek. "That's the paradigm we all work with. Why be fast? So that you can chase after fast prey."

The mantis shrimp flipped that accepted wisdom on its head. Slower shrimps target fast prey, while faster shrimps target slow prey.

"Clearly being fast helps a cheetah catch fast prey," says Patek. "But when you leap up to these much higher accelerations, these animals are doing something else entirely."

It scuttles over to the snail, slowly primes its weapon, then lets rip with a truly super-fast acceleration

She now thinks she has worked out what is going on. Using muscle power to stretch an elastic structure takes time and effort, and the more elastic energy you want to store, the greater that initial expenditure. Similarly, we find it quick and easy to draw the string on a child's toy bow, but pulling back the string on a full-size longbow takes more muscle power and more time.

Mantis shrimp have also come up against this problem, Patek thinks, and have had to make trade-offs.

A spearing shrimp must react quickly when a fish swims near enough to become a target. It needs a weapon that can be primed and fired quickly. So spearing shrimps have sacrificed one kind of speed for another, ditching speed of weapon acceleration for speed of weapon priming.

In contrast, a smashing mantis shrimp does not have this problem. If it spots a tasty-looking snail it can take its time. It scuttles over to the snail, slowly primes its weapon, then lets rip with a truly super-fast acceleration. "It takes a lot of time to be ultrafast," Patek has written.

But when circumstances allow an animal to become super-fast, the rewards can be enormous.

The force of a blow is determined by both the mass of the weapon and its acceleration. A tiny mantis shrimp cannot evolve a particularly massive weapon, but by accelerating the small weapon it has at 100,000m/s/s, it can achieve an astonishingly powerful impact.

The harpoon can reach peak accelerations of about 50,000,000m/s/s

"I have this chart I show in my lectures now which shows the peak force of a mantis shrimp, which is an animal that's about three or four orders of magnitude smaller in body mass than an alligator," says Patek. "But its peak force is right in line with the alligator's jaw. It's just an alternative mechanism for fracturing or puncturing prey, and in the mantis shrimp's case it means it can eat a far larger snail than it could if it were using its jaws to break open the snail."

However, it is not clear whether or not this "you have to be slow to be fast" idea plays out in all circumstance. Impressive though they are, the 1,000,000m/s/s accelerations achieved by trap-jaw ants are no longer world-beating.

Their record has been blown out of the water by the humble jellyfish.

The stinging cells or "nematocysts" that can make jellyfish so dangerous are yet another super-fast structure, built on the principle of elastic energy storage.

Jellyfish are not constrained to attack only slow-moving prey with their ultrafast weapons

The jellyfish allow water to flow into their nematocysts, causing them to swell up like tiny balloons. The nematocyst walls contain elastic proteins that stretch as the cells expand.

When the cell is triggered and that elastic energy is released, a microscopic harpoon-like structure shoots out of the nematocyst. The harpoon can reach peak accelerations of about 50,000,000m/s/s: 50 times the peak trap-jaw ant acceleration.

"I am not aware of any faster process on the cellular level," says Thomas Holstein of the University of Heidelberg in Germany, who helped make those measurements.

Unlike mantis shrimp, jellyfish are not constrained to attack only slow-moving prey with their ultrafast weapons. A fast-swimming fish or shrimp is just as likely to be stung.

"Patek's hypothesis doesn't apply for nematocysts," says Holstein. "They are weapons designed for extremely rapid action in order to hit rapidly-moving targets."

Perhaps that is in part because jellyfish are armed with battalions of nematocysts, rather than carrying a single super-fast weapon. Even if one nematocyst fails to hit a fast-moving target, there are dozens more primed and ready to fire. It might be the ultimate ultra-fast weapon.

I am not aware of any faster process on the cellular level

What makes this all the more astonishing is the fact that the cnidarians, the group of animals that includes jellyfish, is one of the most ancient on Earth. Jellyfish have been a feature of the oceans for over half a billion years. Given that nematocysts are a defining feature of the group, they might have been achieving super-fast accelerations for much of that time.

"This stuff has been around a really, really long time," says Patek. It has just taken time for us to develop the technology to appreciate it.

We can only imagine what secrets remain locked away in the world of super-fast acceleration. But, in a realm where fast is slow, there is a good chance that future discoveries will be every bit as counter-intuitive as those we have already made.

Join over five million BBC Earth fans by liking us on Facebook, or follow us on Twitter and Instagram.

If you liked this story, sign up for the weekly bbc.com features newsletter called "If You Only Read 6 Things This Week". A handpicked selection of stories from BBC Future, Earth, Culture, Capital, Travel and Autos, delivered to your inbox every Friday.