Circular tattoos on the neck of a Peruvian mummy suggest that a medical treatment similar to traditional Chinese acupuncture was being used in south America in ancient times.

I covered the research for New Scientist recently, but I really don't think the story works without pictures of the tattoos so here they are, along with a bit more information about ancient tattoos...

The 1000-year-old mummy was found unwrapped in the sand of the desert at Chiribaya Alta in Southern Peru, in around 1990. She has decorative tattoos representing birds, apes, reptiles and symbols on her hands, arms and lower left leg.

Decorative tattoos are fairly common in human mummies. The oldest known tattoo dates from 6000 BC from the south American Chinchorros culture, and shows a thin pencil moustache tattooed onto the upper lip of a male adult.

But this mummy also has some tattoos that don't appear to be decorative - circles of different sizes placed at irregular positions on her neck. They aren't particularly pretty, and would have been hidden by her hair.

Konrad Spindler, an archaeologist best known for his excavations of Otzi the Iceman, wondered if they might have been part of a medical or therapeutic ritual, perhaps to ease neck pain. He made drawings of the tattoos (see top pic) and brought them back to Europe, along with samples of skin punched from the tattoos.

Maria Anna Pabst from the University of Graz, Austria, and her colleagues used various microscopy techniques to investigate what the tattoos were made of. The decorative markings were made of soot, which is quite common for ancient tattoos. But the dye in the neck circles consisted of partially burned plant material. It's the first time that two different kinds of tattooing materials have been found in the same mummy (Journal of Archaeological Science vol 37 p 3256).

Pabst told me that this is strong evidence the neck circles were meant for a separate purpose from the decorative tattoos. "If you use different materials, they have different functions," she says.

The idea that tattoos could have a medical purpose has been suggested before. For example, Otzi himself has the oldest tattoos ever found on a European mummy - some 15 groups of lines and shapes on his back and legs.

Spindler and his colleagues pointed out in the Lancet in 1999 (vol 354 p 1023) that Otzi's tattooed lines are located very close to classical acupuncture points, and suggested that there might have been a medical system similar to acupuncture practised in Central Europe 5200 years ago.

X-ray scans have shown that Otzi had chronic wear in his hip, knee and ankle joints and his lower spine, and the team concluded that his tattoos matched the acupuncture points that would be used to treat these ailments. They could have been part of the treatment itself, or used as a guide to self-treatment, showing Otzi where to stick the needles when he was in pain.

Of course this evidence is circumstantial, and it's extremely difficult to know what was really intended. But Pabst's comparison of two types of tattoo in the Peruvian mummy adds another useful line of evidence. It is one of only two mummies known with both decorative and what appear to be medicinal tattoos. (The other is a Scythian nomad prince from the Altai mountains, dating from around 500 BC, who has pictures of mythical creatures tattooed on his arms, shoulders, chest, back and right leg, as well as a series of dots down his spine and on his ankle - see pic, left.)

Pabst points out that in her mummy too, the circles match very well with traditional Chinese acupuncture points (see second pic, above), which is a surprise because China and south America are on different land masses. The knowledge must either have been carried far north via the Bering Strait, or developed independently.

The neck therapy would have used needles in the same way as acupuncture, Pabst reckons, but burning medicinal plants into the skin at key sites was perhaps thought to confer extra benefits. For further clues Pabst even showed Spindler's drawing of the tattoo circles to a modern-day shamanic healer in Peru - he said that the designs suggested to him a strengthening ritual, that would have been carried out on an upper class subject.

In case you're wondering, Otzi's tattoos were made of soot - Pabst carried out microscopic analysis of them in 2009 (Journal of Archaeological Science, vol 36, p 2335). But it seems unlikely that much more information will come to light on this topic. Spindler died in 2005, and Pabst has now retired, so the project has sadly come to an end.

What's like rockclimbing, and listening to an allegro molto? Being a bee, according to behaviour researcher Rodrigo De Marco. I've just written a short news story for New Scientist about a study in which De Marco and colleagues used high-speed video inside a hive to glean information about the honeybee's waggle dance. When I interviewed De Marco about his research he had some lovely insights into the mind of a bee, as well as the challenges of decoding its famous dance. His comments didn't fit into the story so I thought I'd post them here as a q&a.

First some background - that honeybees use an ingenious dance to communicate the location of nearby food sources was discovered in 1946 by Karl von Frisch. Each dance consists of a series of "waggle phases" during which the bee shakes its body from side to side (see video). Von Frisch subsequently worked out that the dancer's orientation relative to gravity during each waggle phase gives the direction to the food relative to the sun's position, while the number of side-to-side movements the dancer makes gives the distance.

But in the decades since von Frisch's work, researchers have made little progress in working out other bees decode the information in the dance. Studies have shown that followers can sense vibrations and flows in the air around the dancing bee, however they also touch the dancer directly. De Marco's work, published in Animal Behaviour, is an attempt to solve that mystery...

Q: How does a honeybee move during its waggle-phase?
A: Imagine yourself rock climbing, with your hand and feet held, and moving your backpack from side to side in a controlled manner. Like that! We also know that the dancer does not keep all its feet still while wagging its body from side to side; it moves them in a systematic fashion, pretty much as you would need to move your hands and feet in order to climb up. A single waggle-phase can last from a fraction of a second to several tens of seconds, whereas an entire dance can involve from a single to hundreds of waggle-phases.

Q: You describe the question of how information from the dance is transferred to followers as a "major gap" in understanding. Why has it been so difficult to work this out?
A: To start with, there is a grave lack of understanding of the sensory horizon of honeybees. What is meaningful input for a honeybee? How are different meaningful inputs combined? Answering this question also involves solving several separate puzzles. For example, what is the nature of the uncertainty that the dancers' audience experiences prior to and during dance following? What are the relevant cues and signals associated with the dance? How are such cues and signals being integrated from the sender's and the receiver's side of the communication process? To make matters worse, it is difficult to observe the behaviour of both dancers and followers simultaneously, and even more difficult to observe them both inside and outside the hive. So it is very hard to track the ultimate effect that the dance has on the behaviour of the individual followers.

Q: You used high-speed video to record more than 40 dances, with nearly 400 followers. What did you find?
A: We found that the higher the number of the dancer's wagging movements, the higher the number of the followers' antennal deflections.

Q: So in a sense the followers use their antennae to "count" the number of wagging movements?
A: Using the word counting would imply that followers can compute deflections of a certain size as discrete events, but we do not know whether this is how the deflections are processed by the brain. I think that instead the bees may be sensing the length of time over which the stimulation occurs, as a person might estimate the duration of a piece of music. Imagine the difference between listening to either a short or a long allegro molto.

Q: How do the followers estimate the orientation of the dancer with respect to gravity?
A: We found that the type of stimulation the followers receive depends on their position relative to the dancer. For followers who face the dancer from the side, contacts are more obvious and regular, and involve both antennae, whereas from behind they are more subtle, a bit less regular, and involve one antenna at a time. The followers might be able to use this information, combined with knowledge of their own position, to figure out the dancer's orientation. Imagine a situation in which you can easily orient your body either vertically or horizontally, and that when you find the right orientation, this is signalled to you by the occurrence of music, or a certain kind of motion. You would only need to figure out what the orientation of your body is when the music (or motion) occurs to find out what the relevant orientation is.

Q: Are the antennal deflections sufficient to transmit all the information that the followers need, or are they using other information sources too?
A: It would be unwise to imagine that honeybees rely on a single pathway. I think that we would learn a great deal by looking at how honeybees interact with each outside the hive. Can they follow each other? If yes, to what extent? The dance may be great advertisement, but it might well be that more is needed to bring the audience to the actual target.

Q: Why do you find these dances so fascinating?
A: Because their occurrence and shape can be predicted by a human observer. Hence, they are useful to gather insights about how information flows in nature. There are other communication systems in nature that humans can interpret, but in most cases the function they serve isn't so obvious. With the dance, the ultimate goal is simply to recruit comrades. Bees also have quite a restricted repertoire of behaviours. So we can observe the behaviour of dancers and followers and make very clear predictions about what we expect to see. That gives us a good chance of finding ‘meaningful' correlations. A panacea for ethologists! However, there is a danger here too: the study of the honeybee dance has frequently led to oversimplified interpretations about communication and behaviour. It seems that, sooner or later, students of the honeybee dance tend to find themselves embracing their own hypotheses a bit too kindly, instead of fighting them properly. I still fail to understand why - could it be that the idea that honeybees ‘read' the dance as we do becomes at some point irresistible?

PS De Marco carried out this research at the Freie Universität Berlin, however he is now at Max Planck Institute for Medical Research in Heidelberg, studying behaviour development and the development of neuronal circuits underlying the so-called stress response.

PPS OK so this doesn't have anything to do with the Antikythera mechanism. But De Marco's paper does have "decoding" in the title :-)

OK, now I'm obsessed with Paleodictyon. I mentioned this mysterious hexagonal entity briefly in my last post as an example of a trace fossil drawn by Leonardo da Vinci (see his sketch, left). But what is it - undersea mushroom garden, or the world's weirdest sponge?

Paleodictyon is a fossil found in sedimentary rocks that once formed the deep seabed. It consists of a network of tunnels or tubes that would have sat a few millimetres underground, with vertical shafts leading up to surface. Palaeontologists traditionally interpret it as a burrow system, dug by an unknown maker. But Paleodictyon forms a perfectly regular hexagonal pattern, like a honeycomb (see fossil cast made by Yale palaeontologist Adolf Seilacher, left). What kind of animal could have dug such geometrically perfect tunnels?

Now it turns out that this hexagonal structure may not be a burrow system after all, but the remains of a bizarre creature.

The story starts in 1976, when marine geologist Peter Rona of Rutgers University was using a giant sled carrying cameras and echo sounders to map the seabed of the Mid-Atlantic Ridge, around 3.5 kilometres down. In an area populated by black smoker hydrothermal vents, he found some strange patches of dots. The dots were holes, arranged in a "strikingly symmetric" hexagonal pattern. The patches were quite small, around 5 centimetres across, but there were thousands of them (see one of Rona's photos, below).

The patterns were so regular, and so odd, that Rona apparently even wondered for a moment whether he had stumbled across the footprints of aliens from outer space. Then Seilacher got in touch and pointed out the similarity to Paleodictyon. Seilacher and other palaeontologists had assumed that whatever made Paleodictyon, it went extinct 50 million years ago. But Rona had found the modern-day equivalent.

Between 1990 and 2003, Rona made several trips down into the abyss, in the tiny manned sub Alvin, to scoop up cores of mud that contained the mysterious hexagonal structure. The project was featured in IMAX film, Volcanoes of the Deep Sea, in 2003, and Rona finally published his results in September 2009.

Rona's work was covered in this New York Times article, but I found it frustratingly light on detail regarding Paleodictyon itself so I thought I'd summarise the main findings here.

When Rona studied Paleodictyon in the lab, he confirmed that the holes he had seen on the seabed are indeed the openings of vertical shafts, which lead to a honeycomb pattern of horizontal tunnels beneath the surface, just as seen in the fossil version (see model made by Hans Luginsland, left). The surface of the structure forms a raised shield-shaped dome in the seabed.

Disappointingly Rona found no trace of any organism associated with the tunnel structure - no body parts, biological material, or DNA. Little sediment lands on this part of the seabed, so it may be that fresh-looking holes can persist on the seafloor for hundreds of years, even after the owner/inhabitant is long dead. But intriguingly, when Rona tested a model of the structure in a flume tanks, he found that the shield-shaped profile makes it self-ventilating - in other words, water currents are deflected through the structure in a way that keeps all of the tunnels aerated and supplied with organic particles.

The study has left Rona and Seilacher with completely different views on how the structures are formed. Seilacher sees Paleodictyon is an especially complex member of a group of fossils called graphoglyptids, which palaeontologists believe are all burrows. They come in a dizzying array of shapes and patterns, including hexagons, spirals and tree-like structures (there's a nice paper showing the range of graphoglyptids here). But they all divide a given surface into equal subunits, just as human-made drainage systems do.

Seilacher reckons that graphoglyptids are basically subsurface mushroom gardens, in which foods (bacteria or fungi) are cultivated by unknown animals, probably some kind of worm. So in theory it might be possible to send a robotic sub to sit for hours next to a fresh burrow and wait for the unknown tracemaker to visit its farm.

Rona, on the other hand, thinks the hexagonal pattern of tubes represents the remains of the organism itself. His tests found that levels of bacteria are no higher inside the tunnels than outside - evidence against the farming theory.

It is also hard to imagine how any creature could have dug such a regular pattern. You can get hexagonal patterns in nature from closely packing regular subunits, such as soap bubbles, eggs, corals or honeycomb cells. To weave or burrow them is much harder, yet defects such as elongated sides or missed vertices are very rare. A computer simulation carried out in 1993 showed that the burrower would need "outstanding navigational skills" including the ability to execute 60-degree turns with an accuracy of 2 degrees, and to measure the length of the hexagon sides with an accuracy of less than 1% (Journal of Geological Education 41, 159-163; 1993).

More recently, another study applied graph theory to the tunnel network. The researchers calculated the total tunnel length for networks of varying sizes, and used a theory originally developed by the mathematician Euler to work out the shortest path an animal would need to take to visit all the parts of its burrow. One fossilised example of Paleodictyon is a metre across. Its hexagons have a side-length of just 5 millimetres, with a tube diameter of less than 1 millimetre, so the animal that dug this burrow could not have been more than 5 mm long, or more than 1 mm across. The researchers calculate that the total tunnel length in this fossil is around 230 metres, and that the burrower would need to revisit previous parts of the path at least one third of the time. To maintain its burrow system, the organism would need to move an improbably long distance relative to its own size, perhaps 46,000 times its own length and 250,000 times its width.

Instead, Rona thinks that the hexagonal tube system is all that remains of a strangely adapted form of sponge. Once the animal dies, its body parts get eaten away, leaving just the imprint of its body shape.

Rona suggests the most likely candidate is a hexactinellid sponge. These are also known as glass sponges, and have an internal skeleton made up of a four- or six-pointed silica lattice (see dried skeleton of the deep sea glass sponge Euplectella, above left). Perhaps one of these sponges, flattened into a horizontal layer, might form a hexagonal network of tubes like Paleodictyon. The vertical shafts would then allow aerated water and food particles to circulate through the animal's body.

Both ideas seem quite mind-boggling but I reckon Rona's explanation is the most convincing. But if Paleodictyon is a sponge, not a burrow, what does this mean for other graphoglyptid "burrows"? Are they actually creatures too?

After those lovely X-ray images of the shipwrecked watch in my last post, here are some more pictures. I recently wrote a feature article for New Scientist about how Leonardo da Vinci was deciphering trace fossils hundreds of years before mainstream palaeontologists caught up with him. The feature is here... the images that go with it are missing from the online version though, so I thought I'd post them here instead.

It's fairly well-known that da Vinci worked on body fossils (the direct remains of prehistoric organisms) but trace fossils - the tracks and trails these creatures leave behind - are much trickier to interpret. In da Vinci's time, scholars were arguing over why stone seashells were being found in the rocks on top of mountains. Trace fossils formed a key part of his arguments against the two prevailing theories - that the lofty seashells were the remains of sea creatures deposited during the Biblical flood, or that they had spontaneously grown inside the rock (yes really!)

Palaeontologist Andrea Baucon picked all this out of Da Vinci's cryptic notebooks, in particular the Codex Leicester. The top pic shows an extract - the highlighted text translates as: "Between one layer [of the rock] and the other there remain traces of the worms that crept between them when they had not yet dried. All the sea mud still contains shells, and the shells are petrified together with the mud." Baucon's analysis is published in Palaios (vol 25, p 361).

Baucon believes that da Vinci may also have included some trace fossils in his art. He has identified two possible candidates - slithery-looking tracks in "Madonna of the Yarnwinder" (second pic, just at the bottom of the stick) and "Virgin of the Rocks" (third pic, marked by arrow). When I asked Martin Kemp, an expert in da Vinci's art based at the University of Oxford, he reckoned that the marks in the Madonna are actually threads of yarn, but couldn't rule out that those in the Virgin were meant to represent trace fossils (although he was pretty sceptical).

There is one other example though that does seem quite convincing - a scrawled hexagonal pattern (last pic; left) which appears alongside some sketches of body fossils in another of da Vinci's notebooks, Codex I. Baucon points out that it looks just like Paleodictyon, one of the most common and characteristic trace fossils of the area of Italy where da Vinci was working (last pic; right). It is a network of burrows, originally dug into the sand of the seabed, made up of vertical, hexagonal shafts. The organism responsible is still a mystery. Scientists recently sent the submersible Alvin looking for it at 3500-metre-deep underwater vents. They found modern-day versions of the burrows, but no creatures were inside.

Thanks to Andrea Baucon for sending over the images. 

In September 1653, a small warship called the Swan formed part of Oliver Cromwell's forces as they attacked Duart Castle, a staunchly royalist stronghold in Mull on the west coast of Scotland. A thousand troops disembarked from the fleet's six ships, only to find the castle abandoned. But worse was to come. On 13 September a nasty gale sank three of their anchored vessels.

Robert Lilburne, the senior government commander in Scotland, wrote to Cromwell: "While our men staid in this Island the 13th instant there hapned a most violent storm, which continued for 16 or 18 houres together, in which wee lost a small Man of Warre called the Swan that came from Aire." Most of the soldiers were still onshore, but they had to watch, helpless, as more than twenty of their comrades met their deaths.

The wreck of the Swan was discovered in the 1970s and excavated in the 1990s. Artefacts including an iron gun, the brass lock-plate from a pistol, an ornate sword hilt and a hoard of silver coins were recovered and taken to the National Museum of Scotland in Edinburgh.

Why am I telling you this story now? Well, one of the most intriguing items salvaged from the wreck was a 17th-century pocketwatch. It was covered in barnacles and corroded almost beyond recognition (see top pic). A fuzzy X-ray image taken at the museum showed that some of the gearwheels inside had survived, but didn't provide any useful information about what kind of watch it was, or what state the mechanism was in. The only way to find out more would have been to break the watch open, something conservators hate to do, so it was put to one side.

Until now. Museum scientist Lore Troalen and her conservator colleagues Darren Cox and Theo Skinner have recently used a state-of-the-art X-ray scanning technique to probe the watch's innards. This is the same imaging technique that was used so successfully on the Antikythera mechanism in 2006 - indeed the researchers got the idea from reading about the Antikythera mechanism in Nature.

Troalen and Cox packed the watch in a Tupperware sandwich box and took it to Andrew Ramsey of X-Tek Systems in Tring, Hertfordshire, part of the team who imaged the Antikythera mechanism (you might remember him and his colleagues from my book). The technique, called 3D computed tomography (CT) involves imaging an object from thousands of different directions, then combining all of the data to produce a three-dimensional virtual reconstruction of the object's internal structure.

Troalen says she didn't hold out much hope that the watch would be in great condition, but she hoped at least to find a mark or signature that could help to identify the original owner of the watch.

The results, published in The International Journal of Nautical Archaeology, are stunning. Although the steel parts of the watch - its single hand, and the studs and pins that originally held the mechanism together - have corroded away, the majority of its components are brass, and in extremely good condition (see pics 2, 3 and 4). I've just written an article about the project for Nature, which has a slideshow of more of the X-ray images than I'm showing here, as well as an animated fly-through of the watch's insides.

The X-rays show that the watch was driven by a single gear train, as was typical for watches of the period. Roman numerals marked the hours, and there are plenty of decorative touches, including floral engravings, and an English rose in the centre of the dial. The watch, as well as its virtual reconstruction, are on display until 2011 in the Treasured gallery of the National Museum of Scotland.

Other, more impressive, watches from this period do exist, but researchers are excited about this study because (along with the Antikythera mechanism) it shows how much detail can be preserved in objects after hundreds or even thousands of years under the sea.

Since the work on the Antikythera mechanism, 3D CT is starting to revolutionise the study of fossils (something I've written about before) and hopefully this watch project will be one of many on archaeological finds. The technique is likely to be particularly useful for studying objects, from terrestrial or underwater sites, that are embedded within layers of corrosion products, and that contain small, detailed structures, for example inscriptions or mechanical parts. 

So did Troalen and her colleagues find that signature? They did indeed - that of Nicholas Higginson, a watchmaker who was based in London in the years before the Swan went down (see last pic). The watch would have been expensive, so its most likely owner was the Swan's captain, who was called Edward Tarleton. Archaeologist Colin Martin, who led the original excavation of the Swan, says that the watch was found in the vicinity of the stern cabin, where the captain would have lived, along with a sword, and "quite a bit of booze".

Tarleton didn't go down with the ship, by the way. He ended up as Lord Mayor of Liverpool.

***Thank you to Trustees of NMS for letting me use these images***

UPDATE (13 Oct): I forgot to say that X-Tek is now owned by Nikon Metrology. There's more info about their CT technology here (they normally use it for industrial applications such as checking aircraft turbine blades for weaknesses).