A news item about an Iranian viper with a spider perched on the tip of its tail passed through my morning feed. Not exactly a sensation in itself: nature has plenty of deceivers, and a snake’s tail is a well-known hunting tool. It was the final line that caught me. CT scans showed that the vertebrae in this tail look perfectly ordinary — the lure the snake uses to draw in birds is made entirely of soft tissue and leaves no trace whatsoever in the skeleton.
So this is not a story about a snake. It is a story about how many such lures, trunks, and larynges have dropped out of the paleontological record simply because they were not made of bone.
The Iranian spider-tailed horned viper Pseudocerastes urarachnoides is a snake about half a meter long from western Iran, living on the gypsum cliffs of the Zagros Mountains. By day it sits in a crevice and waits, while its tail, with a flattened “abdomen” and long scaly “legs,” looks like a spider and moves like one. A bird drawn to that spider becomes lunch.
All this was known by 2006, but it began with an oddity. In 1968, an expedition collecting mammals came to Iran; it picked up amphibians and reptiles along the way, and one of the vipers ended up at Chicago’s Field Museum. There, in 1970, it was examined by zoologist Steven Anderson.
The first thing he thought, looking at the specimen through the glass of the jar, was that a solifuge — a large desert spider, a camel spider — had latched onto its tail. Then he saw that the “spider” was not attached to anything. It grew out of the tail. The snake was identified as a Persian horned viper, the growth was deemed a tumor or a parasite, and that was where it stopped: one specimen was not enough to say whether this was a deformity or the norm. “The specimen sat there, but was not forgotten, for nearly four decades,” the original description says.
A second specimen, collected by zoologist Hamid Bostanchi, settled the matter. It had exactly the same “spider” on its tail. Two identical oddities in two different snakes are no longer a disease, but a species trait. In 2006, four authors described the viper in the journal of the California Academy of Sciences and named it for its tail: from the Greek “oura” (tail), “arachne” (spider), and “oides” (resembling).
The description turned out to be more restrained than the snake itself. The authors had to state frankly that they were describing the most elaborate tail ornament known in snakes — “with the possible exception of rattles in rattlesnakes” — from two specimens, one of them badly damaged, and that they could only guess at the tail’s purpose. No one had directly observed it hunting at the time. But the second specimen had a digested lark in its stomach, and as early as 1988 the herpetologist Gasperetti had noted that the tail tip of adult Pseudocerastes is usually dark, which “suggests caudal luring” — in other words, the idea of a lure had been hanging in the air long before anyone tested it.
Two minor details in the literature on this snake still do not line up, and I will leave them as they are. News stories and press releases date the second specimen to 2003, while the type-specimen list in the original description gives 15 May 2001. The holotype number in that list is one thing; in the text of the same paper and in Fathinia’s later work, a completely different number appears. No one has ever corrected it.
Then museum logistics did their job. At the end of 2024, a new CT suite opened at the Field Museum: an object is X-rayed thousands of times and assembled into a three-dimensional model without cutting into the specimen itself. For a holotype — the reference specimen of a species — this matters most of all: it cannot be damaged.
The holotype was among the first things rolled into the scanner. Sarah Ruane, the museum’s curator of herpetology, calls it her favorite specimen and carries it around for displays and tours. The study was Georgios Georgalis’s idea, from the Institute of Systematics and Evolution of Animals at the Polish Academy of Sciences in Kraków. He is a paleontologist who studies snake vertebrae precisely in order to read fossil finds through them, and here he had come across a living species with an obviously extraordinary tail.
The reasoning was clear. In sand boas and rubber boas, the vertebrae at the very tip of the tail are so strange that they are unmistakable. Georgalis wanted to see whether the Iranian viper had anything similar.
It did not. “The spider tip is only external morphology; the tail vertebrae themselves are typical of most vipers,” he says. “There was nothing in this snake’s skeleton to indicate how extravagant it looks on the outside.” The lure is assembled from elongated, heavily modified scales, with no bony foundation at all. In a press statement, he repeats the point more sharply: if all we had today were fossil remains, we would not be able to reconstruct — or even imagine — such a tail.
The qualification matters, and it was added by zoologist Kurt Schwenk of the University of Connecticut, who was not involved in the work. As he puts it, scales turning into spines, horns, and projections is commonplace in lizards and snakes, and beyond the head it almost never affects the skeleton. So the result did not surprise him. Something else did: the viper’s tail vertebrae seemed unusually short. Short vertebrae make a tail more flexible, and flexibility is needed for the lure to trace the kinds of twists that create the impression of a living spider.
So the skeleton did say something after all. It just said something other than what people expected of it.
Behzad Fathinia and his colleagues were able to test the hypothesis in the field. For two and a half years, from April 2011 to June 2013, they watched twelve snakes in their territories and recorded six successful hunts: one for an Isabelline shrike and five for reed warblers. Only once did the camera catch the actual capture — in the other cases, the birds were already in the snakes’ mouths by the time the researchers arrived.
The crucial finding was this: the snake waves its tail even when there are no birds around, and merely speeds up when birds appear. That breaks the textbook definition of caudal luring, in which the behavior is triggered by the prey’s presence. The wording had to become more cautious: without birds, the snake waves less often, but it does not stop.
One more detail: the lure grows. It appears after birth and increases along with the snake’s body length, meaning it is not a ready-made feature but an organ that matures. And it wears down. Fathinia observed one adult female with a bloody tail tip — apparently a bird had managed to reach the “spider” and bite it, rather than the other way around. The lure used for catching becomes a lure itself.
Everyone now knows what it looks like: the Iranian viper’s hunt appeared in the David Attenborough series Seven Worlds, One Planet. Its venom, too, is tuned to the task — in this viper it hits birds harder than the venom of its closest relatives, which is better suited to mammals and amphibians.
Now to the paleontology this text was written for. There is a 2012 book, All Yesterdays, a collection of dinosaur drawings, though half its point lies in a reverse exercise: the authors applied a standard paleoart technique to modern animals. A hippo, elephant, or swan shrink-wrapped in skin over its bones looks like a monster in which the original is unrecognizable. The problem was called “shrink-wrapping”: an artist puts exactly enough soft tissue onto a skeleton for the silhouette to remain identifiable from the bones. Taken to its limit, it produces “zombie dinosaurs,” every bone visible through the skin.
Paleontologists have spent years discussing animals that push this problem into absurdity. The elephant: its skeleton contains no trace of a trunk, not even a hint of its size. Sauropods: their nostrils sit high on the skull, and in the 1970s Walter Coombs proposed that these dinosaurs might have had trunks. Robert Bakker drew a trunked Diplodocus in The Dinosaur Heresies, while the authors of All New Dinosaurs and Their Friends (1980) wrote frankly: we will probably never get direct evidence, but it is interesting to see what it might have looked like.
Evidence did turn up after all — and in the bones. The soft tissues around a nostril, a structure densely threaded with blood vessels, leave openings and channels in the skull. In sauropods, those traces lie in front of and below the bony nostril, meaning their fleshy nostril was at the tip of the snout, not at the end of a trunk. There is no neurological support for a trunk either: in an elephant, the facial nerve where it emerges beneath the jaw joint is about a centimeter wide; in Diplodocus, the root of the same nerve is a millimeter; in Camarasaurus, three. And the teeth are worn in a way that shows the animal stripped branches directly with its jaws, which does not sit well with a trunk.
The conclusion: sauropods did not have trunks. But the same analysis ends with a line that brings us back to the Iranian viper: the absence of a trunk does not mean sauropods lacked other soft-tissue ornaments — inflatable sacs, wattles, and crests. Bone can refute. It cannot always confirm.
Here are a few more cases that show both sides.
The larynx in all tetrapods except most modern birds is made of cartilage, and cartilage fossilizes far less readily than bone. So until 2023, no larynx had ever been found in an extinct non-avian reptile, and the evolution of this organ remained almost unknown.
Then a find appeared: in a juvenile ankylosaur, Pinacosaurus grangeri from Mongolia, the larynx had ossified — with a movable joint between the cricoid and arytenoid cartilages and an enlarged cricoid cartilage. In structure, it resembled a bird’s apparatus more than a crocodile’s. The authors honestly qualify this: perhaps it is a feature specific to ankylosaurids, or perhaps all the other larynges simply did not survive. The story is exactly the same for the avian vocal organ: the syrinx is built from hard cartilaginous rings, and it has been found only once in Mesozoic rocks — in the Antarctic bird Vegavis, which lived more than 66 million years ago.
The second case is feathers. In 2007, six bumps were found on the forearm of a Velociraptor from Mongolia — traces where flight feathers had been attached by ligaments. Here, an asymmetry applies that the study’s lead author put well: the absence of bumps does not mean there were no feathers, but if the bumps are there, the feathers definitely were too.
And the third case, the oldest of all: Iguanodon. Gideon Mantell decided the dinosaur had a horn on its nose — perhaps because he had the modern rhinoceros iguana, with its small snout horn, in mind — and attached to the face the one large bone for which he could find no other place.
That is how Iguanodon, horn on its nose, arrived at Crystal Palace: before the sculpture opened in 1853, several prominent scientists even dined inside the model. Only after discoveries in Belgium in the 1870s did it become clear that the bone was a thumb spike. Even with the object itself in hand, without soft tissue you get the wrong animal. And its true size remains unknowable anyway: the spike was covered by a keratin sheath, which does not preserve.
Something of soft tissue does remain, but through very narrow channels. The first is exceptional deposits, known as Lagerstätten: where sediment accumulated rapidly and without oxygen, not only skeletons but soft tissues survive. That is how we learned about the soft-bodied inhabitants of the Burgess Shale — a little more than five hundred million years old — about feathers from Solnhofen limestone, and about soft tissues from Germany’s Messel. The scale of this window onto the record has been measured: in the Burgess Shale, only about 14% of genera had calcified, and therefore preservable, skeletons in life. The rest of the community — soft, unprotected creatures — exists for us only because suitable conditions came together there once in history.
The second channel is traces. Footprints, burrows, bite marks, coprolites: behavior leaves a mark behind in place of a body. The trouble is that it is nearly impossible to identify the maker from such a mark. Different animals leave similar traces, and paleontologists have to invent separate names for them, unrelated to biological species. Even where behavior has been preserved, we usually do not know exactly who did it.
The third channel is museum shelves. The Chicago viper was recognized only because it never went anywhere: it sat in a jar from 1968, waited for a second specimen, and then for a new CT suite. More than half a century passed between the first photograph of the “spider” on its tail and the scan of its insides, and for most of that time the specimen simply gathered dust. Its keeper calls the snake her favorite exhibit and takes it on tours.