The flower that ruined fortunes, the wombat's cubes, and the book no one can read

The appetite was real, and so were the prices. Orchids changed hands at London auction rooms for sums that could buy a house, and the record climbed as the century wore on. In 1904 a single plant of Odontoglossum crispum named 'Fred K. Sander' went to the collector H. T. Pitt for fifteen hundred pounds, and at the same sale Baron Schroeder outbid him for a rival variety at twelve hundred and fifty.[2] Two decades earlier a prize crispum had already reached a hundred and sixty guineas.[2] The scarcity behind those numbers was manufactured as much as natural, because the plants were hard to keep alive and harder still to multiply.[2]

The trade had a king. Frederick Sander ran a nursery at St Albans with some sixty glasshouses, was appointed Royal Orchid Grower to Queen Victoria in 1886, and kept twenty-three collectors scattered across the tropics to keep the benches full.[4] His reach was industrial. Sander once claimed to have imported more than a million plants of a single New Guinea species in one push, and Colombian customs recorded 88,567 kilograms of orchids leaving the country in 1887 alone.[4][2] The rival dealer William Bull is thought to have taken in about two million plants across two shipments in 1878.[2]

Behind every crate was a lethal job. The men the nurseries sent out, the orchid hunters, raced one another into the forests of South America and Southeast Asia and often did not come back. William Arnold drowned in the Orinoco, Gustav Wallis died of fever in Ecuador, and Albert Millican, who wrote a book about the life, was stabbed to death on a later expedition.[3] A party of eight sent to the Philippines in 1901 came apart almost entirely, with one man killed by a tiger, another burned, and five who simply vanished, leaving a lone survivor to carry home his Phalaenopsis.[3] The Bohemian collector Benedikt Roezl, who worked with an iron hook after a machine took his left hand, once shipped 27,000 living plants of which just two reached Europe alive.[5]

The hunters stripped the forests as they worked. Millican's crews pulled some ten thousand plants from one region while felling around four thousand trees to reach them, and observers of the day guessed at a tree lost for every three orchids that survived the voyage to a European bench.[2] Most did not survive it. One consignment of sixty thousand Odontoglossum yielded about seven thousand living plants, and a shipment of forty thousand Miltoniopsis reached Sander with only three thousand showing any life.[2] The losses deepened the scarcity, and the scarcity lifted the prices.

What nobody could do was raise the plants reliably from seed. Orchid seed is little more than dust, and in the wild it will not sprout unless a particular soil fungus invades it first, a dependence the French botanist Noel Bernard worked out only in 1899.[2] Until Lewis Knudson found a way to grow seedlings on nothing but sugar and agar in the early 1920s, every fashionable orchid had to be torn from a tree somewhere and carried across an ocean, usually in a sealed glass Wardian case that kept a lucky few alive.[2] That advance eventually broke the spell. The mania had already crashed once, in 1894, when so many plants flooded the rooms that lots sold for a few shillings, and cheap propagation finished what the glut began.[2]

The flowers won in the end. Orchids are now grown by the million in laboratories, and the moth orchid on a supermarket shelf is among the most heavily traded flowering pot plants on earth, a market worth well over a billion dollars a year.[7] The wild species paid for the boom. Odontoglossum crispum, the darling of the auction rooms, is now judged endangered in its Colombian cloud forests, and the entire orchid family is listed under CITES, the treaty that governs trade in threatened wildlife.[6][2]

The habit is not a party trick. A wombat's eyesight is poor and its sense of smell sharp, so it stakes out territory by scent, perching its droppings on rocks, logs and mounds where the message carries farthest.[6] Flat-sided pellets stay where they are set instead of rolling off, which may be exactly the point.[5] A single animal can turn out eighty to a hundred of them in a day, each a rounded cube a little over two centimetres on a side.[1]

The shape is built inside an unusually long gut. A wombat's intestine runs some six to nine metres, roughly ten times the animal's body length, and it wrings almost all the water out of what passes through, dropping the moisture content from about eighty per cent in the upper colon to a little over half by the end.[1] Food can take the better part of two to three days to make the trip, kneaded by something on the order of a hundred thousand muscular contractions along the way.[1] By the time it nears the exit it is dry, stiff and ready to hold an edge.[1]

The explanation came from a team with a taste for strange plumbing. Patricia Yang and David Hu of Georgia Tech, working with Scott Carver of the University of Tasmania, first laid out the puzzle at a fluid-dynamics meeting in 2018 and won an Ig Nobel Prize, the award for research that makes people laugh and then think, the following year.[3][4] Their peer-reviewed answer arrived in 2021. Along the last stretch of the intestine, roughly its final sixth, the muscular wall is not uniform but alternates between stiffer and softer bands around its circumference, differing in stiffness by about fourfold, and as the dried mass is squeezed past those bands the rhythmic contractions press flat faces and sharp corners into it.[1] In their model the corners appear in fewer than ten squeezes.[1]

That is the settled part, and a later study adds a second act. In 2023 the same laboratory proposed that as the feces dry inside the gut they crack at regular intervals, much as cooling lava or drying mud splits into columns, and that this cracking fixes the length of each cube once the water content falls below roughly two-thirds.[2] The two accounts fit together rather than compete, with the uneven muscle shaping the square cross-section and the drying deciding where one cube ends and the next begins.[2] A 2025 study turns from how to why, reporting that wombats read one another's deposits like a guest book, each animal's scat carrying a chemical signature that passing neighbours stop to inspect.[5]

The curiosity has a practical tail. Industry usually makes a cube by pouring it into a mould or cutting it from a larger block, and the wombat hints at a third route, coaxing a shape out of a soft material while it is still moving.[3] Whether that idea ever leaves the laboratory is anyone's guess, but the animal has at least shown it can be done.[6]

The book is catalogued plainly as Beinecke MS 408, though the world knows it as the Voynich Manuscript.[1] It is a modest object, about two hundred and forty pages of calfskin vellum, a little under a foot tall, holding some thirty-eight thousand words written left to right in an alphabet that appears nowhere else.[4] Yale has photographed every leaf and put the whole thing online for anyone to page through, which has done nothing to solve it and everything to spread the fascination.[1]

Whoever made it, made it a long time ago. Radiocarbon dating carried out at the University of Arizona in 2009 placed the vellum between about 1404 and 1438, the early fifteenth century, and a separate analysis of the inks and paints found them unremarkable for the period, azurite blue and red ochre with nothing suspicious in the mix.[3] The tests date the skin and the pigment, not the moment the words were set down, but they undercut the easy theory that the whole thing is a modern forgery.[3]

The name comes from the dealer who found it. Wilfrid Voynich, a Polish-born trader in rare books, bought the manuscript in 1912 from a Jesuit college near Rome, and with it came a letter written in the 1660s that is the earliest firm record of its travels.[4] That letter passes along a rumour that Emperor Rudolf II of Bohemia once paid six hundred gold ducats for the book, a story often repeated and never confirmed.[4] The manuscript reached Yale in 1969, and everything about its owners before Rudolf's court is guesswork.[2]

What makes it maddening is the pictures. The book falls into sections that scholars name for what they show: page after page of plants that match nothing in any real herbal, circular charts of stars and zodiac signs, and a strange run of small nude women standing in green pools joined by looping pipes.[2] There are folding cosmological diagrams, a pharmacy section of jars and clipped roots, and a closing stretch of dense text broken into short starred paragraphs like a book of recipes.[2] The drawings promise meaning everywhere and deliver it nowhere.

The writing itself only deepens the problem. Fed through a computer, it carries the statistical fingerprints of a genuine language, with common and rare words falling into the lopsided pattern natural languages show, and a 2013 study found the words clustered by topic the way meaningful text does.[6] Against that, some researchers argue the book is elaborate nonsense, and one showed that a simple sixteenth-century trick with a slotted card could spin out gibberish with the very same features.[4] Bold solutions surface every few years and collapse under scrutiny, among them a 2019 claim that the language was a lost early cousin of the Romance tongues, which the author's own university publicly backed away from.[7] A 2018 attempt to read it as scrambled Hebrew with the help of algorithms met the same sceptical fate.[5] For now the book keeps its secret, which may be the only honest thing anyone can say about it.[2]