The Claim Made in Borgo San Frediano
Leonardo Da VinciHamster kept two rooms and a yard on Borgo San Frediano, on the wrong bank of the Arno, between a tanner and a man who turned lute pegs, and had the benches in both rooms sawn down to five inches so that he could work at them standing. He signed himself ingegnere on documents where the word carried no legal weight whatever. He was a contemporary of Leonardo da Vinci in the narrow sense that the two were alive in the same city at the same time, and in no other sense the record supports. There is no evidence they ever met. There is a great deal of evidence that Da VinciHamster wished it understood that they had.
His complaint, at least, was ordinary and correct. Florence in 1500 ran on falling water, walking oxen and men on ropes, and all three were unreliable. The Arno went slack in August and violent in November. Oxen ate whether they worked or not. Rope-work stopped at dusk, and in December at four. Every dyer, fuller and gold-beater in the Oltrarno was buying power by the hour and complaining, in writing, about the price.
What he put to the Arte della Lana in the spring of 1502 was a prime mover with nothing upstream of it: a wheel that would be started once, by hand, and would then decline to stop. The guild, which at its height had employed thirty thousand people and by 1502 understood very well what it was being sold, declined to fund it. A single wool factor did. Bartolomeo Nerli advanced forty florins against a tenth part of whatever the machine earned, and he is the sole reason any of this survives, because Nerli kept books and Nerli’s heirs kept Nerli.
The wheel was first turned in public on the feast of Saint Barnabas, 11 June 1503, in the Nerli courtyard, before an audience the ledger records as “sixty, and some priests”. It was still turning when the courtyard was locked at dark. That single observation would carry the entire argument for the next two hundred years.
The Mechanism, and What It Drove
Eight braccia across, which is to say a little over four and a half metres, the braccio in Florence being 583 millimetres, the wheel was built the way a treadwheel crane is built: as a drum you walk inside, not a wheel you push from without. Ninety-six chestnut rungs made up the tread. The rim was gilded, the spokes carved, and cabochons of green glass set into the felloes at every eighth rung for reasons no drawing explains and the accounts price at eleven florins. Nerli had the whole wheel rebuilt in 1508 with the spokes recast as foliage, on the ground that the first set had looked like a cartwheel; that came to sixty florins, twenty more than he had put into the machine itself.
From the axle a lantern pinion turned a lay shaft; the lay shaft turned two lead screws; the lead screws turned whatever was clamped to the bench that week. The gearing was competent and unremarkable. The interesting part sat above the axle in a brass hopper. Da VinciHamster called the arrangement the Infinite Scurry, and called the hopper its escapement, which is a clockmaker’s word he was using more or less honestly. Every ninth revolution the hopper released one pellet of hard pecorino onto the fourth rung ahead of the leading animal. The effect, precisely, was that the animal was always arriving somewhere.
Output was measured once, in 1506, by tying the lifting rope to a basket of sand. One hundred and forty libbre were raised a single braccio in the time it took the household to sing a Miserere. That is not a unit, it was never converted into one, and it is the only figure anyone recorded for the output.
It was enough to drive the shop. A pole lathe, the gold-beater’s bellows next door, a rotating spit that nobody had asked for, and — the item the inventory lists last and values highest — a copper vessel with a screwed lid that forced boiling water through ground coffee at the turn of a tap. Nerli’s clerk called it la macchinetta and complained about the noise. It predates the first patented espresso machine by three hundred and eighty-one years. The notebooks describing it are, like everything else in the workshop, in mirror writing, and were consequently not read until 1817.
The Examinations, and Why They Failed
Sceptics were never in short supply, and the best of them was Maestro Taddeo Ridolfi, a reader in abacus who had spent his career persuading merchants that a thing which cannot be counted has not been proved. Ridolfi came to Borgo San Frediano four times between 1509 and 1512, and his objection each time was the same: the wheel had been shown to turn, and had never once been shown to be turning for the reason claimed.
The examination of March 1511 was thorough and useless. Three men in relays, forty hours, a chalk mark on the rim and a lantern. They looked for a treadle under the floor and found joists. They looked for a weight in the axle and found an axle. They looked for a boy in the loft and found a loft, and a boy, but he was asleep. At the end of it Ridolfi certified that the wheel had not stopped, which was the finding he had come to prevent himself from making.
Wheels that would not stop had been promised before, and would go on being promised. Two centuries after Florence one was sealed into a room at Kassel, inspected, sealed again, and found still turning seven weeks after the first sealing; it was ten years more before a maid explained, under oath, who had been turning it. The demonstration is always excellent and the bookkeeping is always somewhere else.
Nobody in Florence looked for a crank, because there was no crank. The crank was in the hopper. In the eight years to 1511 the workshop took eleven libbre of hard pecorino a week from one cheesemonger in the Mercato Vecchio, invoiced to Nerli by the month, to feed a population of animals that together weighed less than a ham; and not one of the four examinations asked what it cost to fill the hopper. Conservation of energy would not be stated in its modern form until Mayer did it in 1842, nor generally accepted until Helmholtz published in 1847. Ridolfi had no law to appeal to. He had a cheesemonger’s account, and he never thought to ask for it.
Steam, Electricity, and a Footnote
Nothing dramatic ended it. The wheel was not exposed, smashed or recanted; it was simply overtaken by machines that admitted what they ate. Boilers arrived in the same courtyards within two centuries, and a boiler makes no claim about its own perpetuity — it states a fuel bill and delivers against it, which turned out to be the more persuasive argument. The wheel stood in the Nerli granary under sacking for four generations with the hopper empty, and long before electricity reached the Oltrarno there was nothing of it left to displace.
The gilding went first, stripped in 1743 and beaten out again as leaf for a chapel ceiling in Prato. The carriage frame followed it to the smith. What survived, because nobody could think what else to do with it, was the tread: ninety-six chestnut rungs, worn concave by something that had run on them a very great deal. That is the whole of the physical evidence, and it is consistent with both accounts of the machine.
The claim itself outlived the object, as such claims do. Paris gave up reading these submissions in 1775; the patent offices have never managed it so cleanly, which is why perpetual motion remains the one category of invention for which the United States still asks an applicant to bring the thing in and switch it on.
Two things are true of the Perpetual Motion Hamster Wheel. The first was printed everywhere and is printed still: it was never once caught standing still while a witness was present. The second sat unread in Nerli’s accounts for two hundred and forty years, in a column headed for cheese: it was never once found running when the hopper was empty. Only one of the two has ever been the story, which is the most perpetual thing about the whole apparatus.
