Leonardo was back in Florence by the beginning of March 1503, hard-bitten from his months at the court of the brigand Duke. On 4 March he drew out 50 gold ducats from his account at Santa Maria Nuova. He records the transaction on a page which also has an enigmatic note: ‘Get the Gonfalonier to cancel the book and the Ser [i.e. notary] to provide me with a written record of moneys received.’64 The sense of ‘cancel’ is legalistic (Latin can-cellare, to cancel a deed with criss-cross lines, from cancellus, a grating). The note is probably connected with the undelivered altarpiece for the Santissima Annunziata, a closing of the accounts, in which case the ‘Ser’ may be Ser Piero da Vinci, who was notary to the Servites of the Annunziata. Is this how Leonardo thinks of his father – ‘il Ser’, the sir or sire? The Gonfalonier is Piero Soderini, who in late 1502 had been elected Gonfalonier of the Florentine republic, in effect its prime minister: an upstanding but unimaginative man, whose dealings with Leonardo would not always be amicable.
On 8 April Leonardo lent 4 gold ducats to ‘Vante miniatore’ – the miniature painter Attavante di Gabriello, a man much the same age as Leonardo, and probably an acquaintance from younger days. ‘Salai carried them to him and delivered them into his own hand: he said he will pay me back within the space of 40 days.’ On the same day Leonardo gave Salai some money for ‘rose-coloured stockings’ – the same colour as his own ‘tunic’ (pitocco), as recorded by the Anonimo Gaddiano: ‘He wore a short, rose-pink tunic, knee-length at a time when most people wore long gowns.’65 One gets a hint of the dandified air of the Leonardo circle, somewhat at odds with the mood of republican Florence. This glimpse of Leonardo’s garb was probably a memory of the Florentine painter whom the Anonimo calls Il Gavina or P. da Gavine.
On 14 June Leonardo withdrew another 50 gold florins from his account. He was perhaps in something of a parlous state: his future uncertain, his savings leaking away. We don’t know where he was living at this time: the doors of the Annunziata were no longer open to him. It is perhaps around now that someone, probably a pupil, writes on a page in the Codex Atlanticus, ‘Tra noi non ha a correre denari’ – ‘Round here the money’s pretty tight…’66
But a new and exciting project beckoned, its purpose nothing less than changing the course of the Arno river. There were in fact two separate projects: the diversion of the lower Arno, which was a purely military tactic intended to cut Pisa off from the sea, and the more grandiose scheme of canalizing the entire river west of Florence to make it navigable. Machiavelli, now Soderini’s right-hand man on military and political affairs, was closely involved in both of them. He had seen something of Leonardo’s work as Borgia’s engineer, and it may be that the Pisa diversion was first hatched up in conversations between them at Imola.
Pisa had been temporarily ceded to the French by Piero de’ Medici in 1494, and had declared independence when the French withdrew from Italy the following year. Florentine efforts to retake the city had been embarrassingly ineffectual: the Pisans could hold out indefinitely while they had food and supplies coming in via their port at the mouth of the Arno. Hence this new stratagem – logical but technically ambitious – of rerouting the river and thereby, in the phrase of one of Machiavelli’s assistants, ‘depriving the Pisans of their source of life’.67
On 19 June 1503 Florentine troops took possession of the fortress of La Verruca, or Verrucola, a knoll which can still be seen on the southern slopes of the Pisan hills, overlooking the flatlands of the lower Arno. Two days later Leonardo was there. The officer in charge, Pierfrancesco Tosinghi, reported back to Florence: ‘Leonardo da Vinci came in person, with some companions, and we showed him everything, and it seems to us that he found La Verrucola satisfactory, and that he liked what he saw, and afterwards he said he had thought of ways to make it impregnable.’68 This brisk dispatch catches Leonardo in action on a summer day in the Pisan hills: his thoroughness (‘we showed him everything’), his enthusiasm (‘he liked what he saw’).
A month later he was in the area again, as part of an official delegation led by Alessandro degli Albizzi. He was there by 22 July, when he dated a sketch of the lower Arno, and on the 23rd he took part in discussions at the Florentine battle-camp, probably at Riglione, as reported by a certain Captain Guiducci the following day:
Yesterday we received Alessandro degli Albizzi, bearing a letter from Your Lordship [Gonfalonier Soderini], together with Leonardo da Vinci and certain others. We studied the plan, and after many discussions and doubts it was concluded that the project was very much to the purpose, and if the Arno can really be turned or channelled at this point, this would at least prevent the hills from being attacked by the enemy.69
Also in the party was one Giovanni Piffero, whom the Signoria later reimbursed to the tune of 56 lire ‘for a carriage with six horses, and the cost of meals, going with Leonardo da Vinci to level the Arno around Pisa’. He is elsewhere named as Giovanni di Andrea Piffero, and was in all probability Giovanni di Andrea Cellini, father of the famous sculptor Benvenuto Cellini, who was indeed a member of the pifferi, or pipers, of the Signoria.70
Captain Guiducci’s dispatch of 24 July 1503 is the earliest specific mention of the Arno diversion, though the ‘plan’ which they studied and the authorizing letter from the Gonfalonier suggest that the project was well advanced on paper. The idea was to channel the river southward to an area of marshland near Livorno called the Stagno. The water would be diverted by weirs down a huge ditch, a mile long and 16 braccia (32 feet) deep, which would then fork into two smaller ditches. The distance between the river and the outflow at the Stagno diversion was about 12 miles, and Leonardo estimated that about a million tons of earth would have to be dug out for the ditches. In an early example of the time-and-motion study, he calculated how much earth a single worker could shift. Because the ditches would be so deep, he reckoned that a single bucket of earth from the bottom would be handled by fourteen workers before it reached the top. After some complex calculations he concluded that the entire project would require 54,000 man-days.71Alternatively, he says, ‘various machines’ could be used to speed things up. One such is the mechanical digger found in the Codex Atlanticus: measurements beside the drawing correspond exactly with those proposed for the Arno ditches, so it seems this machine was specifically designed for the job. Another computation reads, ‘One shovel-load is twenty-five pounds; six shovel-loads made up one barrow-load; twenty barrow-loads made up a cart-load.’72 These are the nuts and bolts of Leonardo’s work as a civil and military engineer.
Digging on the Arno–Stagno canal did not begin for over a year. Leonardo was not actively involved at this later stage: another engineer or maestro d’aqua, one Colombino, was in charge, and some of what was done differed from Leonardo’s plans. Work began on 20 August 1504 ‘at the tower of the Fagiano’, which was demolished to provide building materials for the weir. The project was a total failure, and was abandoned after two months, amid growing unrest about unpaid wages. A report by Machiavelli’s assistant Biagio Buonaccorsi sums up succinctly:
When the decision was finally made, a camp was established at Riglione, and maestri d’aqua were summoned. They said they required two thousand workers, and a quantity of wood to construct a weir to hold the river in, and divert it along two ditches to the Stagno, and they promised to complete the project within thirty or forty thousand man-days [significantly less than Leonardo’s estimate] and in this hope the project was begun on 20 August  with two thousand workmen hired at one carlino a day. In fact the project took much more time and money, and to no profit, for, in spite of these estimates, eighty thousand man-days were not enough to bring it even halfway… The waters never went through the ditches except when the river was in flood, and as soon as it subsided the water flowed back. The whole undertaking cost seven thousand ducats or more because as well as paying the workers it was necessary to maintain a thousand soldiers to protect the workers from attack by the Pisans.73
In early October 1504 disaster struck. There was a violent storm; several boats guarding the mouth of the ditches were wrecked, and eighty lives were lost. The walls of the ditches collapsed, and the whole plain was flooded, destroying many farms. By mid-October, less than two months after the work had begun, the project was abandoned. The Florentine army withdrew, the Pisans filled in the ditches, and the episode was quickly forgotten – another waste of time, money and lives in the long war with Pisa, which continued until 1509.
This abortive plan was part and parcel of Leonardo’s more grandiose dream: the construction of a long canal which would obviate the navigational problems of the Arno and give Florence its own access to the sea. The river below Florence is not navigable because of the rapids between Signa and Montelupo 10 miles west of the city, and is shallow, silted and meandering for much of its course between there and Pisa: one recalls Brunelleschi’s ill-fated Badalone beached on the sandbanks of Empoli. This grand canal is distinct from the military diversion project, but his involvement in the latter probably gave impetus to this larger canalization scheme, which he had been turning over in his mind for some time. The combined effect of them – had they worked – would have been to link Florence to the sea, thus enabling the city to participate in the great adventure of New World trade and exploration, excitement about which the Florentine navigator Amerigo di Vespucci did much to encourage in letters to his Italian patron Lorenzo di Pierfrancesco de’ Medici, a digest of which was published in Florence in 1504.
This gives a topical impetus to an idea that had been around for a while: an Arno canal had been proposed as early as 1347, and Leonardo himself had been thinking about it for ten years or more. There is a sheet of notes on the subject in the Codex Atlanticus which, as it mentions Pisa as a beneficiary of the scheme, appears to have been written before the beginning of Florentine–Pisan hostilities in 1495. He wrote:
Sluices should be made in the valley of La Chiana at Arezzo so that when in the summer the Arno lacks water the canal will not dry up. The canal should be 20 braccia wide at the bottom and 30 at the top, and 2 braccia deep, or 4 so that 2 braccia of water can serve the mills and meadows, which will benefit the area, and Prato, Pistoia and Pisa will gain 200,000 ducats a year, and will contribute labour and money to this useful project, and the Lucchese the same, because the Lake of Sesto will become navigable. I shall make it go through Prato and Pistoia, and cut through Serravalle and debouch into the lake, so there will be no need of locks.74
He costs this out at a rate of 1 lire per 60 square braccia, though this does not include any estimate for the immense work of ‘cutting through’ the mountain at Serravalle.
There are maps and plans which date from 1503–4 and show the proposed route of the canal. In the corner of one is a note reading, ‘They do not know why the Arno will never run in a straight channel: it is because the rivers which flow into it deposit earth where they enter, and take it from the opposite side, and so bend the river.’75 One hears the touch of testiness, the magisterial note which some would have found trying. ‘They’ do not know this elementary fact which he must now for the umpteenth time explain.
The canal remained a dream – the failure of the military diversion was perhaps a discouragement, though the immense costs of the project were doubtless enough to discourage the Signoria anyway. No work was carried out, though the A11 autostrada between Florence and Pisa covers much the same route as the proposed canal, so one might say that Leonardo’s vision was only wrong in so far as it failed to predict the demise of the waterway itself as a medium of transport. His imagination fell short of the internal-combustion engine.
These canalization projects, and others he later undertook in Milan, are another instance of Leonardo’s lifelong fascination with water, its currents, pressures, vortices and refractions, which he noted, analysed, sketched and indeed painted so beautifully – the sullen gleam of the padule behind Ginevra, the marine vistas of the Annunciation, the rock-eroding rivers that meander implacably behind the Madonna of the Yarnwinder and the Mona Lisa.